SAU, Alm.del - 2020-21 - Endeligt svar på spørgsmål 138: Spm. om Danske Banks gældsinddrivelsessag giver ministeren anledning til at undersøge, om inkassovirksomheder, som har inddrevet gæld på vegne af Danske Bank, ligeledes har gjort dette på et uretmæssigt grundlag, til justitsministeren, kopi til skatteministeren

Skatteudvalget 2020-21
SAU Alm.del
Offentligt

Danske Banks svar på Finanstilsynets brev med ref.nr. 6252-0349

FINANSTILSYNET

Århusgade 110

2100 København K

(Sendt via e-mail til

[email protected]

[email protected])

Danske Bank A/S

Holmens Kanal 2-12

DK-1092 København K

Tel. 33 44 00 00

danskebank.dk

10. september 2020

Svar på Fi a stilsy ets brev: A

J.nr. 6252-0349

od i g o

redegørelse o

Da ske Ba k A/S’ gældsi ddrivelsessyste

INDLEDNING

Danske Bank A/S (herefter Danske Bank eller banken) henviser til Finanstilsynets brev af 31. august 2020. I

dette brev anføres, at Finanstilsynet i lyset af de oplysninger, der på det seneste er kommet frem i medierne

om fejl i bankens inkassosystem, har besluttet at undersøge bankens håndtering af de konstaterede fejl i

bankens inkassosystem set i forhold til bankens forpligtelse til at handle redeligt og loyalt over for sine kunder

(jf. § 43 i lov om finansiel virksomhed og bekendtgørelsen om god skik for finansielle virksomheder).

Indledningsvis vil vi gerne understrege, at vi dybt beklager situationen og den usikkerhed, som vores fejl og

de fejlbehæftede data i bankens inkassosystem har skabt

–

ikke mindst for vores kunder. Kunderne såvel som

alle andre interessenter skal naturligvis kunne stole på de oplysninger, de får fra banken og som vores

rådgivning og aftaler er baseret på. Vi beklager meget, at dette ikke har været tilfældet i denne sag. Det er

udelukkende vores ansvar, og vi forpligter os til så hurtigt som muligt at gennemgå og korrigere alle de sager,

der er påvirket af datafejlene, samt sikre, at vores kunder modtager fuld kompensation for de beløb, de har

betalt for meget, eller andre relaterede tab som følge af vores fejl.

Ligeledes kan vi kun beklage over for både kunder, medarbejdere og alle andre interessenter, at disse

problemer ikke på et tidligere tidspunkt er blevet håndteret ordentligt. Der er tale om en kompleks sag, men

det står klart for os, at problemstillingen igennem årene delvist har været kendt i forskellige dele og på

forskellige niveauer af organisationen, herunder blandt ledere. På trods af forsøg på at afhjælpe

problemerne, blev der aldrig taget hånd om de underliggende datafejl, og desværre fik problemerne i denne

sag derfor fik lov til at eksistere gennem mange år.

Vi tager de forhold, der er nævnt i Tilsynets brev, meget alvorligt, og vi forpligter os til fortsat at have en åben

og transparent dialog med Finanstilsynet i forhold til de historiske og aktuelle forhold, der vedrører eller

SAU, Alm.del - 2020-21 - Endeligt svar på spørgsmål 138: Spm. om Danske Banks gældsinddrivelsessag giver ministeren anledning til at undersøge, om inkassovirksomheder, som har inddrevet gæld på vegne af Danske Bank, ligeledes har gjort dette på et uretmæssigt grundlag, til justitsministeren, kopi til skatteministeren

Danske Banks svar på Finanstilsynets brev med ref.nr. 6252-0349

påvirker bankens inkassosystemer. Ledelsen har altid ansvaret for at vide, hvad der foregår i virksomheden,

for at rette op på fejl, der er opstået tidligere, samt for at sikre, at vi har en ledelseskultur, hvor fejl og

problemer bringes frem i lyset og løses.

Hensynet til bankens kunder kommer i første række, og vi vil fortsat arbejde på at sikre holdbare løsninger.

Hvis vi i forbindelse med vores forsatte analysearbejde identificerer forhold, der påvirker oplysningerne i

denne redegørelse, vil vi naturligvis omgående orientere Finanstilsynet herom.

Vi vil gerne understrege, at vi vil kompensere alle de kunder, der er økonomisk påvirket af de fire

hovedårsager

”root auses”

, som vi beskriver i denne redegørelse. Sagen skyldes de beklagelige systemiske

fejl i bankens inkassoproces, der er konstateret i forbindelse med en grundig analyse af bankens

inkassosystemer. Den 10. september 2020 besluttede vi endvidere at tilbyde alle potentielt berørte kunder

at sætte tilbagebetalingen af deres gæld i bero uden omkostninger for dem.

Vi har allokeret og vil fortsat allokere betydelige ressourcer til arbejdet med at rette op på fejlene og

kompensere alle berørte kunder så hurtigt som muligt. De konstaterede systemiske datafejl beskrives mere

detaljeret i denne redegørelse. Vi har i forbindelse med vores analyser af inkassosagen og de fire

hovedårsager indtil videre identificeret en række yderligere faktiske eller potentielle problemstillinger. Disse

er anført i bilag 2.5. Vi vil arbejde videre med disse problemstillingerne, for at opnå en fuld forståelse af dem

og håndtere dem på en passende måde. Nogle af problemstillinger vedrører renter på rykkergebyrer,

fejlbehæftede oplysninger til brug i retssager, mæglersalærer og ukorrekt registrering af skattemæssige

oplysninger. Disse yderligere problemstillinger kan potentielt påvirke et større antal kunder, end dem, der er

påvirket af de fire hovedårsager, som er beskrevet i denne redegørelse.

Bankens direktion blev i maj 2019 orienteret om de systemiske datafejl, der påvirker bankens

inkassosystemer, og iværksatte som følge heraf foranstaltninger, for at afhjælpe konsekvenserne af

systemfejlene, som beskrevet nærmere i denne redegørelse.

BAGGRUND

Banken har i dag to inkassosystemer: i) Debt Collection System (DCS) og ii) Personlige Fordringer (PF).

•

I DCS registreres misligholdt gæld på almindelige bankprodukter som f.eks. kassekreditter, stående

lån, kautioner osv.

I PF-systemet registreres misligholdt gæld, der er opstået i forbindelse med kundens manglende

betalinger på realkreditlån ydet af Realkredit Danmark A/S, hvor pantet i ejendommen ikke har været

tilstrækkeligt til at indfri realkreditlånet.

DCS blev implementeret i 2004 med det formål at centralisere inkassoprocesserne på tværs af banken.

PF-systemet blev implementeret i banken i 1979.

Tidligere foregik inddrivelsesprocessen i de enkelte filialer, der havde et overblik over de lokale kunders

gældsforpligtelser og håndterede eventuelle misligholdte gældsforpligtelser via manuelle processer for at

sikre korrekt tilbagebetaling af gæld. Opkrævningen af inkassogæld før 2004 blev understøttet af PF-

systemet og flere ældre systemer (f.eks. papirarkiver, regneark og Lotus Notes) med individuelt tilpassede

processer og kontrolforanstaltninger. Inkassosystemerne før 2004 havde nogle iboende systemiske

svagheder, men en række manuelle processer for at afbøde effekten af sådanne svagheder var i langt de

fleste sager etableret.

Danske Banks svar på Finanstilsynets brev med ref.nr. 6252-0349

I 2004 blev data vedrørende al udestående misligholdt gæld til banken centraliseret og migreret til DCS. Efter

denne migrering af data blev alle nye inkassosager åbnet direkte i DCS (undtagen de sager, der havde deres

oprindelse i PF-systemet som anført ovenfor).

Da data blev migreret til DCS blev de fejl, der eksisterede i relation til visse kunders gæld, ligeledes overført.

Med tiden har man i banken i større og større omfang handlet i tillid til oplysningerne i DCS, de medarbejdere,

der har været involveret i inkassoprocessen, har haft mindre erfaring og der har i et mindre omfang været

anvendt fysiske kundefiler som referencepunkt. Alt dette har bidraget til, at der fortsat skete fejl i

håndteringen af visse kunders gæld. Som beskrevet nærmere i svaret på spørgsmål 1 har der historisk været

viden om visse problemer i banken, og banken har implementeret forskellige processer for at afhjælpe eller

afbøde konsekvenserne af fejlene. Til trods for eskalering af problemstillingen har banken ikke gjort nok for

rettidigt at håndtere problemet i sin helhed. En større ændring af forældelsesloven i 2008 besværliggjorde

endvidere bankens inkassoprocesser. Vi har i svaret på spørgsmål 2 nedenfor beskrevet de hovedårsager, der

har påvirket inkassosystemernes funktioner negativt og dermed også vores kunder.

I maj 2019 iværksatte banken en proces, der skulle sikre, at der blev foretaget en grundig undersøgelse af

alle aspekter af systemfejlene, herunder især med det formål at identificere alle kunder, der er blevet

påvirket økonomisk med henblik på at disse snarest muligt kan blive fuldt kompenseret.

Som nærmere beskrevet i svarene på spørgsmål 3 og 4 har banken fra og med juni 2019 taget en række tiltag

for at afbøde de negative virkninger af de systemiske fejl. Som anført i svaret på spørgsmål 5 nedenfor viser

bankens analyse, at i en gruppe på 106.000 kunder kan potentielt 10.000-15.000 kunder være økonomisk

påvirket. Systemfejlene er komplekse. På grund af kompleksiteten og de individuelle og skiftende forhold

skal alle potentielt berørte kunders gæld genberegnes manuelt. De første grupper af berørte kunder er blevet

kompenseret, og banken har lagt en plan for at sikre fuld kompensation af alle berørte kunder inden den 1.

juli 2021 som anført i svaret på spørgsmål 7.

Der henvises nedenfor til bankens svar på de enkelte spørgsmål, som Finanstilsynet har stillet:

FINANSTILSYNETS ANMODNING OM BANKENS REDEGØRELSE FOR FORLØBET

MED FEJLENE I GÆLDSINDDRIVELSESSYSTEMET

SPØRGSMÅL 1: HVORNÅR KONSTATEREDE BANKEN FEJLENE FØRSTE GANG?

Medarbejdere på forskellige niveauer i organisationen, herunder ledere, har på forskellige tidspunkter haft

et større eller mindre kendskab til problemerne i inkassosystemerne.

Som nævnt ovenfor blev bankens direktion i maj 2019 informeret om de systemiske fejl, der påvirker bankens

inkassosystem. Nedenfor gives et overblik over kendskabet til de systemiske datafejl i banken.

Inkassosystemet (Debt Collection System - DCS)

Den planlagte implementering af DCS foregik i perioden frem til 2004. Det er bankens opfattelse, at der i et

vist omfang var kendskab til, at der var nogle iboende fejl i de data, der blev migreret til DCS. Derfor var

banken allerede fra implementeringen af DCS i 2004 opmærksom på, at der kunne være datafejl i DCS.

Danske Banks svar på Finanstilsynets brev med ref.nr. 6252-0349

Det er bankens opfattelse, at der i flere år efter implementeringen af DCS i september 2004 var et relativt

højt niveau af manuel håndtering af kundefilerne. Den manuelle håndtering betød, at data i DCS, f.eks. data

om hovedstolen på en kundes gæld, i praksis normalt blev kontrolleret ved at sammenholde med

oplysningerne om gælden i den fysiske kundefil. Dette betød, at data normalt blev tilrettet, før fejlene

påvirkede kunderne. Det er foreløbigt ikke klart for banken, hvor effektiv denne manuelle håndtering var.

Det er imidlertid bankens opfattelse, at der blev gjort en indsats for at sikre, at de fejlbehæftede data i DCS

normalt blev kontrolleret ved at sammenholde med kundefilerne, da man var opmærksom på de systemiske

datafejl i DCS.

I 2007 blev der etableret et "korrektionsteam" med det formål løbende at korrigere eventuelt fejlbehæftede

sager manuelt i DCS.

Bankens interne revision havde en række bemærkninger i årene 2005-2016, hvoraf nogle var relateret til de

problemer, som banken nu står overfor. Vi anerkender, at vi ikke gjorde nok for at håndtere hovedårsagerne

til revisionens bemærkninger fuldt ud eller fulgte op på handlingsplanerne, som opsummeret nedenfor:

I de første to år efter implementeringen af DCS blev der udarbejdet to interne revisionsrapporter (fremsendt

i januar 2005 samt i januar og marts 2006), hvoraf det fremgik, at der var problemer med datakvaliteten i

DCS og behov for væsentlige korrektioner, og det blev anbefalet, at inkassoafdelingen (Debt Collection

Service) foretog en gennemgang af samtlige sager med forkerte registreringer, og at der udførtes periodiske

kontroller for at sikre, at problemerne ikke var vedvarende.

Revisionsrapporter fra 2010 (fire udførte revisioner) og 2011 (to udførte revisioner) identificerede en række

problemstillinger i Group Recovery Debt Management (GRDM) i form af manglende eller utilstrækkelige

kontroller og processer, forskelle i de registrerede forældelsesfrister for krav samt ufuldstændige eller ikke-

behandlede sager om misligholdt gæld. I 2012 konstaterede revisionen, at GDRM ikke behandlede

udestående gæld korrekt, at datagennemgange og beregninger af nedskrivninger på udlån var

utilstrækkelige, og at renteberegninger var ukorrekte.

Der blev ved revisionen af området i 2015 og 2016 ikke konstateret problemer i det omfang, som nu er

konstateret.

En planlagt revision af området i 2019 blev ikke gennemført på grund af den nu igangværende interne

undersøgelse.

I perioden 2006-2016 overvejede man at gennemføre visse gældskampagner, hvor visse debitorer blev

tilbudt et betydeligt nedslag i deres gæld. Fire kampagner blev sendt til godkendelse i banken i 2007, 2009,

2014 og 2016, og selv om det er bankens opfattelse, at det ikke var alle kampagnerne, som blev gennemført,

var der, i det mindste i forbindelse med nogle af kampagnerne, en opfattelse i banken af, at kampagnerne

ville reducere omfanget af det manuelle arbejde med at genberegne de pågældende kunders gæld og

reducere risikoen for at kravene blev forældet.

I 2008 blev forældelsesloven ændret i Danmark, hvilket betød, at hovedstolen på lån generelt havde en

forældelsesfrist på 10 år, og at renter og omkostninger havde en forældelsesfrist på 3 år (med undtagelse af

omkostninger tilkendt af retten). Tidligere var forældelsesfristen henholdsvis 20 år og 5 år. Desuden gjorde

de nye regler det mere vanskeligt at suspendere forældelsesfristen. Disse nye forældelsesregler forværrede

problemerne med de fejlbehæftede data i DCS.

I 2009 lanceredes et initiativ kaldet LEAN. LEAN havde til formål at strømline eksisterende

forretningsprocesser i et forsøg på at blive mere effektiv i sagsbehandlingen. Bankens øgede fokus på LEAN

Danske Banks svar på Finanstilsynets brev med ref.nr. 6252-0349

betød, at medarbejderne i inkassoområdet var under et stigende pres for at handle i tillid til data i DCS i

forbindelse med gældsinddrivelse, frem for at bruge tid på at sammenholde data med fysiske kundefiler.

Overgangen til i højere grad at forlade sig på data i DCS betød, at den tidligere foretagne manuelle kontrol af

de fejlbehæftede data i DCS, der (i det mindste til en vis grad) reducerede risikoen for, at datafejlene i sidste

ende påvirkede kunderne, nu blev foretaget i langt mindre udstrækning.

Ved udgangen af 2016 nedlagde banken korrektionsteamet, idet banken på det tidspunkt var af den

opfattelse, at korrektionsteamet havde gennemført korrektionen af de gamle sager. Der var en opfattelse af,

at "nye sager", dvs. sager oprettet efter den 27. september 2004, hvor DCS blev implementeret, var korrekte.

På trods af denne opfattelse blev der over tid identificeret sager med fejl. I interne eskaleringer i 2016 og

2017 blev det oplyst, at også "nye sager" kunne være fejlbehæftede. Det blev i den forbindelse foreslået at

oprette e task for e til hå dteri g af pro le er ed " ye sager” ed

ukorrekte data, da banken

–

efter

egen opfattelse

–

havde korrigeret de gamle sager i 2016. Uanset dette forslag blev der ikke tildelt nye

ressourcer til en nyt task force (et nyt korrektionsteam).

I perioden fra marts 2018 til september 2018 nåede den nye ledelse i inkassoafdelingens juridiske funktion

en forståelse for, at der var sager, som fortsat krævede manuel korrektion. I september 2018 blev den øverste

ledelse i GRDM gjort opmærksom på problematikken med de fortsatte fejl i DCS. GRDM undersøgte sagen

og så på mulige it-løsninger. Ledelsen i GRDM anerkendte også, at datafejlene i DCS kunne påvirke bankens

planer om yderligere outsourcing af inkassoprocessen. Det er bankens opfattelse, at potentielle bekymringer

måske (i december 2018) blev videregivet til en Sponsor Komite i banken, der havde ansvaret for arbejdet

med at forbedre bankens processer for udenretlig inkasso. Det ser ikke ud til, at arten af de underliggende

datafejl blev drøftet på mødet.

I januar 2019 blev der udarbejdet et notat, der beskrev problemerne med datakvaliteten i DCS. I marts 2019

konkluderede ledelsen i GRDM, at problematikken med fejlene i DCS krævede eskalering, og man

informerede den øverste ledelse i COO-organisationen i marts 2019. Banken gennemførte herefter en

risikovurdering, der førte til indberetning af problemer med datakvaliteten i DCS via bankens operationelle

risikoinformationssystem (Operational Risk Information System - ORIS) den 24. maj 2019. I indberetningen

blev det anført, at man havde identificeret, at DCS havde visse problemer med datakvaliteten, og at de

iboende fejl i inkassoprocessen ikke blot relaterede sig til manuelle fejl. ORIS er et værktøj til den formelle

registrering og håndtering af operationelle risikohændelser, og alt efter den enkelte sags omfang og alvor

udløser det en eskaleringsproces internt i banken, herunder krav om at involvere Finanstilsynet i mere

alvorlige tilfælde. Hændelsens omfang og alvor blev vurderet som værende tilstrækkelig til at udløse både

intern eskalering til direktionen og bestyrelsen og orientering af Finanstilsynet. Finanstilsynet blev skriftligt

underrettet herom den 6. juni 2019. Direktionen og bestyrelsen blev ligeledes informeret den 6. juni 2019.

I ORIS-indberetningen til Finanstilsynet blev det oplyst, at banken havde oprettet et projekt, der skulle

fokusere på:

•

Årsagsanalyser

”root ause a alyses”

, der skulle foretages for at sikre, at fejlene ikke opstod i alle

nye inkassosager;

Oprydning i data i porteføljen (inkl. en gennemgang af alle sager og tilbagebetaling af inddrevne

beløb, som vi ikke var berettiget til)

Implementering af et nyt system og nye procedurer og

Kommunikation til kunderne og tilbagebetaling til dem hurtigst muligt

Danske Banks svar på Finanstilsynets brev med ref.nr. 6252-0349

Det oprettede projekt blev kaldt "Data Quality Project". I Data Quality Project blev der truffet forskellige

foranstaltninger for at analysere og håndtere konstaterede problemstillinger. Eksterne advokater og

konsulenter blev hyret for at bistå med at identificere, forstå og afhjælpe problemerne. Data Quality Project

blev efterfølgende, mere præcist den 4. november 2019, omdannet til "Programme Athens" efter en

præsentation fra Plesner Advokatpartnerselskab (herefter Plesner) den 29. oktober 2019, hvor det blev klart

for direktionen, at problemerne med datafejlene var mere systemiske og havde større konsekvenser end

tidligere antaget. Vi henviser til svarene på spørgsmål 3 og 4 vedrørende de forskellige foranstaltninger og

tiltag, der blev taget i regi af Data Quality Project og Programme Athens.

Personlige Fordringer (PF)

GRDM konstaterede den 4. juli 2019, at der var indtruffet en ikke-finansiel hændelse og foretog derefter

registrering i ORIS. Det blev opdaget, at de kendte problemer med datakvaliteten i et vist omfang var

systembaserede snarere end at der var tale om menneskelige og manuelle fejl. PF-systemet havde

utilstrækkelig funktionalitet, der resulterede i dårlig datakvalitet.

PF-systemet har eksisteret siden 1979. Der har været problemer med systemet over en lang periode.

Fejlene og de systemiske mangler i data blev opdaget den 4. juli 2019.

Siden den 4. juli er de risici, der er forbundet med PF-systemet, blevet varetaget af Danske Banks såkaldte

Data Quality Programme, der blev etableret for at håndtere og afhjælpe de identificerede risici i DCS-

systemet.

SPØRGSMÅL 2: HVAD BESTÅR FEJLENE KONKRET I, OG HVORDAN HAR DET PÅVIRKET KUNDERNE?

De konstaterede systemiske fejl i data skyldes fejlbehæftet migrering af data, manglende afstemning af data

og forkert brug af bankens inkassosystemer.

Som følge af et utilstrækkelig governance-setup i GRDM har de systemiske fejl i data resulteret i et stort antal

fejl i inkassosager samt utilstrækkelig dokumentation til brug for revisionen af området. GRDM havde en

utilstrækkelig governance og kontroller til i alle tilfælde at kunne identificere fejlene i gældsopkrævning og

korrigere dem.

De systemiske datafejl er inddelt i fire kategorier

–

eller

”root auses”.

Disse hovedårsager danner

udgangspunktet for bankens arbejde med at analysere data og korrigere fejl. Vedhæftet dette svar er:

•

Draft System and Process Flaw Document (udkast til dokument om system- og procesfejl af 28.

november 2019), bilag 2.1, som giver en mere detaljeret beskrivelse af de systemiske datafejl, der er

konstateret i DCS.

•

En intern præsentation af 26. november 2019, bilag 2.2, som giver en oversigt over de systemiske

datafejl i DCS

Et notat om hovedårsagerne og Danske Banks forpligtelse til at yde kunderne erstatning, bilag 2.3.

Dataanalyse og overvejelser i forbindelse med inkasso, bilag 2.4

Danske Banks svar på Finanstilsynets brev med ref.nr. 6252-0349

2.1 Hovedårsag 1: Hovedstol, renter og gebyrer slået sammen

Berørte systemer: DCS & PF

Generelt består kundens gæld af en række forskellige fordringer, herunder hovedstol, renter, omkostninger

og gebyrer.

De systemiske datafejl beskrevet under hovedårsag 1 opstår som følge af, at påløbne renter, gebyrer og

omkostninger fejlagtigt er lagt sammen til ét beløb, der er registreret som hovedstolen. Konsekvensen af de

problemer, der er kategoriseret under hovedårsag 1 er, at systemerne ikke kan skelne mellem de forskellige

typer af fordringer, selvom der gælder forskellige forældelsesfrister for disse krav. Dertil kommer, at der blev

pålagt rente på rykkergebyrer forud for en domstolsafgørelse.

Den potentielle effekt af at pålægge renter på rykkergebyrer er ikke medtaget i beregningerne vedrørende

hovedårsag 1. En separat arbejdsgruppe er i gang med at undersøge forholdene yderligere.

Hvordan har problemerne kategoriseret under hovedårsag 1 påvirket kunderne?

Hovedårsag 1 gør det vanskeligt at identificere de forskellige forældelsesfrister for henholdsvis hovedstol,

renter, gebyrer og omkostninger og kan resultere i en unøjagtig beregning af restgælden, renter osv.

Hovedårsag 1 kan resultere i

•

inddrivelse af forældede krav

fejlbehæftet indberetning til skattemyndighederne i relation til rentefradrag og

opkrævning af mere gæld end det faktisk skyldte beløb.

2.2 Hovedårsag 2: Forkert registrering af forfaldsdatoen

Berørte systemer: DCS

Forældelsesfristen regnes fra den dato gælden forfalder til betaling. Ved overførslen af en fordring til DCS

registreres forældelsesfristen imidlertid til datoen for overførslen til systemet i stedet for den faktiske

forfaldsdato. Registreringen medfører, at forældelsesfristen fremstår længere, end den faktisk er.

Dette betyder, at det var vanskeligt ved brug af DCS at identificere og kontrollere, på hvilket tidspunkt

forældelsesfristen skulle afbrydes for at forhindre, at en fordring blev forældet.

Hvordan har problemerne kategoriseret under hovedårsag 2 påvirket kunderne?

Banken kan have opkrævet gæld, der på tidspunktet for inddrivelsen var forældet,

•

Kunder kan have indgået aftaler om at afvikle gæld, der helt eller delvist var forældet,

Banken kan have indledt uretmæssig retsforfølgelse af kunder og

Banken kan have inkasseret gæld, der var forældet.

Danske Banks svar på Finanstilsynets brev med ref.nr. 6252-0349

Der er en øget risiko for, at frivillige forlig eller retsafgørelser, der er indgået eller afsagt på eller omkring det

tidspunkt, hvor forældelsesfristen udløb som registreret i DCS, rent faktisk allerede var forældet på

tidspunktet og derfor ikke burde være blevet opkrævet af banken.

2.3 Hovedårsag 3: Kautionister og meddebitorer behandlet ens

Berørte systemer: DCS

Hovedårsag 3 er en konsekvens af en fejlagtig og forkert migrering af data tilbage i 2004.

Forud for implementeringen af DCS i 2004 havde Danske Bank en række decentrale inkassoafdelinger og -

systemer.

I 2001 overtog Danske Bank BG Bank, og dermed også BG Banks gældsportefølje. I BG Bank var der visse

iboende svagheder i det gamle inkassosystem. I det gamle system kunne man ikke ved registreringen af

kautionister og meddebitorer i systemet teknisk set skelne mellem en kautionist og en meddebitor. BG Bank

havde imidlertid og før overtagelsen etableret manuelle procedurer til håndtering af langt de fleste af de

iboende svagheder i det gamle system.

Set ud fra et dataperspektiv medførte disse systemiske fejl en række udfordringer med datakvaliteten i

systemerne, og fejlene kunne i sidste ende resultere i en ukorrekt beregning af gælden. Disse interne

problemer blev migreret med over i DCS, da eksisterende kundesager blev overført til DCS i 2004. Disse

systemiske datafejl påvirker kun BG Bank-sager fra før 2004 som følge af migreringen. De sager, der stammer

fra denne migrering, kan dog stadig være påvirket af de problemer med datakvaliteten, som man tidligere

via manuelle processer havde forsøgt at imødekomme.

Da oplysningerne fra BG Bank blev overført til DCS, blev alle debitorer, meddebitorer og kautionister

behandlet ens, og det var ikke muligt i DCS at knytte meddebitorer eller kautionister til den gæld, der var

kautioneret for.

Hvordan har problemerne kategoriseret under hovedårsag 3 påvirket kunderne?

Banken kan have opkrævet den fulde gæld fra alle debitorer, meddebitorer og kautionister. Banken kan

derfor have opkrævet den samme gæld mere end en gang.

•

Hvis der f.eks. ikke blev foretaget en manuel justering i DCS ville banken, på trods af at der var indgået

et frivilligt forlig med debitor, fortsat forsøge at inddrive det fulde beløb fra en kautionist og/eller en

meddebitor.

2.4 Hovedårsag 4: Den fulde hovedstol blev opkrævet hos meddebitorer

Berørte systemer: DCS

Hovedårsag 4 er tilsvarende en konsekvens af en ukorrekt migrering af data i 2004.

Danske Banks svar på Finanstilsynets brev med ref.nr. 6252-0349

Forud for implementeringen af DCS var det grundet de begrænsninger, der var i det gamle system,

nødvendigt at åbne separate konti for hovedskyldneren og eventuelt den eller de tilknyttede kautionist(er),

hvor hver enkelt konto viste det fulde beløb som gæld.

Den fulde gæld blev derfor registreret på mere end en konto.

I henhold til bankens forretningsgange skulle debitorkonti justeres manuelt efter, at en kunde havde betalt

et afdrag, eller at der var tilskrevet renter, påløbet omkostninger eller lignende. Der skulle foretages en

tilsvarende justering på konti tilhørende meddebitorer eller kautionister.

Da de eksisterende konti blev overført til DCS, kunne kautionisterne og meddebitorer ikke længere

identificeres i systemerne. Det betød, at hvis en debitor betalte en gæld eller indgik et frivilligt forlig med en

afviklingsordning, så ville der ikke blive foretaget en tilsvarende justering på den pågældende kautionists

og/eller meddebitors konto.

Hvordan har problemerne kategoriseret under hovedårsag 4 påvirket kunderne?

Banken kan have opkrævet den samme gæld mere end en gang fra hver kautionist, meddebitor og

hovedskyldneren.

•

F.eks. kan en kunde have indfriet sin gæld, men gælden er stadig registreret hos meddebitor og/eller

kautionist.

Yderligere problemstillinger

Desuden har banken i løbet af Data Quality Project og Programme Athens identificeret en række yderligere

faktiske eller potentielle problemstillinger. Oplysninger om disse forhold fremgår af bilag 2.5. Det skal

bemærkes, at disse problemstillinger og antallet af berørte kunder fortsat undersøges, og der er behov for

yderligere analyser. Der kan derfor være kunder, der er berørt af de faktiske eller potentielle

problemstillinger, der er anført i bilag 2.5, men som ikke indgår i beregningerne i denne redegørelse. Vi er i

gang med at undersøge problemstillingerne og vil tage de nødvendige tiltag for at rette op på problemerne.

SPØRGSMÅL 3: HVORDAN HÅNDTEREDE BANKEN INFORMATIONEN OM DE KONSTATEREDE FEJL,

HERUNDER HVILKE TILTAG, BLEV DER IVÆRKSAT OG HVORNÅR SAMT HVORDAN MAN VIL UNDGÅ NYE FEJL

FREMADRETTET?

Der er et vist overlap mellem spørgsmål 3 og 4 med hensyn til de tiltag, vi har iværksat, da vi blev

opmærksomme på systemfejlene.

Vi har i besvarelsen af dette spørgsmål 3 fokuseret på det projekt, der i perioden efter ORIS-indberetningen

blev etableret for at afhjælpe problemerne, og de governance-rammer, der blev etableret i forbindelse med

Programme Athens.

Vi henviser til vores svar på spørgsmål 4, der indeholder en liste over de forskellige foranstaltninger og tiltag,

der blev iværksat, da det stod klart for direktionen, at problemerne med datafejlene var mere systemiske og

havde mere alvorlige konsekvenser end det tidligere havde været opfattelsen.

Danske Banks svar på Finanstilsynets brev med ref.nr. 6252-0349

3.1 Iværksatte foranstaltninger og tiltag

Efter ORIS-indberetningen den 24. maj 2019 blev der oprettet en operational task force bestående af

teamleads og lokale specialister og projektet gik under navnet "Data Quality Project" som nævnt ovenfor.

Parallelt hermed har Group Risk Management og Compliance gennemført en række workshops med henblik

på at afdække og analysere problemerne nærmere og for bedre at forstå kompleksiteten i sagen.

Resultaterne blev overdraget til advokatfirmaet Plesner, der blev engageret for at gennemføre en ekstern

advokatundersøgelse med det formål at identificere de væsentligste årsager til, at Danske Bank inddrev gæld

på et uberettiget grundlag.

Desuden blev der omkring dette tidspunkt iværksat en række forebyggende foranstaltninger for at mindske

risikoen for uretmæssig inddrivelse af gæld. Der henvises til spørgsmål 4 for yderligere oplysninger.

Den 27. september 2019 instruerede banken Plesner om at indlede en undersøgelse af datafejlene og de

potentielle problemer og konsekvenser, der opstod som følge af datafejlene, samt om at bistå med det

projektet, der var etableret for at afhjælpe problemerne.

Den 24. oktober 2019 fremlagde Plesner sine foreløbige resultater og redegjorde for visse potentielle

problemer over for Data Quality Project. Præsentationen fra telefonmødet er vedlagt som bilag 3.3. På

grundlag af de foreløbige resultater og de konstaterede mulige problemer anbefalede Plesner, at der blev

foretaget yderligere analyser, og det blev aftalt, at de konstaterede potentielle systemfejl skulle undersøges

nærmere og især, at det skulle verificeres, om konkrete kundesager i DCS rent faktisk var berørt af de

potentielle problemer og i hvilket omfang. Plesners efterfølgende arbejde har især resulteret i de to notater,

der forklarer de konstaterede systemfejl og som er vedhæftet som bilag 2.1 og 2.3.

Som et resultat af de konklusioner Plesner fremlagde den 29. oktober 2019 stod det klart for direktionen, at

problemerne med datafejlene var mere systemiske og havde mere alvorlige konsekvenser, end det tidligere

havde været opfattelsen. På grund af den nye opfattelse af problemets alvor blev der to dage senere, den

31. oktober 2019, foretaget en ny indberetning til Finanstilsynet.

Som en konsekvens af, at der var konstateret et alvorligt problem, oprettede Danske Bank den 4. november

2019 Programme Athens for at skabe en formel governance-struktur. Der henvises til bilag 3.4, som

indeholder en oversigt over den etablerede governance-struktur.

Den 11. november 2019 bad banken revisions- og konsulentvirksomheden Ernst & Young (EY)om at hjælpe

med Programme Athens og med at analysere og verificere de systemfejl, der var konstateret af Plesner.

Desuden har EY hjulpet med at kvantificere effekten af de potentielle fejl - både med hensyn til antallet af

berørte kunder, men også med hensyn til de økonomiske konsekvenser.

SPØRGSMÅL 4: HVORDAN SIKREDE BANKEN SIG FRA DET TIDSPUNKT, FEJLENE BLEV KONSTATERET, AT

TILSVARENDE FEJLAGTIGE INDDRIVELSER FRA DETTE TIDSPUNKT IKKE LÆNGERE KUNNE FINDE STED?

Ud over det projekt, der blev oprettet efter ORIS-indberetningen for at afhjælpe problemerne, har banken

implementeret følgende foranstaltninger og tiltag:

4.1 Indførelse af nye manuelle kontroller vedrørende genberegning

Den 17. juni 2019 blev det besluttet at iværksætte en række yderligere forebyggende foranstaltninger i form

af følgende nye manuelle kontroller:

Danske Banks svar på Finanstilsynets brev med ref.nr. 6252-0349

1) ingen nye inkassosager må oprettes fra den dato, uden at der først er foretaget en manuel beregning.

Derfor skulle alle nye sager efter denne dato behandles i henhold til den opdaterede forretningsgang,

der dækker afhjælpningen af de væsentligste problemer;

2) ingen nye sager kan indbringes for retten, uden at der først er foretaget genberegning, og

3) banken vil ikke afslutte igangværende sager (dvs. sager, der allerede er oprettet i DCS inden den 17.

juni 2019), f.eks. i tilfælde af ønske om ekstraordinær tilbagebetaling, uden at der først foretages en

manuel genberegning for at sikre, at kunden ikke har betalt banken mere, end banken har krav på.

4.2 Korrektionsteam oprettet

Det nye krav om manuel kontrol resulterede i en øget arbejdsbyrde. Der blev derfor oprettet et

korrektionsteam den 17. juni 2019 til at korrigere data i alle nye sager. Det oprindelige team på 4

medarbejdere blev efterfølgende udvidet med 21 medarbejdere for yderligere at fremskynde

genberegningsprocessen. EY blev også hyret til at bistå med genberegningen af sagerne.

Pr. 9. september 2020 er i alt 6.250 sager blevet genberegnet af korrektionsteamet.

4.3 Udbedring af DCS

I juli 2019 blev der foretaget en ændring i kodningen af DCS for at sikre, at rykkergebyrer ikke længere bliver

indsat i et datafelt i DCS, uden at forældelsesfristen er korrekt. Dette blev gjort for at forhindre, at denne

systemfejl resulterede i uretmæssig inddrivelse af gæld i nye sager.

4.4 Ophør af igangværende retssager

Ud over den nye manuelle kontrol, der foretages med henblik på at forhindre, at nye sager bliver indbragt

for retten, før der er foretaget en ny beregning, blev det den 17. december 2019 besluttet at trække alle

igangværende retssager tilbage eller genberegne/korrigere dem, herunder sager vedrørende gældssanering

og konkursboer. Korrektionsteamet har genberegnet de fleste af de igangværende sager, og alle

igangværende sager, hvor domstolen endnu ikke havde truffet afgørelse, blev trukket tilbage fra domstolene

i februar 2020.

4.5 Beslutning om at betale kompensation til kunderne

Styregruppen har besluttet at kompensere alle kunder for deres tab som følge af de konstaterede årsager.

Der henvises til vores svar på spørgsmål 7 for yderligere oplysninger om det antal af kunder, der på

nuværende tidspunkt har modtaget kompensation, og den aktuelle plan for yderligere udbetaling af

kompensation.

4.6. Beslutning om at tilbyde alle kunder at sætte deres tilbagebetaling af gæld i bero

For yderligere at reducere risikoen for overopkrævning sætter Danske Bank omkring 17.000 kunders

inkassosager i bero, indtil de er blevet genberegnet. Det sker i forlængelse af den igangværende indsats for

at rette op på de fejl, der er konstateret i vores gældsinddrivelsessystem.

Der er tale om de sager, hvor der er afdraget 60 pct. eller mere af hovedstolen, og hvor der derfor er en

højere risiko for, at kunderne bliver opkrævet for meget, inden sagerne når at blive genberegnet. Inddrivelsen

genoptages, når den enkelte sag er gennemgået, og eventuelle fejl er rettet. Herudover vil der desuden ikke

blive tilskrevet renter i den periode, hvor sagen er sat i bero.

Danske Banks svar på Finanstilsynets brev med ref.nr. 6252-0349

Inddrivelsen vil fortsætte for de 35.000 kunder, som har påbegyndt tilbagebetaling, men har tilbagebetalt

mindre end 60 pct. af den oprindelige hovedstol, og hvor risikoen for overopkrævning dermed er meget lille.

Selvom inddrivelsen opretholdes, vil der heller ikke i disse sager blive tilskrevet renter, før en genberegning

har fundet sted.

For at sikre størst mulig valgfrihed kan kunder, hvis inkassosager sættes i bero, vælge at fortsætte betalingen.

Ligeledes kan kunder, hvor inddrivelsen som hovedregel opretholdes, vælge at sætte betalingerne i bero,

indtil sagen er genberegnet.

Ændringerne vil træde i kraft med udgangen af denne måned.

4.7. Plan for indførelse af yderligere tekniske sikkerhedsforanstaltninger og forbedringer

Banken er i øjeblikket i gang med en plan for it-implementering, som omfatter en række tekniske

sikkerhedsforanstaltninger og forbedringer af de pågældende it-systemer med henblik på at styrke de

eksisterende kontroller samt implementere yderligere kontroller.

4.8 Vurdering af beregninger i DCS og PF

Efter implementeringen af systemforbedringerne blev de to systemer, DCS og PF, vurderet af EY, og det

blev bekræftet, at beregningerne sker korrekt baseret på korrekt datainput og korrekt anvendelse af

funktionaliteten.

4.9 Fremadrettede tiltag - nyt it-system indkøbt

Banken havde allerede inden Programme Athens taget skridt til at købe et nyt it-system til erstatning af DCS.

Banken har nu købt et nyt system, der er planlagt til at erstatte DCS.

SPØRGSMÅL 5: HVOR MANGE KUNDER ER BERØRT AF FEJLENE?

Som anført ovenfor er der to særskilte inkassosystemer, som er blevet påvirket af de fire konstaterede

hovedårsager, nemlig DCS- og PF-systemerne. For de fleste af de berørte kunder er deres sag behandlet enten

i DCS eller PF-systemet, mens et mindre antal kunder er behandlet af begge systemer. Dette svar beskriver

vores aktuelle estimat af det samlede antal kunder, der er påvirket på tværs af begge systemer.

Det skal bemærkes, at svarene i denne redegørelse omhandler det antal kunder, der er påvirket af en eller

flere af de hovedårsagerne. Redegørelsen vedrører ikke antallet af kunder, der kan være påvirket af de

yderligere problemstillinger, som banken blev opmærksom på i forbindelse med årsagsanalysen (beskrevet i

bilag 2.5).

Der henvises i øvrigt til EY rapporten i bilag 2.4.

5.1 Det samlede antal potentielt berørte kunder

Det samlede antal potentielt berørte kunder på tværs af både DCS- og PF-systemet, som kan have betalt for

meget tilbage på deres gæld, er 106.000. Der henvises til

figur 1

nedenfor.

402.000 kunder (ca. 600.000 konti) er behandlet i DCS- og PF-systemerne, og der er talt om både privat- og

erhvervskunder. Det samlede antal potentielt berørte kunder omfatter både sager, der stadig er aktive DCS,

samt sager, der er afsluttet (fordi gælden er tilbagebetalt, eller sagen på anden måde er lukket af Danske

Bank) i perioden mellem 2004 og nu.

Danske Banks svar på Finanstilsynets brev med ref.nr. 6252-0349

Af de i alt 402.000 kunder blev 333.000 behandlet i DCS, 49.000 i PF-systemet, og de resterende 20.000 i

begge systemer.

Ifølge nedenstående kategorisering er der i alt 106.000 kunder, som muligvis er berørt af fejlene, og som

risikerer at skulle betale eller have betalt for meget tilbage til banken. Af hensyn til overskueligheden skal det

bemærkes, at dette omfatter både kunder, der vil modtage en økonomisk kompensation efter

gennemgangen af deres sag, og kunder, hvor godtgørelsen vil blive modregnet i anden gæld til Danske Bank.

5.2 Kategorisering

Kunder i andre jurisdiktioner

Af de 333.000 kunder, der udelukkende blev behandlet i DCS, var 152.000 kunder en del af gældsporteføljer

i andre jurisdiktioner (herunder Norge, Sverige, Finland og Storbritannien). Da DCS også anvendes i andre

lande end Danmark blev der foretaget en intern undersøgelse for at få bekræftet, om de identificerede

hovedårsager også berørte inkassokunder i de andre lande. En kombination af afbødende foranstaltninger,

herunder manuel afstemning, software-løsninger og nationale forældelsesfrister har betydet, at de

konstaterede systemfejl kun har berørt kunder i Danmark.

PF-systemet blev ikke anvendt til behandling af kunder i lande uden for Danmark.

Kunder, der ikke har foretaget betalinger

105.000 kunder af de resterende kunder havde ikke betalt af på deres gæld, da de blev overført til DCS/PF,

dvs. de havde hverken betalt hovedstol, gebyrer eller rente forud for overførslen til DCS eller PF. Disse kunder

er således ikke genstand for økonomisk godtgørelse.

Kunder, der har været aktive i mindre end 3 år.

I alt har 24.000 kunder kun været aktive i mindre end 2,5 år i relation til DCS-systemet og 3 år i relation til PF-

systemet, og de er ikke overflyttet til DCS- eller PF-systemet fra et ældre inkassosystem. Disse kunder kan

ikke kategorises under hovedårsag 1 og risikerer ikke at have betalt renter, omkostninger eller hovedstol, der

er forældet, idet forældelsesfristen er tre år for renter og omkostninger og 10 år for hovedstol samt

omkostninger tilkendt af retten.

Kunder, der ikke er ramt af hovedårsagerme

Endelig er der i alt 15.000 kundesager, som ved stikprøvekontrol er blevet bekræftet som værende behandlet

korrekt i både DCS- og PF-systemet. I nogle sager, der registreres i DCS, er renterne korrekt adskilt fra

hovedstolen i modsætning til dem, der ellers er observeret som et resultat af hovedårsag 1, hvor renterne er

forkert aggregeret med hovedstolen.

Danske Banks svar på Finanstilsynets brev med ref.nr. 6252-0349

Figur1 Kategorisering af de kunder, der er omfattet af potentielle beløbskorrektioner. Kunderne er inddelt i følgende kategorier:

Kunder i alt; Kunder uden for Danmark; Kunder i Danmark; Kunder, der ikke har betalt; Kunder, der har betalt; Kunder aktive i >3 år

og registreret i DCS/PF efter 2004; Tilbagebetalt hovedstol med evt. forældede gebyrer og renter; Kunder registreret korrekt i DCS/PF;

og Kunder med behov for beløbskorrigering.

5.3 Estimeret antal kunder, hvis sager potentielt skal korrigeres

Vi har gennemgået 17.000 kunder og identificeret 900 kunder, som har krav på kompensation. Ud fra disse

resultater har vi modelleret den resterende gruppe af kunder, og vores aktuelle skøn er, at det betyder, at

mellem 10.000 og 15.000 kunder kan være økonomisk påvirket.

5.4 Typiske godtgørelser

Den typiske kompensation

–

enten ved modregning i anden gæld eller ved kontant udbetaling

–

ligger på

1.000-2.000 kr. Beløbet kan dog i nogle sager være meget højere. Banken erkender, at der er kunder (især i

PF-systemet), der er berettiget til en relativt højere kompensation, og der vil også være kunder, der kun er

berettiget til et mindre beløb.

5.5 Yderligere sager, der skal korrigeres

Herudover vil en række kunder, som ikke har lidt noget økonomisk tab (dvs. kunder, der ikke har betalt af på

deres gæld og kunder, der har haft aktive inkassosager i mindre end 3 år) få deres sager genberegnet.

SPØRGSMÅL 6: HVORDAN HAR BANKEN HÅNDTERET DE BERØRTE KUNDER, HERUNDER HVORDAN

OPGØRES OG BEREGNES KOMPENSATIONEN FOR DEN ENKELTE KUNDE, OG HVILKE OPLYSNINGER GIVES

KUNDERNE OM FEJL OG BEREGNINGER?

6.1 Hvordan har banken håndteret de berørte kunder?

De systemiske datafejl, der potentielt har ramt kunderne, er opdelt i tre hovedkundegrupper:

ii.

Kunder med forkert registreret gæld i systemerne, men som har ikke betalt af på deres gæld.

Kunder, hvor der er foretaget forkert opkrævning; der kan dog foretages modregning i kundens

udestående gæld.

Danske Banks svar på Finanstilsynets brev med ref.nr. 6252-0349

iii.

Kunder, hvor der er foretaget forkert opkrævning, og som er berettiget til at modtage kompensation.

For så vidt angår ii) og iii) følger banken følgende principper for afhjælpning og kompensation af kunder:

Banken vil

•

Sættekunden i en økonomisk situation

, som om fejlene aldrig var sket.

Godtgøre kunderne baseret på alle tilgængelige data.

Baseres kompensationen på en model for direkte tab.

Etablere en proces for behandling af kundesager, der involverer indirekte tab.

Kompensere kunderne for eventuelle skattemæssige tab.

Ud af de 106.000 kunder, der potentielt er berørt, og som derfor har mulighed for at få kompensation, har

banken til dato undersøgt 17.000 kundesager ved at genberegne den enkelte kundes gæld og sammenligne

det genberegnede beløb med data, der er registreret i DCS (se svaret til spørgsmål 7 nedenfor for yderligere

oplysninger). Genberegningen af de 17.000 kundesager har vist, at der er opkrævet for meget i 930 tilfælde,

og disse kunder vil modtage godtgørelse, og/eller der vil ske modregning i kundens eventuelle udestående

gæld.

6.2 Hvordan fastsættes og beregnes kompensationen for den enkelte kunde?

Kundens gæld genberegnes ved hjælp af en datamodel, der er udviklet af EY for Danske Bank. Modellen

opdeler den samlede gæld i renter, gebyrer og hovedstol og anvender korrekte forældelsesfrister for hver

postering, der går tilbage til datoen for den første tilbagebetaling på kundens lån.

Modellen er baseret på bankens principper for kompensation ved, at der laves et skøn, der er til kundens

fordel, når den underliggende datakvalitet ikke giver mulighed for et mere præcist skøn (et konservativt skøn

over opkrævningen af for store beløb).

Formålet med modellen er at beregne godtgørelsen til kunder, der er ramt af hovedårsag 1 og 2, ved 1) at

genskabe de saldi, der oprindeligt blev registreret i DCS, men samtidig justere for renter og gebyrer, der ikke

er korrekt sammenlagt med hovedstolen (afhjælpning af hovedårsag 1) og ved 2) at rette de anvendte

forældelsesfrister (afhjælpning af hovedårsag 2). Når saldiene i DCS er korrigeret, genberegner datamodellen

de beløb, som kunden skylder i dag, på baggrund af alle de tilbagebetalinger, som kunden har foretaget; de

renter, der er påløbet korrekt på de forskellige dele af den samlede gæld; alle gældende forældelsesfrister,

der kan være udløst i lånets løbetid; den anerkendte gæld og/eller potentielle korrektioner, der er foretaget

i lånets løbetid (f.eks. når en debitor får gældssanering). Til sidst vil det output, der genereres af

datamodellen, være forskellen mellem, hvad saldiene er i DCS, og hvad saldiene burde have været i DCS, hvis

de ikke var påvirket af de fire hovedårsager. Forskellen mellem de faktiske saldi i DCS og de saldi, der

genberegnes af datamodellen, anses for at være det beløb, som Danske Bank har opkrævet for meget over

for kunden.

Kunder, der har betalt for meget, vil modtage kompensation svarende til forskellen mellem, hvad de skulle

have betalt tilbage på deres udestående gæld, og hvad de faktisk har betalt med tillæg af rente for den

periode, hvor pengene skulle have været til kundens rådighed.

Rentekompensationen beregnes fra datoen for kundens første overbetaling af lånet og indtil datoen for

bankens godtgørelse af kunden. Datoen for første overbetaling er sat som startdato for beregningen af

Kunder, som har lidt et økonomisk tab som følge af en af de fire hovedårsager, vil få mulighed for at indsende

dokumentation for yderligere potentielle følgetab.

Danske Banks svar på Finanstilsynets brev med ref.nr. 6252-0349

rentekompensationen, idet dette er den første dato for, hvornår banken har opkrævet for meget, og dermed

det tidspunkt, hvor kundens krav mod Danske Bank opstår.

Rentekompensationen beregnes på baggrund af det samlede beløb for, hvad kunden har betalt for meget. I

henhold til rentelovens § 5 er renten for rentekompensationen den officielle udlånsrente (offentliggjort to

gange årligt af Nationalbanken) plus et tillæg på enten 7 % eller 8 % p.a. (i overensstemmelse med

definitionen af morarente). Valget mellem 7 % og 8 % er afhængig af, hvornår kundens krav mod banken er

opstået, og er her defineret som, hvornår den første tilbagebetaling af gælden har fundet sted. For kunder,

der har foretaget deres første tilbagebetaling før den 1. marts 2013, er der et tillæg på 7 %, mens der for

kunder, der har foretaget deres første tilbagebetaling efter den 1. marts 2013, er et tillæg på 8 %.

I den aktuelle version er datamodellen ikke i stand til at registrere kunder, der potentielt har været ramt af

for store tilbagebetalinger som følge af hovedårsag 3 og 4. Grunden til, at hovedårsag 3 og 4 endnu ikke er

indarbejdet i modellen er kompleksiteten af beregningen af kundernes godtgørelse, idet dette kræver brug

af fysiske kundefiler. Der arbejdes fortsat på både at identificere og løse problemerne i forbindelse med

genberegning af gæld for kunder, som også er ramt af hovedårsag 3 og 4. Kunder, der risikerer at have betalt

for meget som følge af hovedårsag 3 og 4, undersøges og gennemgås i øjeblikket manuelt, og denne proces

vil om nødvendigt finde sted sammen med kunden.

6.3 Hvilke oplysninger modtager kunderne om fejl og beregninger?

Alle kunder, der er konstateret som værende ramt af hovedårsagerne, vil blive kontaktet direkte af Danske

Bank, så vi kan oplyse, hvad der er gået galt, hvad Danske Bank gør for at sikre, at kunden får den rette

kompensation, og hvad kunden kan forvente, at der dernæst sker.

SPØRGSMÅL 7: HVOR MANGE KUNDER HAR PR. 1. SEPTEMBER 2020 MODTAGET KOMPENSATION, OG

HVOR MANGE KUNDER FORVENTER BANKEN AT UDBETALE KOMPENSATION TIL, OG HVORNÅR

FORVENTES ALLE KUNDER AT VÆRE BLEVET KOMPENSERET?

7.1 Hvor mange kunder har pr. 1. september 2020 modtaget kompensation?

Pr. 1. september 2020 havde 326 kunder modtaget kompensation. Kompensationsbeløbene til disse kunder

udgør samlet 395.638 kr. Disse kunder er blevet gennemgået i et segment på 17.000 sager (som alle er

oprettet i DCS efter 2009). Vi er nu i gang med at korrigere sagerne, og arbejdet vil skride hurtigere frem i

resten af 2020.

Kompensationsbeløbene fordeler sig således:

•

25 kunder har modtaget mere end 2.000 kr. i kompensation

70 kunder har modtaget mere end 1.000 kr. (de 25 kunder indgår i dette antal) i kompensation

256 kunder har modtaget mindre end 1.000 kr. i kompensation

7.2 Hvor mange kunder forventer banken at skulle betale kompensation til?

På nuværende tidspunkt vurderer banken, at 10.000-15.000 kunder er berettiget til kompensation i form af

en godtgørelse eller en reduktion af deres restgæld.

7.3 Hvornår forventes det, at alle kunder har fået kompensation?

Danske Banks svar på Finanstilsynets brev med ref.nr. 6252-0349

Banken har udarbejdet en plan for kompensation af alle berørte kunder. Danske Bank forpligter sig til at rette

alle kundesager og kompensere alle kunder senest den 1. juli 2021.

SPØRGSMÅL 8: FINANSTILSYNET ANMODER ENDVIDERE OM AT MODTAGE RELEVANTE DELE AF BANKENS

INTERNE UNDERSØGELSER OG DEN UNDERSØGELSE, SOM BANKENS COMPLIANCE-FUNKTION OG

EKSTERNE ADVOKATER UDFØRER.

Finanstilsynet henviser i anmodningen til "relevante" dele af de forskellige undersøgelser, der er udført.

Banken har fortolket dette således, at Finanstilsynet gerne vil modtage de dele af undersøgelsesrapporterne,

der

(i) er relevante for Finanstilsynets undersøgelse af banken som anført i brev af 31. august: bankens

håndtering af de konstaterede fejl i bankens inkassosystem set i relation til bankens forpligtelse til at handle

redeligt og loyalt over for sine kunder (jf. § 43 i lov om finansiel virksomhed og bekendtgørelsen om god skik

for finansielle virksomheder), og

(ii) redegør for fakta og juridiske vurderinger vedrørende de forhold, som Finanstilsynet har spørgsmål til

ovenfor,

8.1 Interne undersøgelser

Banken har bedt EY om at vurdere og varetage kompensationen af kunderne, der er ramt af de konstaterede

systemiske datafejl i GRDM.

Vedlagt dette brev er følgende dokument:

•

Draft

–

EY report: Data analysis and considerations within debt collection of 1 July 2020 (bilag 2.4)

8.2 Eksterne advokatundersøgelser

Som nævnt ovenfor har banken bedt advokatfirmaet Plesner om at analysere systemfejlene, identificere

relevante risici og bistå med at afhjælpe systemfejlene.

Vedlagt dette brev er følgende dokumenter:

•

Draft System and Process Flaw Document of 28 November 2019 (bilag 2.1)

Workshop presentation of 26 November 2019 (bilag 2.2)

Draft e ora du o

ai root auses a d Da ske Ba k’s o ligatio to o pe sate usto ers

(bilag

2.3)

Presentation from conference call on 24 October 2019 (bilag 2.4)

--oo0oo--

Frans Woelders

Chief Operating Officer

Rob de Ridder

Head of Customer Service Delivery

Danske Banks svar på Finanstilsynets brev med ref.nr. 6252-0349

BILAG

2.1

2.2

2.3

2.4

2.5

3.3

3.4

Draft System and Process Flaws Document of 28 November 2019

Workshop presentation of 26 November 2019

Draft memorandum on main root causes and

Da ske Ba k’s o ligatio to o pe sate

customers of 13 December 2019

EY report: Data analysis and considerations within debt collection of 1 July 2020

List of potential derived or correlated issues

Presentation from conference call on 24 October 2019

Overview of the governance structure of Programme Athens

Early draft of 28 November 2019

The descriptions in this document are not finally verified

Certain parts are currently in Danish to speed up the verification process

STRICTLY CONFIDENTIAL

–

ATTORNEY-CLIENT

PRIVILEGE

PROJECT COLLECTION

DESCRIPTION OF THE KEY SYSTEM AND PROCESS FLAWS IN DCS AND CERTAIN OTHER SYSTEMS

TABLE OF CONTENTS

INTRODUCTION .................................................................................................................... 4

TIMELINE OF MAIN EVENTS ................................................................................................... 5

2.1

2.2

2.3

2.4

2.5

2.6

2.7

2.8

Pre-2004: Legacy Danske Bank systems

..................................................................... 5

2004: Introducing DCS

.............................................................................................. 5

2008: Change in statute of limitations

........................................................................ 6

2009: Correction team

............................................................................................... 6

2009: LEAN focus

...................................................................................................... 6

Lithuania outsourcing

................................................................................................ 6

2019: Operational Risk Event

..................................................................................... 6

2019: Correction Team

.............................................................................................. 7

GENERAL DESCRIPTION STATUTE OF LIMITATIONS ................................................................ 7

3.1

3.1.1

3.1.2

3.1.3

3.2

Statute of limitation ................................................................................................... 7

Loans

Interest and fees

Transitional period

Overview of relevant limitation periods ....................................................................... 8

DESCRIPTION OF KEY SYSTEM AND PROCESS FLAWS WITHIN DCS ......................................... 8

4.1

4.2

4.3

4.4

Introduktion til DCS-systemet .................................................................................... 8

Hvordan DCS formentlig var tiltænkt at fungere .......................................................... 9

Hvad der faktisk sker i DCS ...................................................................................... 10

Praktisk eksempel ................................................................................................... 11

HISTORIC ISSUES PRE-DATING DCS ..................................................................................... 14

5.1

5.1.1

5.1.2

5.1.3

5.2

5.2.1

5.2.2

5.2.3

5.2.4

Legacy Danske Bank systems ................................................................................... 14

Generel introduktion

Fejlregistrering af advokatomkostninger

Migrationsfejl

Legacy BG Bank systems ......................................................................................... 17

Generel introduktion

Registrering af debitorer og kautionister

Skift af kontonumre

Regionale forskelle i håndteringen af kontotype 3

5.2.5

5.2.6

Eksempel på et hændelsesforløb

Tilgodehavender til ekstern inddrivelse

PROJECT COLLECTION

DESCRIPTION OF THE KEY SYSTEM AND PROCESS FLAWS IN DCS AND CERTAIN OTHER SYSTEMS

INTRODUCTION

We have been instructed by Danske Bank A/S ("Danske Bank") to assist with certain data quality

issues identified within the debt collection systems in Group Recovery & Debt Management

("GRDM"), which have resulted in cases of collection of time-barred debt.

The main scope of our work has been:

•

to understand and describe the key system and process flaws that were identified in the

systems and associated processes;

to assess and describe the key legal and regulatory risks resulting from the identified

system and process flaws; and

to assist Danske Bank in identifying possible ways of solving the issues resulting from the

system and process flaws.

In order to deliver on these tasks, we have performed a detailed fact-finding exercise involving a

number of interviews of relevant members of staff, participated in meetings held by GRDM's 'Cor-

rection Team', and observed how case officers navigate and use the debt collection systems. We

have not made an analysis of how the relevant IT systems are coded. The descriptions contained

in this document are therefore based on the information we have received about how the relevant

debt collection systems work and examples from actual cases in the systems.

[Comment: Section about the verification of the description that has been carried out by GRDM to

be inserted here. We are also working on including examples for all the system flaws to make it

possible to see an actual case from DCS in which the specific flaw has occurred.]

The purpose of this document is to describe the key system and data flaws in the Debt Collection

System ("DCS") and in the historic debt collection systems from which data in the DCS is based.

We note that some of the flaws described in this document remain unresolved or have only recently

been resolved, whilst some of the historic flaws only transpired in connection with the migration of

data into the DCS in 2004 or occurred in the systems that preceded the DCS. The end result,

therefore, is that a high number of customers in GRMD - both current and historic - may therefore

be impacted by the flaws.

We have also included a section setting out a high-level timeline of certain events that are relevant

either for the purposes of understanding some of the system and processing flaws or to understand

the scope or impact of the flaws.

Furthermore, we have also included a high-level summary of the applicable rules on limitation

periods that have applied during the relevant period. As most of the data flaws relate to incorrect

handling of the statutory limitation periods applicable to the various elements that make up the

outstanding debt (interest, collection costs and the principal amount of the loan) it is relevant to

have an understanding of these rules and how they changed in 2008.

The varying elements (interest, costs, and principal amount) that make up an outstanding debt

need to be tracked separately within debt collection systems in order for the systems to be able to

apply the applicable varying time-barring periods. Accordingly, the debt collection systems are

meant to break up an outstanding debt into different data fields, each representing an element of

the outstanding debt, and each with their own varying periods of limitation.

For the purposes of understanding some of the system and process flaws described below, it is

necessary to understand that when a debt is 'written-off', from an accounting perspective, it is still

tracked and considered to be owed by the customer, and attempts are made by Danske Bank to

collect such outstanding debt.

This document is structured as follows: In Part 2 immediately below we have set out the timeline

of main events that have occurred in the relevant time period. In Part 3 we have set out the high-

level summary of the relevant statute of limitation rules. Part 4 contains our description of the

system and process flaws within DCS and Part 5 contains our description of the system and process

flaws pre-dating DCS.

TIMELINE OF MAIN EVENTS

This section contains a chronological timeline of what we understand to be the key events in the

process leading up to the issues Danske Bank is facing today in respect of the system and process

flaws. The list does not include events that may be relevant from an accountability perspective.

2.1

Pre-2004: Legacy Danske Bank systems

BG Bank was formed in a merger in 1996 between Bikuben and Girobank. During 1999 the debt

collection portfolio of Girobank was transferred / converted to the recovery system in Bikuben.

To our knowledge the collection system in Bikuben can be accessed back to approx. 1984. The

recovery system from Girobanken's 'Jurix' cannot be accessed.

In 2001 Danske Bank acquired BG Bank, including BG Banks debt collection portfolio and depart-

ment.

Initially BG Bank's debt collection department operated independently from a number of other

decentralised Danske Bank debt collection departments.

It can be assumed that the majority of the older collection cases originating in BG-Bank and the

Danske Bank legacy collection system were incorrectly registered. These issues were further exac-

erbated, through the compounding of issues, when the cases were subsequently integrated into

DCS, cf. below.

2.2

2004: Introducing DCS

In 2004 a new debt collection system (DCS) was implemented in an effort to centralise debt col-

lection processes across Danske Bank. Data on all outstanding defaulted debt was therefore mi-

grated into DCS.

There appears to have been some knowledge of the existing issues within the earlier decentralised

systems, which in part led to the introduction of DCS, which in turn was meant to be a new and

improved system. Attempts to implement the new system spanned over a period of time, and we

understand that this is due to the fact that already at this stage it was known that DCS had some

inherent issues that needed to be fixed prior to "go-live". We have been informed that the "go-live"

date was pushed back once, but that the management decided to implement the DCS despite

awareness of at least some of the system and process flaws because it was not considered an

option to push the "go-live" date a second time. It remains unclear to us, what efforts were made

to rectify the issues within DCS before implementing, or whether the full extent of the issues known

today, were also known then.

All existing cases, irrespective of where they originated, were transferred to DCS on 27 September

2004. All new cases following this date were opened directly within DCS (except from cases origi-

nating in the PF-system).

It is relevant to note already now, that rather than flaws in how DCS treats and calculates out-

standing debt, the key problems relate to the point of data entry into DCS. In other words, when

an existing, or a new case, was or is transferred to, or created, in DCS.

2.3

2008: Change in statute of limitations

In 2008, new rules on the statute of limitations were implemented in Denmark and going forward

principal amounts on loans have in general a time-barring period of 10 years and interest and costs

carry a time-barring period of 3 years. Prior to this, the time-barring periods were 20 years and 5

years, respectively. Moreover, the new rules made it much more difficult to suspend the time-

barring period. In order to suspend the time-barring period a large portion of the claims was sent

to external debt collection agencies.

2.4

2009: Correction team

Perhaps in the knowledge that DCS contained some inherent data quality issues, and in conjunction

with the new statute of limitations, a Correction Team was established in 2009 to manually correct

the cases in DCS on a continuous basis. However, in many instances this was done without the

case handler being able to reconcile the most recent calculation of debt owed against an original

account statement. By inference, therefore, some but not all cases that underwent correction, were

not corrected to the accurate amount. The Correction Team managed to review some, but far from

all cases in DCS, before it was disbanded in around 2011. Plesner does not know how many cases

were corrected at this stage, or the extent to which these cases were corrected accurately.

2.5

2009: LEAN focus

In 2009 Danske Bank launched an initiative known as LEAN, which aimed to streamline existing

business processes in an attempt to become more efficient in case handling. It is not currently clear

to us exactly when the focus on LEAN ended but it is our understanding that the strong focus on

LEAN lasted for a longer period. In the case of DCS, the focus on LEAN meant that GRDM was

instructed to only rely on the data as presented by DCS when performing debt collection (such as

when entering into settlement agreements, seeking court orders etc.) and not to spend time looking

at physical files. Previously greater efforts were made to ensure that amounts reflected in DCS

were reconciled against a previous account statement or other evidence on file for that customer,

e.g. by reviewing the physical customer file that contained information on the correct principal

amount owed by the customer etc. It is our impression that the focus on LEAN further exacerbated

some of the issues present today, as presumably some of the case errors would have been caught

earlier.

2.6

Lithuania outsourcing

Danske Bank operates an off-shore office in Lithuania that mainly performs operations and other

administrative tasks. For a period of time since 2004, staff in Lithuania conducted case handling on

behalf of GRDM using DCS. The focus on LEAN meant that staff in Lithuania were incentivised to

produce results over quality in their case handling. Further to LEAN, this may have meant collection

of debt has occurred without any controls or reconciliation to ensure that the data produced by

DCS was correct.

2.7

2019: Operational Risk Event

On 24 May 2019 an ORIS was escalated internally at Danske Bank confirming that it had discovered

that DCS suffered from underlying data quality issues, and that the inherent issues with the debt

collection processes were not merely manual errors, as apparently previously thought. The Danish

FSA was notified on 6 June 2019.

2.8

2019: Correction Team

The Correction Team was re-established in August 2019, following the ORIS, with a view to man-

ually review and correct all cases in DCS that had been mishandled. We have been informed by

members of the Correction Team that some cases can be dealt with relatively quickly, whilst other

cases can take days to recalculate and complete - and in some cases it might not even be possible

to perform a recalculation due to the lack of a physical customer file.

GENERAL DESCRIPTION STATUTE OF LIMITATIONS

When describing the applicable statute of limitation rules, it is relevant to distinguish between the

following three time periods:

1) The period before 1 January 2008. In this period the time-barring rules were governed by

two limitation acts covering loans, fees and interest (a) Danske Lov 5-14-4 ("DL"), and (b)

the 1908 Act of Limitation (Act No. 274 of 22 December 1908) (the "1908 Act").

2) The period from 1 January 2008. In this period the Act of Limitation (Act No. 522 of 6 June

2007) (the "2008 Act") governed, and continues to govern, the time-barring rules in re-

spect of loans, fees and interest.

3) The transitional period for obligations incurred before 1 January 2008 that were unresolved

on 1 January 2008 and therefore became subject to the specific transitional rules in the

2008 Act.

The purpose of the 2008 Act was to replace DL and the 1908 Act and the 2008 Act remains the

applicable legal framework.

3.1

3.1.1

Statute of limitation

Loans

Before 1 January 2008

Prior to the 2008 act, the limitation period on loans was regulated by DL. Under the DL, the limita-

tion period was twenty (20) years from the

establishment of the claim

, e.g. the establishment of a

loan. Such limitation could be interrupted by either (i) the debtor's perception of the claim, ii) by

the creditor reminding the debtor about the obligation, (iii) a notice to an estate or iv) request of

appropriation (

anmodning om udlæg

). If the limitation period was interrupted before the loan be-

came time-barred, a new limitation period would begin.

From (and including) 1 January 2008

Under the 2008 Act, the limitation period for loans is ten (10) years

from the due date of the loan

The statute of limitations was thereby changed from twenty (20) years to ten (10) years. The new

legal requirements for interrupting the statute of limitation period, under the 2008 Act, are to either

(i) obtain the acknowledgment of the obligation by the debtor or (ii) by taking legal action against

the debtor.

3.1.2

Interest and fees

Before 1 January 2008

Prior to the 2008 act, the limitation period on interest and fees were regulated by the 1908 Act.

Claims on interest, which was due at specified intervals but not considered as repayment of a debt

owed, carried a statute of limitation period of five (5) years

from the due date

. Claims on interest

become time-barred either independently or if the principal amount became time-barred.

From (and including) 1 January 2008

Under the 2008 Act, interest and fees will as a general rule become time-barred after three (3)

years

from the due date

However, it should be noted if claims on interest and/or fees are attributed within the maximum of

an overdraft facility (

kassekredit

), the limitation period is ten (10) years

from the due date

Moreover, if a debtor acknowledges the debt on interest and fees, wholly or partially, the limitation

period will be 10 years

from the due date

3.1.3

Transitional period

When the 2008 Act entered into force on 1 January 2008, the general transitional rule was that the

new 2008 Act applied to previously established obligations, which were not timed-barred under the

DL or the 1908 Act before the date of entry into force.

The transitional rules were relevant if:

•

the applicable rules led too time-barring after 1 January 2008; and

the 2008 Act led to time-barring before 1 January 2011.

In order to avoid unintended consequences, a transitional period was introduced in the 3-year

period, after which time-barring only occurred if it followed from the previous as well as the newly

applicable rules - and if so the latest limitation date of the two possible limitation dates should

apply as the limitation date in this transitional period.

The transitional rules imply that if a reminder was sent to the debtor prior to 1 January 2008 the

interruption will have effect under the 2008 Act (even though such an interruption cannot be made

under the 2008 Act). A reminder sent before 1 January 2008 in relation to a claim, which according

to the 2008 Act is subject to a limitation period of ten (10) or tree (3) years, triggers a limitation

period of respectively ten (10) or tree (3) years for the claims (and not the twenty (20) and five

(5) years known under DL and the 1908 Act.)

For claims established on 1 January 2008, or later, the transitional rules will not be relevant.

3.2

Overview of relevant limitation periods

We have in annex 1 to this document set out an overview of the various limitation periods relevant

to the debt portfolio. Any reference to Danske Bank shall be read as a reference to any previous

lender having taken such action or has been unknown to the relevant fact.

4.1

DESCRIPTION OF KEY SYSTEM AND PROCESS FLAWS WITHIN DCS

Introduktion til DCS-systemet

DCS er det centrale gældsinddrivelsessystem i Danske Bank.

Når et udestående gældsforhold overdrages fra en filial til GRDM, flyttes kontoen samt ansvaret

for at inddrive gælden samtidigt til DCS og GRDM.

De identificerede system- og procesfejl er i al væsentlighed resultat af flere forskellige "data-entry"

flaws. Som konsekvens af disse data-entry flaws vil alle senere løbende beregninger forudsæt-

ningsvist være forkerte i et eller andet omfang. En fejl på registreringstidspunktet vil således på-

virke restgældens størrelse, hvis ikke sådanne entry-flaws manuelt over tid korrigeres.

De identificerede fejl er proces/systembundne, og den følgende beskrivelse skal dels illustrere da-

tastrømmene, når en sag oprettes i DCS, og dels illustrere hvor og hvorfor fejlene opstår.

I DCS opdeles den udestående gæld i seks forskellige datafelter. Differentieringen mellem disse

datafelter er nødvendig bl.a. af hensyn til fremtidig beregning af renter, overvågning, kontrol og

afbrydelse af forskellige forældelsestidspunkter, samt for indberetning af relevante rentefradrag

over for Skattestyrelsen.

Figur 1: Datafelter i DCS

Felt nr.

Beskrivelse

Omkostninger

Forældelsesregel

i DCS

F.eks. tilkendte sagsomkostninger. Datafeltet er typisk tomt,

når sagen oprettes i DCS.

Beregnet ikke betalte renter

10 år

Renter påløbet i filialen - der ikke er betalt og før sagen op-

rettes i DCS.

Hovedstol

3 år

Resthovedstol

Anerkendte renter

10 år

Renter påløbet efter sagens oprettelse i DCS, og som kunden

har anerkendt ved f.eks. indgåelse af frivilligt forlig. Ren-

terne fra datafelterne 5 og 6 overføres til dette datafelt efter

en 'anerkendelse'.

Løbende renter

10 år

Totale renter påløbet efter sagens oprettelse i DCS.

Løbende renter i år

3 år

Renter påløbet i indeværende år.

3 år

Når en debitor foretager et afdrag på et udestående gældsbeløb er der etableret følgende dæk-

ningsrækkefølge i DCS (i forhold til datafelterne): 1, 2, 5, 4 og 3.

Omvendt hvis der foretages modregning er dækningsrækkefølgen: 1, 2, 3, 4 og 5.

[Beskrivende tekst om felt 6.]

4.2

Hvordan DCS formentlig var tiltænkt at fungere

Følgende afsnit viser alene, hvordan DCS formentlig var tiltænkt at fungere med blik på at eksem-

plificere de fejl, som opstår i DCS. I afsnit 4.3 nedenfor angives således, hvad der rent faktisk sker

i DCS.

Når en sag oprettes i DCS registreres hovedstolen som værende resthovedstolen og eventuelle

anerkendte renter. Det samlede beløb registreres i datafelt 3.

Datafelterne 5 og 6 anvender beløbene som angivet i datafelterne 2, 3 og 4 til at beregne løbende

renter. Renterne i datefelterne 5 og 6 holdes adskilt fra datafelterne 3 og 4, idet forældelsesfristen

for disse fordringer er 3 år, hvorimod forældelsesfristen for hovedstolen og anerkendte renter er

10 år.

Opnås en ny dom eller indgås et nyt frivilligt forlig, akkumuleres renterne (datafelterne 5 og 6) og

overføres til datafelt 4. Samtidig overføres ethvert beløb i datafelt 2 til datafelt 3, hvorved beløbet

betragtes som en del af hovedstolen. Beløbet i datafelt 2 overføres til datafelt 3 - og ikke datafelt

4 - da der allerede er søgt om skattefradrag for renterne i datafelt 2. Danske Bank søger herved at

undgå registrering af dobbelt rentefradrag for kunden.

Datafelt 1 indeholder tilkendte omkostninger, der f.eks. tilkendes som led i Danske Banks retslige

inddrivelse af et tilgodehavende. Beløbet forbliver i datafelt 1, og rykkes ved en dom eller en for-

ligsaftale ikke til datafelt 3.

4.3

Hvad der faktisk sker i DCS

Bemærk, at følgende beskrivelse og illustrationer er eksempler på de hyppigste fejl, der er identi-

ficeret i forbindelse med tilretningsprocessen. Det skal understreges, at vi ikke er bekendt med,

hvorvidt sagerne er repræsentative. Der opstår fortsat nye "problemer", hvor der ikke umiddelbart

findes en logisk forklaring. F.eks. har vi fået forklaret, at inddrivelsen af krav over for danske kau-

tionister systemmæssigt er oversendt til Norge.

Der er ikke i Danske Bank et fuldstændigt overblik over alle fejl i DCS. Hertil kommer, at der fortsat

identificeres fejl, hvis oprindelse ikke umiddelbart lader sig forklare.

Figur 2: Sekvens af begivenheder for et nyt lån i en filial og før sagens oprettelse i DCS

Reminder

notice sent:

DKK 100

1 May

Transfer

into DCS

1 Jan

1 Feb

1 Mar

1 Apr

1 Jun

1 Jul

1 Aug

1 Sep

1 Oct

1 Nov

DCS

Key:

Interests

DKK 2,500

Interests

DKK 2,500

Interests

DKK 2,500

DKK 1,000 Monthly Contribution Paid

DKK 1,000 Monthly Contribution Unpaid

Debitors sidste rettidige betaling blev i ovennævnte eksempel foretaget i august.

Almindeligvis fremsendes 3 rykkere, før en sag overføres til DCS.

4.3.1

Delvist betalte renter

Tilskrevne renter dækkes forud for hovedstolen. Ovennævnte scenarie beskriver den fejl i DCS, der

opstår, når et afdrag delvist dækker allerede tilskrevne renter. I eksemplet bliver der foretaget to

afdrag på DKK 1.000 i hhv. juli og august, der tilsammen dækker DKK 2.000 af en rentetilskrivning

på DKK 2.500.

De resterende DKK 500 er fortsat "ubetalte renter" i andet kvartal. De DKK 500 (og eventuelt

senere ubetalte renter) burde overføres til datafelt 2 i DCS. Resthovedstolen skal overføres til da-

tafelt 3. Imidlertid overføres de DKK 500 fejlagtigt til datafelt 3 i DCS, og beløbet registreres derved

som en del af resthovedstolen. Med andre ord registreres der i dette eksempel DKK 500 for meget

i datefelt 3. Eventuelle senere ubetalte renter (f.eks. de renter, der tilskrives i september) overføres

korrekt til datafelt 2 i DCS.

DCS-systemet synes ikke at kunne skelne mellem delvist betalte renter og delvist betalte ydelser,

hvorfor de behandles på samme måde, hvilket ikke er korrekt.

Ved oprettelsen af sagen i DCS indeholder datafelt 3 således fejlagtigt delvist ubetalte renter, der

er påløbet før gælden blev overført til DCS. Sammenlægning af delvist betalte renter og hovedsto-

len i datafelt 3 indebærer, at DCS anvender forældelsesreglerne for resthovedstolen (10 år) på

delvist betalte renter (i vores tilfælde DKK 500). Derfor bliver DKK 500 "markeret" som forældet

efter 10 år. Reelt burde renterne forældes efter 3 år.

Denne iboende fejl i DCS-systemet medfører i yderste konsekvens, at Danske Bank ultimativt op-

kræver et større beløb end den faktiske gæld.

4.3.2

Behandling af rykkergebyrer

Forældelsesfristen for rykkergebyrer er i udgangspunktet 3 år.

Rykkergebyrer, der er tilskrevet før sagens oprettelse i DCS, er - siden implementeringen af syste-

met i 2004 - blevet overført direkte til datafelt 3. Rykkergebyrerne er derfor blevet inkluderet som

en del af hovedstolen til beregning af de løbende renter. Rykkergebyrer burde ikke registreres i

datafelt 3, men snarere i et selvstændigt datafelt.

Systemopsætningen har på dette punkt i praksis betydet, at sagsbehandlerne skulle foretage en

manuel korrektion af hovedstolen, når de relevante gebyrer blev forældet. Denne korrektion skulle

foretages 3 gange (én gang for hvert rykkergebyr), idet rykkerne blev sendt på forskellige tids-

punkter (typisk med en måneds mellemrum) og derfor også forældes på forskellige tidspunkter.

Denne manuelle sagsbehandling er ikke i alle tilfælde blevet foretaget, og den manuelle korrektion

har tillige været fejlbehæftet.

I juli 2019 blev denne fejl systemmæssigt korrigeret, således at ubetalte rykkergebyrer blev adskilt

fra hovedstolen. Ubetalte rykkergebyrer overføres på nye sager til datafelt 2. Ændringen i juli 2019

har alene virkning for fremtidige sager, og der er ikke foretaget en automatisk korrektion af de

allerede eksisterende sager i DCS. [Det er for nuværende uklart, hvorvidt registreringen i relation

til forældelse og indberetning af rentefradrag for kunderne foretages korrekt.]

4.3.3

Beregning af forældelsesperioden

Forældelsesfristen for hovedstol, påløbne renter og gebyrer regnes fra forfaldstidspunktet. Imid-

lertid beregner DCS forældelsesfristen ud fra oprettelsestidspunktet i DCS. Som konsekvens inde-

bærer dette, at man i DCS nødvendigvis ikke kan identificere og kontrollere det korrekte tidspunkt

for afbrydelse af forældelse.

Det indebærer, at frivillige forlig eller retsafgørelser i praksis kan indeholde allerede forælde renter,

gebyrer og eventuelt hovedstole. Dette gælder navnlig, hvis et frivilligt forlig indgås omkring tids-

punktet for forældelsesfristens udløb registreret i DCS. Hertil kommer, at ikke-betalte ydelser under

(i hvert fald) gældsbreve selvstændigt forældes. Ikke-betalte ydelser forældes således før den

samlede restgælds forældelse.

Såfremt hovedstolen allerede er forældet på tidspunktet forud for indgåelsen af et frivilligt forlig,

har Danske Bank fortabt retten til at kræve hovedstol og renter betalt til sig.

4.4

Praktisk eksempel

De i punkt 4.3 beskrevne problemstillinger er de primære systemfejl i DCS (foruden de beskrevne

data-entry flaws fra BG Bank og "Legacy Danske Bank" i punkt 5).

Correction-teamet gennemgår fortsat nye og gamle sager, og der fremkommer forskellige typer af

systemmæssige problemstillinger, hvis nærmere oprindelse ikke umiddelbart kan forklares. Dette

illustreres via følgende eksempel, som viser fejl i DCS-systemet, der pt. ikke præcist kan beskrives.

Figur 3: Kontooversigt

Date

08.11.2007

11.12.2007

31.12.2007

03.01.2008

14.02.2008

04.03.2008

31.03.2008

07.04.2008

22.04.2008

10.06.2008

Text

Transport

Credit

Interest

Overdraft Interest

Credit

Interest

Overdraft Interest

Reminder Notice

Overdraft Interest

Debit

Credit

2.500,00

2.151,96

6,16

2.500,00

2.177,78

8,61

100

1.738,22

Balance

-91.697,27

-89.197,27

-91.349,23

-91.355,39

-88.855,90

-86.355,39

-83.855,39

-86.033,17

-86.041,78

-86.141,78

-86.241,78

-86.291,78

-88.030,00

*Overtræksrenter er renter tilskrevet efter ubetalte månedlige ydelser.

Den 4. marts 2008 foretog kunden i eksemplet det sidste rettidige afdrag på gælden. På tidspunktet

udgjorde gælden samlet DKK 83.855,39. Dette beløb burde rettelig overføres til datafelt 3 i DCS

(hvor DKK 200 i rykkergebyrer - fejlagtigt - også blev overført, da sagen er fra før juli 2019, hvor

denne fejl blev rettet).

Efter den 4. marts 2008 er der 4 gange tilskrevet renter for samlet DKK 3.974,61. Dette beløb

burde være overført til DCS-systemets datafelt 2. I alt udgør det samlede krav DKK 88.030,00.

Uanset ovenstående er alle tilskrevne renter frem til den 22. april 2008 fejlagtigt overført til datafelt

3, og kun renterne tilskrevet den 10. juni 2008 er overført til datafelt 2. Dermed henstår der en

urigtig hovedstol på DKK 86.241,78, der er DKK 2.186,39 for højt (ekskl. rykkergebyrer på DKK

200) (beløbet er markeret med gult i både figur 3 og 4).

Renterne registreret i datafelt 3 registreres som konsekvens fejlagtig og ikke som forældet efter 3

år.

Figur 4: Screenshot af sag i DCS

Figur 4 illustrerer tilretningen af sagen. I datafelt 3 burde der have stået DKK 83.855,39 svarende

til udeståendet pr. 4. marts 2008.

Figur 5: Screenshot af sag i DCS

Figur 5 illustrerer den korrekte registrering af renterne tilskrevet den 10. juni 2008 i datafelt 2.

Fejlens opståen kan ikke umiddelbart forklares.

Uanset ovenstående medfører denne iboende fejl i DCS-systemet i yderste konsekvens, at Danske

Bank via DCS-systemet ultimativt opkræver et større beløb end den faktiske gæld.

Figur 6

: Illustration DCS-systemets behandling af sagen anvist i figur 3

Datafelt 1

DKK 0,00

Datafelt 2

DKK 1.788,22 (which includes the two sets of interests

incurred on the 10

of June)

Datafelt 3

DKK 86.241,78 (which includes the correct amount of

DKK 83.855,39, but also the interest incurred on the

of March as well as the two reminder notice costs)

Datafelt 4

N/A

Datafelt 5

N/A

Datafelt 6

N/A

5.1

5.1.1

HISTORIC ISSUES PRE-DATING DCS

Legacy Danske Bank systems

Generel introduktion

Forud for DCS-systemet var inddrivelsessystemet internt i Danske Bank decentralt og baseret på

følgende tre overordnede konti:

1. Almindelig konto

f.eks. indlån, udlån og kreditter

2. Inkassokonto

Overgang fra filial til inkassoafdelingen

3. Afskrevet fordringer

Uerholdelige fordringer

I Danske Banks decentrale gældsinddrivelsesenheder var der ikke etableret en systemmæssig be-

talingsrækkefølge.

Når en given debitor misligholdte sin betalingsforpligtelse overførtes den misligholdte fordring fra

kontotype 1 til kontotype 2, dvs. til "inkassokontoen". Medmindre der med kunden blev indgået et

frivilligt forlig eller man i Danske Bank opnåede dom blev fordringen afskrevet, og fordringen blev

overført til kontotype 3. Hvis en betalingsforpligtelse efter et frivilligt forlig eller en dom fortsat blev

misligholdt blev tilgodehavendet tilsvarende afskrevet og overført til kontotype 3. Kontoen beva-

rede det samme kontonummer indtil afskrivelsen, hvorefter der blev udfyldt et E-blad, og følgende

informationer blev registreret i den fysiske kundemappe:

•

Oplysning om gammelt og nyt kontonummer,

Navn på debitor og meddebitorer,

Navn på kautionister og deponenter,

•

Retsbehandling, og

Dato for afskrivning

Registreringsmæssigt var det den samlede saldo på kontotype 2, der blev overført til kontotype 3.

Det vil sige, at når en misligholdt fordring blev afskrevet anførtes den samlede saldo som et enkelt

beløb i kontotype 3 (inkl. renter, omkostninger, gebyrer m.v.). Illustreres simplificeret nedenfor:

Kontotype 2

Lån (inkl. renter og om-

kostninger/gebyrer):

DKK 155.000 (hovedstol)

+ DKK 10.000 (renter)

+ DKK 5.000 (omk./geby-

rer)

Kontotype 3

DKK

170.000

DCS

DKK

170.000

(registreret i kontotype 3

i DCS)

Danske Bank opererede ikke med én kontotype 3 pr. gældspost. I stedet havde man i Danske Bank

en "samlekonto" for en debitors samlede afskrevne gæld. Denne integrationsmodel har bl.a. med-

ført, at de forskellige forældelsesfrister og starttidspunkter for fordringernes respektive forældel-

sestidspunkter blev umulig at identificere. Problemet videreførtes ved indførelsen af DCS. Hvis

denne registrering manuelt ikke var blevet tilrettet ville den samlede saldo fra kontotype 3 blive

overført til DCS-systemet.

Figur 7: En sag med en debitor med flere gældsposter til inddrivelse

Kontotype 1

(f.eks. gældsbreve)

DKK 100.000

Kontotype 2

(f.eks. kassekredit)

DKK 50.000

Kontotype 2

DKK 100.000

Kontotype 2

DKK 50.000

Kontotype 3

DKK 150.000

5.1.1.2

Dobbeltregistrering

Såfremt en fordring registreret i kontotype 3 havde to eller flere debitorer eller kautionister, blev

der oprettet en konto for hver debitor med hver sit kontonummer. Gælden blev herved registreret

dobbelt.

Som anført ovenfor opererede Danske Bank ikke med én kontotype 3 pr. gældspost - men derimod

en kontotype 3 pr. debitor. Registreringen medførte som konsekvens, at bankens tilgodehavende

fremstod forkert, og hvis der ikke internt i Danske Bank blev foretaget en manuel korrektion, risi-

kerede man herved, at Danske Bank opkrævede et større beløb end det faktiske skyldte.

Kontotype 1

Debitor 1 - egen gæld

DKK 100.000

Kontotype 1

Fællesgæld

DKK 150.000

Kontotype 1

Debitor 2 - egen gæld

DKK 20.000

Kontotype 2

DKK 100.000

Kontotype 2

DKK 150.000

Kontotype 2

DKK 20.000

Kontotype 3

Debitor 1

DKK 250.000

Kontotype 3

Debitor 2

DKK 170.000

Figur 8: En sag med 2 debitorer, hvor hver især har en egen gæld samt en fællesgæld

Når en konto blev afskrevet og overført til saldotype 3, blev der for hver skyldner oprettet et E-

blad. På dette E-bland skulle løbende indbetalinger registreres - hvis de fandt sted.

Disse E-blade foreligger kun i fysisk form (ringbind), og blev ikke registreret i et elektronisk system.

Alene ud fra E-bladet kunne man konstatere, hvorvidt der var andre skyldnere (meddebitorer), der

hæftede for fordringen. En meddebitor skulle registreres med et grønt kryds på forsiden. Dette var

metoden, hvorved man søgte ikke at inddrive det samme beløb to (eller flere) gange.

5.1.2

Fejlregistrering af advokatomkostninger

Registrerede tilkendte sagsomkostninger er ikke i overensstemmelse med de faktisk tilkendte sags-

omkostninger fra f.eks. en inkassosag.

De danske domstole fastsætter standardiserede sagsomkostninger i inkassosager, der sjældent

modsvarer de faktiske advokatomkostninger. Tilkendte sagsomkostninger kan tilskrives restgæl-

den. Omkostninger herudover er kunden uvedkommende, og er en omkostning Danske Bank som

kreditor må bære.

Det er konstateret, at Danske Bank har registreret de faktiske advokatomkostninger frem for de

tilkendte sagsomkostninger. Hvorvidt denne praksis er fastholdt også efter indførelsen af DCS er

uvist.

Dette må dog beskrives som en manuel fejl. Hvorvidt det dog skyldtes en forkert intern procedure

eller blot en forkert praksis i p.t. uvist.

5.1.3

Migrationsfejl

Ved konvertering til DCS i 2004, blev saldoen for hver kontohaver (kontotype 3) overført til datafelt

3 i DCS.

Integrationen skete uagtet, at saldoen kunne indeholde forskellige hovedstole, renter og omkost-

ninger, der potentielt kunne være forældede.

Renter og omkostninger burde ikke registreres i datafelt 3 i DCS - (medmindre renterne er aner-

kendt).

Alene den totale saldo overgik til DCS-systemet.

Oplysninger om tidligere kontonumre eller navne på meddebitorer/kautionister blev ikke registreret

i DCS. Det var ikke muligt løbende at registrere indbetalinger parallelt mellem flere meddebitorer.

Eksempel (se figur 8)

Hvis debitor 1 indbetalte DKK 150.000 (et beløb der dækker egen gæld samt DKK 50.000 af

fællesgælden) ville DKK 170.000 fortsat stå til inddrivelse ved debitor 2. Såfremt der ikke blev

foretaget en manuel korrektion, ville der som konsekvens fortsat blive gjort forsøg på at inddrive

det fulde beløb ved debitor 2 - selvom gælden allerede delvist var blevet betalt.

5.2

5.2.1

Legacy BG Bank systems

Generel introduktion

Inddrivelsessystemet i BG Bank var baseret på følgende tre overordnede konti (idet der for kasse-

kreditter dog ikke var behov for oprettelse af en kontotype 2):

1. Almindelig konto

f.eks. indlån, udlån og kreditter

2. Inkassokonto

Overgang fra filial til inkassoafdelingen

3. Afskrevet fordringer

Uerholdelige fordringer

I BG Bank var der etableret en systemmæssig betalingsrækkefølge, hvor afdrag først anvendtes til

at inddække allerede påløbne renter forud for hovedstolen. Påløbne renter blev først dækket på

kontotype 1 og derefter kontotype 2. Påløbne renter blev kun dækket på kontotype 3, idet omfang

at renterne var anerkendt af debitor.

Når en debitor misligholdte sin betalingsforpligtelse, overførtes den misligholdte fordring fra kon-

totype 1 til kontotype 2, dvs. til "inkassokontoen". Filialen oprettede kontotype 2 udfyldte samtidig

et stamblad, hvor følgende informationer blev registreret i den fysiske kundemappe:

•

Navn på debitorer,

Navn pa kautionister og deponenter,

Kontonummer på konti der skulle til inkasso,

Kontotype - gældsbrev, kassekredit eller almindeligt overtræk,

Oprindelig lånebeløb på hver enkelt konto,

•

Saldo ved overførsel til inkasso på hver enkelt konto (samlede saldo),

Oplysning om eventuel sikkerhedsstillelse,

Beskrivelse af sagsforløb i filialen samt oplysninger om debitors situation, og

Dato for overførsel til inkasso.

Systemmæssigt kan man ikke tilgå disse informationer.

Når inkassosagen var blevet behandlet og skulle afskrives, blev fordringen overført til kontotype 3,

dvs. til kontoen med afskrevne fordringer.

Inkassoafdelingen udarbejdede forud for afskrivningen et stamblad kaldet

’’Indstilling til afskriv-

ning”,

hvilket stamblad indeholdte informationer om:

•

Oplysning om gammelt og nyt konto nr.

Saldo på krav på hver enkelt konto (det samlede udestående),

Hvad der var sket i sagen,

Retsbehandling,

Dato for afskrivningen, og

Oplysning om sagsforløb.

Registreringsmæssigt var det den samlede saldo på kontotype 2, der blev registreret på kontotype

3. Det vil sige, at når en misligholdt fordring blev afskrevet anførtes den samlede saldo som et

enkelt beløb i kontotype 3.

Ved overførslen fra kontotype 2 til kontotype 3 blev der således ikke sondret mellem hhv. hovedstol,

renter og omkostninger. Illustreres simplificeret nedenfor:

Kontotype 2

Lån (inkl. renter og om-

kostninger/gebyrer):

DKK 155.000 (hovedstol)

+ DKK 10.000 (renter)

+ DKK 5.000 (omk./geby-

rer)

Kontotype 3

DKK

170.000

DCS

DKK

170.000

(registreret i kontotype 3

i DCS)

Før 1992:

Tilgodehavender der blev afskrevet (overført til kontotype 3) blev overført til en såkaldt "hoved-

bogskonto". Denne "hovedbogskonto" var en samlekonto for samtlige af de i BG Bank værende

"afskrevne fordringer". Rentetilskrivning kunne således ikke foretages ligesom eventuelle afdrag

ikke direkte kunne registreres.

En kunde i BG Bank modtog årligt en oversigt over de løbende mellemværende. Imidlertid kunne

gæld, der var blevet tilskrevet kontotype 3, ikke indeholdes i disse årlige oversigter. Gælden til-

skrevet kontotype 3 blev således skrevet ud af årsoversigten - og kunden kunne således ikke iden-

tificere den pågældende gæld til banken. Kunden modtog således ikke systematisk en årlig oversigt

over gældsforholdets udvikling. Udarbejdelsen af en årsoversigt vedr. gæld tilskrevet kontotype 3

krævede manuel sagsbehandling og blev ikke konsekvent generet.

Efter 1992:

Medio 1992 blev det EDB-mæssigt muligt at håndtere de afskrevne fordringer på kontotype 3 på

separate konti, hvorfor gælden kunne registreres med en samlet saldo på enkeltstående konti.

Kunderne modtog herefter også årsoversigter vedr. gæld tilskrevet kontotype 3.

I de tilfælde hvor en konto havde flere end to interessenter (meddebitorer), var det ikke muligt at

sende årsoversigter til alle øvrige interessenter. Det var kun muligt at sende årsoversigt til én

person, hvis der var mere end to interessenter tilknyttet en konto.

Perioden 1999-2004

I perioden 1999-2004 var der i BG Bank etableret et note-baseret system. I dette system var der

oprettet en række underrubrikker på kontotype 3, hvor man sondrede mellem renter, anerkendte

renter, omkostninger, gebyrer og resthovedstol. Ved overgangen til DCS blev disse underrubrikker

i al væsentlighed ikke registreret i DCS. I BG Bank havde man mange forskellige registreringsmu-

ligheder i disse underrubrikker, hvilke mange registreringsmuligheder migrationen ikke tog højde

for.

Den endelige migration til DCS var mangelfuld, og alene dele af de registrerede anerkendte renter

blev overført til DCS's datafelt 4. Efter 2004 er underrubrikkerne i BG Banks system ikke ajourført.

5.2.2

Registrering af debitorer og kautionister

På de afskrevne engagementer på kontotype 3, hvor en eller flere kautionister helt eller delvist

hæftede for engagementet, blev sådanne kautionister systemmæssigt (urigtigt) registreret som en

meddebitor. Denne systemmæssige registrering var nødvendig for, at der kunne generes årsover-

sigter. Årsoversigten anvendtes bl.a. til at afbryde forældelse efter Danske Lov 5-14-4.

Hvis der systemmæssigt blev registreret mere end to meddebitorer, kunne der imidlertid ikke ge-

nereres årsoversigter for alle interessenter. Det var kun muligt at genere én årsoversigt, hvis der

var mere end to interessenter. Der blev således kun udsendt årsoversigt til én interessent (den der

var registreret med skattekode 1) med den risiko, at forældelse ikke blev afbrudt over for kautio-

nister.

Registreringen indebærer, at man ikke systemmæssigt kan se, at meddebitoren rent faktisk er en

kautionist - en sådan information fremgår alene af den fysiske kundemappe. Som konsekvens

medfører dette, at kautionisten - hvis datagrundlaget i systemet lægges til grund - kunne/kan blive

opkrævet uberettigede beløb, f.eks. hvis der bliver indgået en akkordordning med hovedmanden.

Et sådant restkrav kan i udgangspunktet ikke gøres gældende over for en kautionist. Danske Bank

har konstateret, at sådanne krav er inddrevet.

5.2.2.1

Et krav pr. konto - flere debitorer

Såfremt en fordring registreret i kontotype 3, havde to eller flere debitorer, og kravene over for de

respektive debitorer blev behandlet forskelligt, og der derfor var forskellige pengemæssige krav

overfor debitorerne, blev sagen ikke registreret korrekt. I BG Bank kunne der kun registreres et

krav pr. konto.

Hvis debitor 1 eksempelvis anerkendte en resthovedstol + renter, og debitor 2 tilsvarende aner-

kendte kravet, kan der potentielt rejses forskellige krav over for de respektive debitorer - i hvert

fald over tid i takt med forældelsen.

Eksempel: Debitor 1 underskriver et frivilligt forlig

Resthovedstol

Renter

TOTAL

DKK

100.000,00

15.000,00

115.000,00

Såfremt debitor 2 ikke tilsvarende anerkendte rentebeløbet forældes rentekravet efter den almin-

delige forældelsesfrist.

Systemet kunne alene håndtere ét krav. Systemmæssigt blev det højeste af de mulige krav regi-

streret. Dvs. i ovenstående tilfælde registreres kravet med DKK 115.000,00 - og dette krav ville -

hvis der ikke blev foretaget en manuel korrekt - også blive gjort gældende over for debitor 2.

Samme systemfejl gør sig gældende for kautioner. Dette er navnlig relevant i forhold til beløbsbe-

grænsede kautioner. Hvis der eksempelvis var vedtaget en beløbsbegrænset kaution på DKK

50.000 i ovenstående tilfælde, ville kravet overfor kautionisten blive registreret som værende DKK

115.000,00.

5.2.3

Skift af kontonumre

Frem til ca. år 1999/2000 var det ikke muligt at opretholde et oprindeligt kontonummer, når gælden

blev sendt til hhv. kontotype 2 og kontotype 3. Systemmæssigt oprettedes et nyt kontonummer,

og der var registreret en transport. Dette indebærer, at den historiske udvikling i engagementet

blev vanskeligt at spore.

5.2.4

Regionale forskelle i håndteringen af kontotype 3

En debitor med flere gældsposter til inddrivelse blev håndteret på forskelligvis regionalt i Danmark.

Overordnet kan det siges, at man på Sjælland opererede med en kontotype 3 pr. gældspost, mens

man i Jylland anvendte kontotype 3 som en "samlekonto" for en debitors samlede afskrevne gæld.

Se illustration nedenfor:

Kontotype 1

(f.eks. gældsbreve)

DKK 100.000

Kontotype 2

(f.eks. kassekredit)

DKK 50.000

Nyt kontonummer

Kontotype 2

DKK 100.000

Kontotype 3

DKK 150.000

Denne integrationsmodel har bl.a. medført, at de forskellige forældelsesfrister og starttidspunkter

for fordringernes respektive forældelsestidspunkter blev umulig at identificere. Problemet videre-

førtes ved indførelsen af DCS.

Hvis denne registrering manuelt ikke var blevet tilrettet ville den samlede saldo fra kontotype 3

blive overført til DCS-systemet.

5.2.5

Eksempel på et hændelsesforløb

Kontotype 1

Kontotype 2

Kontoen blev således "afskrevet" den 17. december 1996 med en saldo på DKK 117.536,24. Den

samlede saldo på DKK 117.536,24 blev overført til kontotype 3. Det konstateres, at der allerede

inden overførslen til kontotype 3 ikke var korrigeret for tilskrevne forældede renter (den gældende

forældelsesfrist for renter på daværende tidspunkt var 5 år regnet fra forfaldsdagen, og allerede

den 17. december 1996 var der tilskrevet DKK 14.170,15 i forældede renter.

Saldoen på kr. 117.536,24 har været registreret på kontotype 3 frem til konverteringen til DCS i

2004, og blev registreret som hovedstol i DCS.

Med virkning pr. 31. december 2018 blev sagen korrigeret, og hovedstolen blev nedskrevet til kr.

59.179,39, svarende til den sidste foretagne indbetaling den 1. august 1989. De kr. 59.179,39 kan

i dag tillægges 3 års rente (forudsat at forældelsen af hovedstolen er behørigt afbrudt).

Hvis debitor i dette scenarie eksempelvis den 31.12.2017 havde underskrevet et frivilligt forlig på

baggrund af kontotype 3, ville de 117.536,24 blive tillagt yderligere 3 års renter af en allerede

forkert opgjort restgæld.

I denne sag er der forudsætningsvist indgået enten (i) et frivilligt forlig eller (ii) opnået dom på

baggrund af en urigtigt opgjort restgæld.

Hvis der ikke var foretaget en manuel korrekt af denne sag, kunne der potentielt være fejlagtigt

inddrevet en beløb, der oversteg den reelle restgæld med mere end 100 %.

Resthovedstolen på alle tilgodehavender er forudsætningsvist - hvis sagen ikke er blevet manuelt

korrigeret - systemmæssigt registreret forkert. Det indebærer bl.a., at Danske Bank undertiden har

opkrævet mere end det skyldte beløb. Den forkerte opgørelse af resthovedstolen ses derfor også

reflekteret i f.eks.:

•

5.2.6

Indfriede fordringer,

Frivillige forlig,

Fogedopgørelser,

Domsfundamenter,

Advokatopgørelser, og

Gældssaneringssager, konkursboer, rekonstruktion, dødsboer etc.

Tilgodehavender til ekstern inddrivelse

Såfremt en fordring har været til inddrivelse ved ekstern advokat er det konstateret, at når kravet

tilbagesendes fra advokaten - f.eks. fordi der er opnået skylderklæring - registreredes fordringen

på ny i systemet. I kontotype 3 registreredes den samlede resthovedstol inkl. renter i 5 år og

fogedomkostninger. Den eksterne advokat havde typisk tilskrevet yderligere renter frem til over-

sendelsestidspunktet, hvilket tilsvarende blev registreret i kontotype 3 - dog som anerkendte om-

kostninger. Efterfølgende sagsbehandlingsskridt i tillid til systemet vil være forkerte.

5.2.7

Migrationsfejl

Ved konverteringen til DCS-systemet i 2004 blev det besluttet, at fastholde registreringen af kauti-

onister som meddebitorer som anført i punkt 5.2.2.

Ved konverteringen til DCS-systemet blev kun saldoen fra kontotype 3 konverteret til DCS. Elektro-

niske oplysninger vedr. kontotype 1 og kontotype 2 fremgår således ikke af DCS-systemet, og

sådanne oplysninger kan kun tilgås via den fysiske kundemappe. Tilsvarende fremgår der ikke

informationer om, at saldoen på kontotype 3 også udgøres af eksempelvis renter, gebyrer og om-

kostninger. I enkelte tilfælde er anerkendte renter blevet overført til DCS. Anerkendte renter kunne

i BG Bank registreres på flere forskellige måder. Selve migrationen til DCS inkluderede ikke alle

registreringsmetoder.

ANNEX 1 - OVERVIEW OF TIME-BARRING RULES

This is a high-level non-exhaustive overview of the key rules on statute of limitation.

CLAIM

LIMITA-

TION

PERIOD

SUSPENSION

CAUSE

SUSPEN-

SION RE-

SULT

INTERRUP-

TION CAUSE

INTERRUP-

TION RE-

SULT

PRIOR to 1 JANUARY 2008

Loans, over-draft ac-

counts, document of

indebtedness, settle-

ment out of court

and guarantees of

these obligations

20 years from

establishment

N/A

i) Debtor's Per-

ception

(

anerkendelse

)

of the claim

ii) Notification

to debtor

iii) Notice to an

estate

iv) Request of

appropriation

(

anmodning

om udlæg

)

Interest, fees, costs

and guarantee obli-

gations of these

5 years from

the due date

i) Unaccounta-

ble knowledge

of the claim or

the debtor's

residence.

ii) Substantive

negotiations

POST 1 JANUARY 2008

Loans, over-draft ac-

counts, document of

indebtedness, settle-

ment out of court

and guarantees of

these obligations

10 years from

the due date

i) DB's lack of

knowledge or

ought to know

of the claim or

of the debtor

Postponement

of limitation

period

i) Debtor's Per-

ception of the

claim

ii) DB's legal

actions to-

wards the

debtor.

Postponement

of limitation

period

i) Debtor's Per-

ception of the

claim

ii) DB's legal

actions to-

wards the

debtor.

New 3 years

limitation pe-

riod

New 10

years limita-

tion period

Postponement

of limitation

period

i) Debtor's Per-

ception of the

claim

ii) DB's legal

actions to-

wards the

debtor

New 5 years

limitation pe-

riod

New 20

years limita-

tion period

Interest, fees, costs

and guarantee obli-

gations of these

3 years from

the due date

i) DB's lack of

knowledge or

ought-

knowledge of

the claim or of

the debtor

Project Collection

Workshop 26 November 2019

Strictly confidential

–

legal privilege

Introduction

Key impact of system and process failures

Systems impacted

•

The Debt Collection System

("DCS")

Legacy Danske Bank system

(pre 2004)

BG Bank (pre 2004)

(PF and Pantebreve

–

to be

verified)

•

Affected areas

All customers within GRDM

•

General debt collection

Debt relief

Bankruptcy

Foreclosure sale

Decedent estate

Etc.

•

Risks

Regulatory sanctions and

liability

Prudent practice and conduct

FSA

TAX

GDPR

Reputation

•

Remediation

Reimbursement of suffered

losses

Developing a compliant debt

collection framework

Maintaining the reputation of

Danske Bank

High impact flaws

Examples of high impact flaws (pre and post 2004)

•

The field in the DCS that uses the statute of limitation for the principal amount may contain interest and/or fees although the timing barring period for these

types of obligations is shorter than the time barring period for the principal amount. This is an ongoing flaw, but with a particularly high impact in connection

with the migration of data into the DCS in 2004.

•

The time limitation for the principal and interest is generally counted from the time the debt is entered into DCS rather than from the time the debt was due.

•

Time-barred interest, costs and fee payments have incorrectly been included in settlement agreements entered with customers, and also in successful claims

made against customers in courts, estates etc.

The key impact of the system and process flaws

•

Customers in risk of overpaying

•

Customers have overpaid

•

Customers have paid debt that was time-barred

•

Danske Bank has collected debt that was not owed

•

Danske Bank has submitted inaccurate tax reports to the Tax Agency

•

Inability to accept payments from debtors who want to make total repayment on their outstanding debt

➢

It is not possible generally to "auto-correct" the incorrect debt calculations

DCS

Main flaws within DCS (irrespectively of the various migration flaws between DCS and legacy Danske Bank systems)

➢

Principal amount containing interest; ongoing issue

–

FLAW:

The data field in respect of the principal amount may incorrectly contain unpaid interest that have accrued prior to the debt being transferred to the DCS.

CONSEQUENCES:

•

Different types of amounts have been aggregated into a single amount that is then reflected as the 'principal amount' in the DCS, and subsequently treated as

such for time-barring purposes.

Inaccurate calculation and reporting of interest.

Potential collection of time-barred claims, and subsequently in risk of over-payment.

➢

Principal amount containing costs; (fixed on cases entered into DCS post July 2019)

–

FLAW:

When creating a new case in DCS, related costs (collection fees) that remain unpaid are added to the principal amount.

CONSEQUENCES:

•

Incorrect handling af time-barring.

•

Inaccurate calculation and reporting of interest.

•

Potential collection of time-barred claims, and subsequently in risk of over-payment.

➢

Time-barring counting period; ongoing issue

–

FLAW:

DCS starts to count the time-barring periods from the date a case is opened in the DCS

CONSEQUENCES:

•

The exact time for time-barring is untraceable.

•

Potential collection of time-barred claims, and subsequently in risk of over-payment.

BG Bank

Improper and incorrect migration of data into DCS (applicable to historic BG Bank cases)

➢

Single amount migration

–

FLAW:

At the migration between BG Bank and DCS the total debt registered in BG Bank was registered as a principal amount (field number 3 in DCS).

CONSEQUENCES:

•

Different types of amounts merged into a single amount registered as the 'principal amount' in the DCS.

Inaccurate calculation and reporting of interest.

Potential collection of time-barred claims, and subsequently in risk of over-payment.

➢

Guarantors and co-debtors - Settlement agreements

–

FLAW:

In BG Bank it was not possible to distinguish between guarantors and co-debtors. This limitation has been transferred into DCS.

–

•

CONSEQUENCES:

Unjustifiable collection of debt.

•

For example where a settlement agreement (for a lower amount than the initial debt) has been agreed with a debtor, attempts to collect the full amount

would still be made against co-debtors.

Inaccurate calculation and reporting of interest.

Potential collection of time-barred claims, and subsequently in risk of over-payment.

BG Bank

Improper and incorrect migration of data into DCS (applicable to historic BG Bank cases)

➢

Regional differences

–

FLAW:

Debt collection processes differed across regions of Denmark. The integration in DCS may be affected.

–

In Jutland it was primarily decided that once a debt was written-off it would be combined with other written-off debt (even where the types of debt may have

varied) belonging to the same customer

Impossible to track differing periods of limitation.

Different types of amounts merged into a single amount registered as the 'principal amount' in the DCS.

Inaccurate calculation and reporting of interest.

Potential collection of time-barred claims, and subsequently in risk of over-payment.

CONSEQUENCES:

•

➢

Double registration of debt

–

FLAW:

Joint liability debtors were registered on one single account. It was not possible to take into account any differences that might arise between the debtors in

respect of the outstanding debt.

–

By example, if one co-debtor entered into a settlement agreement, the system would also apply the new principal amount (which includes earlier interest and

collection costs) to the remaining co-debtor(s) and the guarantor, both of whom are not bound by the settlement agreement

Unjustifiable collection of debt.

Inaccurate calculation and reporting of interest.

Potential collection of time-barred claims, and subsequently in risk of over-payment.

CONSEQUENCES:

•

Legacy Danske Bank

Improper and incorrect migration of data into DCS (applicable to historic Danske Bank cases)

➢

Single amount migration

–

FLAW:

At the migration between Danske Bank and DCS the total debt registered in Danske Bank was registered as a principal amount (field number 3 in DCS).

CONSEQUENCES:

•

Different types of amounts merged into a single amount registered as the 'principal amount' in the DCS.

Inaccurate calculation and reporting of interest.

Potential collection of time-barred claims, and subsequently in risk of over-payment.

➢

Double registration of debt

–

FLAW:

Where a debt was assigned to multiple debtors and/or a guarantor it was necessary, due to system limitations, to open an account for each debtor and

guarantor (where applicable) each reflecting the debt owed.

CONSEQUENCES:

•

Unjustifiable collection of debt - the relevant debt was recorded more than once.

Different types of amounts merged into a single amount registered as the 'principal amount' in the DCS.

Inaccurate calculation and reporting of interest.

Potential collection of time-barred claims, and subsequently in risk of over-payment.

The link ('green X') connecting co-debtors and/or guarantors was not incorporated into the transfer and therefore co-debtors and guarantors became untraceable.

•

Potential double collection.

Key financial risks

Comments

Repayments to the customers

Depreciation of debt that is actually time

barred

Tax related liabilities

Sanctions from Tax Agency/SØIK

Sanctions from DFSA/SØIK

•

The obligation to repay customers will result in a net loss

•

Debt registered in DCS as 'owed', but which is actually time-barred will need to be depreciated

•

There is a risk that re-opening customers' annual tax assessments will result in "late payments" of the relevant

taxes. Such a cost will likely be borne by Danske Bank

•

Risk of being fined for having reported incorrect data to the Tax Agency (Skattestyrelsen)

•

Breach of conduct of business rules (god skik) for collecting debt that was statute-barred, or otherwise not owed

by the customer

•

Governance failures for not resolving the issues despite escalation efforts made by employees and the

continuation of debt collection whilst knowing that the calculations might not be correct

•

Governance failures for not having sufficient systems in place

Damages for losses suffered by

customers and other third parties

Fine for GDPR-non compliance

Work to fix issues may be costly

•

Customers and other third parties may claim against Danske Bank for damages for the consequences of

incorrect debt collection

•

The DCS contains incorrect data on customers' debt and other personal data in breach of GDPR.

•

Due to the number of customers impacted by the issue, the costs in fixing the issue will likely be sizeable. Work

on re-calculating debt correctly will take a lot of effort and resource. The need to resolve the impact on

customers' tax returns may be costly. Finding a new solution for debt collection will naturally also be costly (we

understand this aspect is already ongoing)

Q&A

Plesner draft - 13 December 2019

STRICTLY CONFIDENTIAL

–

LEGAL PRIVILEGE

MEMORANDUM

ON MAIN ROOT CAUSES AND DANSKE BANK'S LEGAL OBLIGATION TO COMPENSATE CUSTOMERS

TABLE OF CONTENTS

INTRODUCTION .................................................................................................................... 3

DEFINITIONS ........................................................................................................................ 4

MAIN ROOT CAUSES .............................................................................................................. 4

3.1

3.2

3.3

3.4

Root cause 1: Interest, fees and costs incorrectly added to principal amount field .......... 4

Root cause 2: Incorrect information of Due Date ......................................................... 5

Root cause 3: Guarantors from BG Bank have been treated like co-debtors ................... 6

Root cause 4: Missing link between principal debtor, co-debtors and guarantors ............ 6

OBLIGATION TO COMPENSATE CUSTOMERS ........................................................................... 7

4.1

4.2

4.3

4.4

4.5

4.6

Right of restitution .................................................................................................... 7

Time-barring............................................................................................................. 8

Set-off right or obligation to pay out debt collected on an unjustified basis? .................. 9

Interest on Compensation amounts ............................................................................ 9

Regulatory conduct of business requirements ............................................................ 10

Specific situations assessed ..................................................................................... 10

ISSUES NOT CONSIDERED ................................................................................................... 11

Annex 1 - List of agreed terms

MEMORANDUM

ON MAIN ROOT CAUSES AND DANSKE BANK'S LEGAL OBLIGATION TO COMPENSATE CUSTOMERS

INTRODUCTION

We have been instructed by Danske Bank A/S ("Danske Bank") to prepare this memorandum set-

ting out:

1) the main root causes that in our view have caused Danske Bank to collect debt on an

unjustified basis from customers within the Group Recovery & Debt Management

("GRDM") business area; and

2) our assessment of to which extent Danske Bank is legally obliged to Compensate the cus-

tomers from which Danske Bank has collected debt on an unjustified basis due to the root

causes.

The purpose of this memorandum is to serve as the basis for a data analysis to be carried out by

Ernst & Young ("EY") that will enable EY to make a rough estimate (subject to certain assumptions

and qualifications) of:

1) the total amount of debt Danske Bank on an unjustified basis has collected as a result of

the root causes; and

2) the total amount Danske Bank will need to Compensate the customers due to an unjusti-

fied collection of debt.

The exact scope of the estimate to be made by EY will of course be agreed between EY and Danske

Bank and the above just reflects our preliminary understanding of the scope of the estimate.

Together with EY we have identified four main root causes, which in our view are those with the

highest impact in terms of volume. These are the root causes described in this memorandum. We

note that we have also identified a number of other root causes in connection with our work on

this project. These other root causes are described in our separate draft memorandum entitled

"Description of key system and process flaws" dated 28 November 2019 (the "System and Process

Flaw Document").

The descriptions of the root causes set out in this memorandum have been simplified in order to

make it more operational and easier for EY to use the descriptions as the basis for their data

analysis work. We refer to the System and Process Flaw Document for more detailed and accurate

description of the root causes identified.

We have together with this memorandum also provided a separate memorandum on the relevant

Danish statute of limitation rules that are relevant to this mater.

This memorandum has been prepared under time constraints as we have just had a few days from

the initial instruction and only one day from final agreement on the exact scope of our work. We

have therefore not had sufficient time to assess and consider all the aspects that it, in our view, is

relevant to assess and consider before taking any decisions regarding remediation actions. How-

ever, we believe that our assessments and descriptions set out in this memorandum should be

sufficient for the purposes of EY's data analysis described above.

This memorandum is structured as follows: Part 2 sets out the definitions used in this memorandum

by reference to Annex 1. In Part 3 we have set out our descriptions of the four main root causes

identified. Part 4 contains our assessment of the extent of Danske Bank's obligation to Compensate

the customers and guarantors. Finally, we have in Part 5 listed the issues that we have not consid-

ered or taken into account in relation to this memorandum but which it, in our view, is relevant to

take into account or consider at a later point in time.

DEFINITIONS

Capitalised terms used but not defined in this memorandum have the meaning set out in Annex 1.

3.1

3.1.1

MAIN ROOT CAUSES

Root cause 1: Interest, fees and costs incorrectly added to principal amount field

Description of root cause

Interest, fees and costs have incorrectly been added to the principal amount field in the DCS (field

3). This root cause has been triggered by various events.

This root cause can be divided into:

(i) two ongoing flaws in DCS that have continued to occur for a longer period of time; and

(ii) a one-off migration event that occurred in 2004 when data was transferred into the DCS.

3.1.2

3.1.2.1

Ongoing flaws in the DCS

Interest incorrectly added to principal amount field in the DCS

When a debt collection case is created in the DCS, the data field in respect of the principal amount

may incorrectly contain unpaid interest that have accrued prior to the debt data being transferred

to the DCS. This means that the interest amount and the principal amount have been aggregated

into a single amount that is then reflected as the 'principal amount' in the DCS.

It is our understanding that this flaw mainly occurs in cases of partial repayments of interest.

However, there continues to be some uncertainties exactly when this flaw is triggered and what its

exact impact is in respect of the unpaid interest. Further, we also understand that EY has identified

some cases where debt data has been transferred to the DCS and where interest have been added

directly to the principal amount even though no partial repayments of interest has occurred.

Period of time

This flaw has to our understanding occurred on an ongoing basis since 2004 and continues to

occur.

3.1.2.2

Fees and costs incorrectly added to principal amount field in the DCS

When a debt collection case is created in the DCS, the standard collection fees (in Danish:

rykker-

gebyrer

), which is typically 3 x DKK 100, are generally incorrectly added to the principal amount

field in the DCS. This means that the collection fees and the principal amount have been aggregated

into a single amount that is then reflected as the 'principal amount' in the DCS.

It is also our understanding that certain other costs in some cases may incorrectly be added to the

principal amount field, e.g. certain lawyers' fees.

Period of time

The flaw has to our understanding occurred on an ongoing basis since 2004 and until July 2019.

3.1.3

3.1.3.1

One-off migration event in 2004

Interest, fees and costs added to principal amount field in the DCS

When data in 2004 was migrated into the DCS, the total debt (the principal amount, interest, fees

and costs) registered in each of Danske Bank's and BG Bank's systems was incorrectly transferred

as single field integration into the principal amount field in the DCS.

Period of time

This flaw originates from the migration of data in 2004, and this flaw should therefore only occur

on migrated cases. However, it likely impacted a high percentage of the cases being transferred to

the DCS.

3.1.4

Consequences of root cause

The key consequences of this root cause are:

•

Since the DCS is programmed to apply a 10-year time-barring period to the principal

amount field, the incorrectly added interest, fees and costs - in respect of which only a 3-

year time-barring period apply - will not be treated as time-barred by the DCS before 10

years.

Danske Bank will unjustified have collected time-barred interest, fees and costs.

The DCS will unjustified have calculated additional interest based on too high an amount

since interest is calculated on the basis of the principal amount stated in field 3.

Danske Bank may have reported incorrect data on customers' principal amount of debt

and interest payments to the Tax Agency (in Danish:

Skattestyrelsen

Danske Bank's data on the total amount owed by the customers in the DCS will reflect a

higher amount than actually owed by the customers.

•

3.2

Root cause 2: Incorrect information of Due Date

Pursuant to the Limitation Act, the time-barring period in respect of a principal amount, accrued

interest, fees and costs should be counted from the Due Date of the relevant debt.

In order to correctly keep track of when a debt becomes time-barred, it is as a starting point

necessary to know:

(i) the time-barring period that applies to the relevant type of debt; and

(ii) when the Due Date of the relevant debt occurred.

However, information on the correct Due Date is not necessarily transferred to the DCS. Rather,

the due date registered in the DCS, and which therefore is used to calculate when the debt is time-

barred, is the date on which the debt data is transferred into the DCS.

This means that the DCS starts to calculate the relevant time-barring period from an incorrect date

later in time than the due date, and consequently the debt is registered time-barred at a later point

in time than it actually should be under the Limitation Act. This is mainly a risk in respect of time-

barring of the principal amount where Danske Bank (or a debt collection agency) may have to

interrupt the time-barring of the principal amount. If Danske Bank takes such action close to the

expiration time of the time-barring period, the principal may in reality already be time-barred even

though the DCS shows that there still is some time before the principal is time-barred.

Period of time

This flaw has to our understanding occurred on an ongoing basis since 2004 and continues to

occur.

3.2.1

Consequences of root cause

The key consequences of this root cause are:

•

The debt shown in the DCS may actually be time-barred because the Due Date used to

calculate the beginning of the time-barring period is a later date than what is actually the

case.

•

Danske Bank will unjustified have collected time-barred debt.

Danske Bank's data on the total amount owed by the customers in the DCS will reflect a

higher amount than actually owed by the customers.

3.3

3.3.1

Root cause 3: Guarantors from BG Bank have been treated like co-debtors

One-off migration event in 2004

In BG Bank's system it was not possible to distinguish between guarantors and co-debtors. BG

Bank therefore manually kept track of whether an individual or entity was a guarantor (in Danish:

kautionist

garant

) by stating this in the relevant physical files.

Since there was no electronic information in BG Bank's system to identify whether a specific indi-

vidual or entity was a guarantor, no information about this was migrated into the DCS system in

2004. Therefore, after the migration to DCS, guarantors in BG Bank has incorrectly been reflected

and treated like co-debtors in the DCS system. Further, since the information in the DCS to an

increasing extent over the years has been relied upon rather than the physical files, the risk of this

data flaw materialising has increased over time.

The flaw means in practice that Danske Bank may not have observed the special requirements that

follow from an individual or entity being a guarantor. For example, Sections 47-48 of the Danish

Financial Business Act set out a number of notification obligations that must be observed by Danske

Bank in respect of guarantors in order for Danske Bank to maintain a claim against the guarantors.

The requirements apply both to commercial relationships and to non-commercial relationships,

although the rules are stricter in respect of the requirements vis-à-vis non-commercial relationships.

Period of time

This flaw only occurs on cases migrated into DCS in 2004.

3.3.2

Potential ongoing issue

We understand that EY has received information that this root cause may also have occurred after

2004 in the DCS. It is not currently clear to us if any such subsequent occurrences are limited to

manual errors only (i.e. non-systemic or processing errors). We have not received information

about this root cause being an "ongoing" issue so this is something that should be investigated

further.

3.3.3

Consequences of root cause

The key consequences of this root cause are:

•

Danske Bank may fully or partially have lost its rights against guarantors due to not having

observed the legal requirements applicable in respect of guarantors.

Danske Bank may incorrectly have reported data to the Tax Agency in respect of the guar-

antors as if they were co-debtors.

Danske Bank may have collected debt from guarantors which in fact was not covered by the

guarantee, e.g. if the guarantee was limited, or if debt-arrangements have been agreed with

the principal debtor.

3.4

3.4.1

Root cause 4: Missing link between principal debtor, co-debtors and guarantors

One-off migration event in 2004

In Danske Bank's pre-DCS system it was not possible to distinguish between guarantors and co-

debtors. Where a debt was assigned to multiple debtors and/or a guarantor the procedure was to

open an account for each debtor and the guarantor (where applicable) each reflecting the debt

owed. The relevant debt was accordingly recorded more than once.

Prior to DCS, these types of accounts were managed manually by making amendments to physical

ledgers associated with an account following account activity such as a payment, newly accrued

interest, costs etc. When Danske Bank received a payment from a debtor, the principal amount

would be reduced in line with the payment received on his/her ledger. Where a co-debtor or guar-

antor was attached to the same debt, the physical ledger would be marked with a green X, and

the payment received would also need to be registered, and the principal amount reduced, on the

corresponding physical ledger of that co-debtor or guarantor.

When the DCS was implemented and existing accounts were transferred into the DCS, the link (the

green X) connecting co-debtors and/or guarantors was not migrated into the DCS and therefore

the link between co-debtors and guarantors was lost. In other words, if a debtor made a payment

against a debt, a corresponding reduction that would be required on the account of a co-debtor

and/or a guarantor would not be made.

Period of time

This flaw only occurred when the migration of data took place in 2004.

3.4.2

Potential ongoing issue

We understand that EY has received information that this root cause may also have occurred after

2004 in the DCS. It is not currently clear to us if any such subsequent occurrences are limited to

manual errors only (i.e. non-systemic or processing errors). We have not received information

about this root cause being an "ongoing" issue so this is something that should be investigated

further.

3.4.3

Consequences of root cause

The key consequences of this root cause are:

•

Danske Bank may have collected debt that was not owed because the debt was recorded

more than once (potentially double payment).

Danske Bank have reported incorrect data on customers' principal amount of debt and in-

terest payments to the Tax Agency.

Danske Bank's data on the total amount owed by the customers in the DCS may reflect a

higher amount than actually owed by the customers (because repayments e.g. by co-debtors

may not be registered in respect of the other co-debtors).

4.1

OBLIGATION TO COMPENSATE CUSTOMERS

Right of restitution

Under Danish law, if a person (or entity) has paid an amount due to a mistake, misunderstanding

or as a result of an error, the starting point is that the person has a right of restitution (

condictio

indebiti

) against the other person (or entity) to which the payment has been made.

Based on case law from the Danish courts, the starting point with the payer's right of restitution

may in some cases change to the effect that the payer losses the right of restitution. The typical

examples of cases where the payer has lost the right of restitution, regardless of the starting point,

is where the payment (i) is made due to a mistake attributable to the payer; and (ii) the receiver

of the payment was in good faith.

In situations where the fault of the mistaken payment lies with the receiver of the payment (as it

will be the case in the situation at hand in respect of Danske Bank), it is clear from case law from

the Danish Supreme Court that the payer will be entitled to restitution.

Based on this, it is generally our view that the customers (or guarantors) that have paid amounts

that were time-barred or that were not actually owed to Danske Bank have a right of restitution.

The fact that the customers (or guarantors) have actually paid the amounts without any objections

does not alter this.

In cases where it due to the specific circumstances would be unreasonable or very burdensome for

Danske Bank to Compensate a customer, Danske Bank may have an argument to avoid having to

Compensate the customer. However, we generally do not believe Danske Bank will be able to avoid

Compensations based on this argument. Accordingly, at least for the purposes of EY's data analysis,

we suggest that this possible exemption is disregarded.

4.2

4.2.1

Time-barring

General background to the applicable time-barring rules

Currently, the Danish time-barring rules are set out in the Limitation Act, which came into force on

1 January 2008. Before the Limitation Act entered into force, the Danish time-barring rules were

set out in the two acts DL 5-14-4 and the 1908 Act. Both of these two acts were replaced by the

Limitation Act on 1 January 2008.

The transitional rules applied during the period from 1 January 2008 to 1 January 2011. All claims

established before 1 January 2008 will be time-barred by now, unless Suspension or Interruption

have continuously been and are duly constituted. All relevant claims are now regulated by the

Limitation Act exclusively.

For further details on the time-barring rules, please refer to our separate memorandum on this.

4.2.2

The time-barring rules applicable to the customers restitution claims against Danske Bank

Under the Limitation Act, the general rule on time barring of debt provides that debt will be time-

barred after three (3) years, pursuant to the 3(1) of the Limitation Act. Accordingly, this will be the

time-barring period unless a different time-barring period is specified in the Limitation Act. The

period begins from the Due Date pursuant to Section 2(1) of the Limitation Act.

The principal amount of a loan is subject to a time-barring period of ten (10) years.

In the examples where a customer has paid a debt to Danske Bank that was actually time-barred

or was not actually owed, and the customer therefore gets a restitution claim against Danske Bank,

as stated in Part 4.1 above, the time-barring will as a starting point be three (3) years as no special

time-barring period is specified in respect of such claim. This will also be the case in situations

where the payment is not made by the customer but by way of distribution of a divided from a

bankruptcy estate, an estate of a deceased person or from a customer in debt relief on the basis

of an unjustified claim for the debt submitted by Danske Bank.

Regardless that the time-barring period is three (3) years in respect of the customers' restitution

claims, the time-barring period is suspended if the customer (or the bankruptcy estate or the estate

of deceased person etc.) is not and ought not to be aware of the restitution claim. However, the

absolute limitation period - regardless of whether the time-barring period is suspended - is ten (10)

years, pursuant to Section 3(2), no. 4.

In our view, customers' or guarantors' restitution claims against Danske Bank will be subject to the

general time-barring period of three (3) years, but the time-barring period will be suspended for a

maximum of ten (10) years until the customer becomes or ought to become aware of its restitution

claim.

There may be special cases where Danske Bank perhaps would be able to argue that the customer

ought to have been aware that the payment(s) made to Danske Bank were not correct as they

were time-barred or not actually owed. However, we only expect this to be relevant in very few

cases.

Therefore, for the purposes of EY's data analysis, we recommend that it is assumed that the time-

barring period of customers' restitution claims against Danske Bank will be suspended which means

that the model should consider all restitution claims that have a Due Date less than 10 years back

will not be time-barred.

We note that there may be special cases where Danske Bank from a legal perspective would be

legally obliged to go a bit further back in time than 10 years, but this will most likely only be a in a

limited number of cases. We do therefore not believe this is something that is relevant for EY to

take into account in their data analysis at this point in time.

4.2.3

Special rules on time limitation in respect of certain claims under criminal cases

As described above, claims in respect of restitution (

condictio indebiti

) and liability for damages

against Danske Bank will be time-barred at the latest after 10 years of the unjustified collection.

In that connection we note that it may be possible for customers to file an already time-barred

claim against Danske Bank according to the rules on Ancillary Proceedings.

In these Ancillary Proceedings, time-barring of claims does not prevent a defendant from being

liable to pay damages or Compensation to the injured in connection with the criminal case where

the accused is found guilty. Such a claim can also be asserted under a separate lawsuit within 1

year of final decision in a criminal case where the defendant is found guilty, or within 1 year of the

defendant adoption of a fine or other criminal sanction.

The decisive point is that a criminal case is actually carried out, and by implication that the specific

injured can be identified - however, it is unclear whether a more general decision will be sufficient

to initiate such a process. The rules on time-barring is regulated in respect of the time-barring

period for criminal charges. For the identified potential criminal offence there is a limitation period

of 10 years.

Claims that have been time-barred before the time-barring of the criminal charges cannot by im-

plication be involved in such Ancillary Proceedings. At this moment initiation of Ancillary Proceed-

ings seems unlikely.

4.3

Set-off right or obligation to pay out debt collected on an unjustified basis?

If (i) the customer has a valid claim (for restitution or liability for damages) against Danske Bank,

and (ii) Danske Bank has a valid claim against the customer, it is our preliminary assessment that

Danske Bank will be entitled to set-off the customers claim against Danske Bank in Danske Bank's

claim towards the customer.

Due to the limited time for our input, we have not been able to analyse whether such a set-off

would be permitted under the applicable regulatory conduct of business requirements, including

prudent debt collection practice, good business practice etc. Further, we have also not conducted

an analysis of any tax-related impact.

4.4

Interest on Compensation amounts

In the situations where Danske Bank is obliged to Compensate the customer for a loss, interest will

as a starting point accrue from the date the customer paid the debt Danske Bank was not entitled

to collect.

The rate of interest is determined pursuant to Section 5 of the Danish Interest Act (at present

8.05%).

It is possible it will be relevant to distinguish between situations where Danske Bank is able to set-

off the customer's claim against other debt owed by the customer and situations where Danske

Bank will actually need to pay out an amount to the customer. We have not had time to analyse

this in detail yet.

4.5

Regulatory conduct of business requirements

Section 43 of the Financial Business Act provides that Danske Banks must be operated in accord-

ance with honest business principles and good practice within the field of activity. This obligation

is fleshed out in Executive Order on Good Business Conduct.

We have considered whether this requirement entails a requirement for Danske Bank to Compen-

sate customers in respect of those of the customers' claims against Danske Bank that have become

time-barred. Our preliminary view is that that no such Compensation requirement will apply in that

case as we would expect the time-barring rules of the Limitation Act to prevail over any obligation

to compensate customers pursuant to the conduct of business requirements.

We have also considered whether Danske Bank will have an obligation to notify those of the cus-

tomers that have been impacted by the root causes and in respect of which Danske Bank has

collected debt Danske Bank was not entitled to. Our preliminary view is in that respect that Danske

Bank will be obliged to make such notification unless it would unreasonably cumbersome. We

recommend that this is analysed in further details and any decision in this respect should probably

be discussed with the Danish FSA.

4.6

4.6.1

Specific situations assessed

Introduction

We have below set out our assessment in respect of Danske Bank's obligation to Compensate

customers (or guarantors) in the following four situations:

1) if Danske Bank has collected debt that was time-barred;

2) if Danske Bank has collected debt that was not owed;

3) if a customer has suffered a loss due to a claim from the Tax Agency resulting from an

incorrect reporting of data by Danske Bank; and

4) if Danske Bank is not able to evidence with a customer's debt has become time-barred.

4.6.2

Situation 1 - Collection of debt that was time-barred

The customer will have a restitution claim against Danske Bank in a situation where Danske Bank

has collected debt that was time-barred.

The customer's claim is subject to the general three (3) years' time-barring period, but the period

will be suspended for ten (10) years because the customer will typically not be aware of the resti-

tution claim.

4.6.3

Situation 2 - Danske Bank has collected debt that was not owed

The customer will have a restitution claim against Danske Bank in a situation where Danske Bank

has collected debt that was not actually owed.

The customer's claim is subject to the general three (3) years' time-barring period, but the period

will be suspended for ten (10) years because the customer will typically not be aware of the resti-

tution claim.

4.6.4

Situation 3 - Loss due to claim from the Tax Agency resulting from incorrect data reporting

Since Danske Bank has reported incorrect data in respect of the customers' debt owed to Danske

Bank and the interest payments made, the Tax Agency may claim damages from customers that

have received too high a tax deduction as a result of the incorrect data reported by Danske Bank.

The customers will in turn get a recourse claim against Danske Bank for the same amount as the

amount claimed by the Tax Agency.

In our view, the Tax Agency will most likely be able to resume the tax assessment of customers 10

years. In the cases where the Tax Agency has a claim going up to 10 years back, the Tax Agency

will also have a right to claim interest on the amount up to 10 years back. Detailed rules apply in

respect of determining the restitution level, see Section 62a of kildeskatteloven.

[@EY, in respect of the impact assessment, please note that Section 5(8) of Ligningsloven provides

that: "

Renteudgifter mv., jf. stk. 1, kan først fradrages ved opgørelsen af den skattepligtige ind-

komst i det indkomstår, hvori betaling sker, hvis renteudgifter mv. for tidligere indkomstår i samme

gældsforhold ikke er betalt inden udgangen af indkomståret

".]

4.6.5

Situation 4 - Unable to evidence if debt is time-barred or not

In situations where Danske Bank is not able to evidence whether a customer's debt is time-barred

or not, e.g. because Danske Bank cannot prove when the Due Date of the debt is, Danske Bank

will in our view have to waive the claim for the debt.

ISSUES NOT CONSIDERED

We have considered the following issues out of the scope of this memorandum:

•

Any claims for indirect or consequential damages, e.g. customers that have suffered an

indirect loss as a result of Danske Bank having collected an amount Danske Bank was not

entitled to;

Any compensation that Danske Bank may be obliged to pay as a result of Danske Bank

having entered into a settlement agreement or repayment agreement with a customer on

the basis of Danske Bank stating that the customer's debt was X amount while it actually

was a lower amount, e.g. if it has appeared to the customer that the customer received a

discount of 25% while the customer actually - due to the incorrect debt information - only

received a discount of 10%. The customer may in such situation - depending on the spe-

cific facts - be able to claim to be put in a situation as if a 25% discount had actually been

granted. This may be relevant to analyse further at a later point in time;

Issues arising out of or in connection with PF or the "Home issue"; and

Any questions of foreign law.

•

ANNEX 1 - LIST OF AGREED TERMS

Below we have prepared a non-exhaustive list of defined terms relevant to this project. The list should be

updated on an ongoing basis based on the agreement between Danske Bank, EY and Plesner regarding which

terminology to use in connection with the project.

"Ancillary Proceeding"

"Barring of Claim"

"Co-debtor"

"Compensation"

"Debt Relief"

"Document of Indebt-

edness"

"Due Date"

"Dunning notice"

means a judicial enforcement of a civil claim filed and growing out of the prose-

cution of a criminal offence. In Danish "

adhæsionsproces

means preclusion of a claim. In Danish "

præklusion

means one of two or more debtors who are joint and separate liable to the same

debt.

means set-off, refund or repayment.

means in court debt-restructuring. In Danish "

gældssanering

means an instrument of debt and serves as a legally enforceable evidence of a

debt and the promise of its timely repayment. In Danish "

Gældsbrev

means the last day for payment without unpaid amounts being subject to a late

payment charge or additional collection efforts. In Danish "

forfaldsdag

means a notification sent to a customer, stating that it is overdue in paying an

account receivable and where a fee off DKK 100 is attributed to the debt. In

Danish "

Rykkerbrev

means executive order no. 330 of 7 April 2016, as amended.

"Executive Order on

Good Business Con-

duct"

"Guarantor"

"Interruption"

"Limitation Period"

"Liquidator"

"Overdraft

Account"

and "Overdraft Facil-

ity"

"Perception"

"Proff of Claim"

"Root cause"

"Suspension"

"Time-barred"

means a person who guarantees to pay a borrower's debt in the event the bor-

rower defaults on a loan obligation (totally or partially).

means a break in the continuity of the limitation period.

means a certain period limited by the applicable rules on statute of limitation

after which debt collection cannot be enforced.

means either (i) a person appointed by the shareholders to liquidate a company,

(ii) a person appointed by the probate court to liquidate a company.

means a credit agreement made with a financial institution that permits an ac-

count holder to use or withdraw more than they have in their account, without

exceeding a specified maximum negative balance. In Danish "

Kassekredit

means the debtor's admission of the existence of the debt. In Danish "erkendelse

af gæld".

means a notice of claim in an estate (bankruptcy, decedent estate, debt relief,

etc).

means the actual reason for the occurrence of a problem.

means postponement of the limitation period.

means barred by the passage of time under a statute of limitations. In Danish

forældet

DRAFT

Phase 2:

Data analysis and considerations within debt collection

Project Athens

–

01-July-2020

Version 0.3

Project Athens

Introduction

•

The following analysis has been performed in accordance with instructions from Danske Bank A/S (Danske Bank) and founded upon the description of main root causes

prepared by Plesner Advokatpartnerselskab (Plesner

set out in their memorandum of

December

9 the Main Root Cause Memorandum

•

Ernst & Young (EY) have been tasked by Danske Bank to estimate and execute the redress of customers affected by the flaws identified within Group Recovery & Debt

Management (GRDM), specifically relating to data quality issues in the two centralized debt collection systems; Debt Collection System (DCS) and Personlig Fordring (PF)

•

The flaws have affected the customers registered in the two systems which can be segmented into three main customer groups;

ii.

iii.

Customers with debt incorrectly registered in the systems, but have not made any repayments

Customers subject to incorrect collection of debt, however the over-collection

can be offset in the customers outstanding debt

Customers subject to incorrect collection of debt that are eligible to receive compensation

•

The scope of the data analysis and the focus of this report has been to identify affected customers that are likely to be eligible to receive compensation as a direct

consequence of the main root causes (customer group (iii)), as well as estimate the total financial impact of such remediation. Further, EY has been tasked to assist

Danske Bank in executing the outpayment to affected customers

•

The reasoning behind the approach outlined abo e relies on Danske Bank s ambition to compensate customers, that ha e suffered

a direct financial loss due to root

causes, as soon as possible while being mindful of the important task at hand of rectifying every customer that have been impacted by root causes, although it has not led

to a direct net financial loss (i.e. customer groups (i) and (ii))

•

The remediation of customer groups (i) and (ii) will focus on correction of cases while notifying affected customers respectively that their debt has been corrected or that

their claim against Danske Bank have been offset in their outstanding debt

•

The programme governance structure has been revised in June 2020 to ensure a clearer remediation focus and speed of execution with a programme timeline for

completion set to end of Q3 2021, subject to changes pending dialogue with authorities and further calculations. See pp. 23 and 24 respectively

•

The remediation of customer groups (i) and (ii) will be executed by a dedicated workstream

Portfolio trimming and Clean up that ill identif , rectif and inform the

selected customers impacted by root causes (i.e. debt incorrectly registered)

Athens Remediation Analysis - July-2020 - Version 0.3

Project Athens

Key terminology and definitions

•

Compensation

‒

Means either set-off, refund, redress, restitution or repayment (dependent on context)

•

Due date

‒

The last day for payment without unpaid amounts being subject to a late payment charge or additional collection efforts

•

Over-collection

‒

Amount of the collected debt that Danske Bank did not have a legal justification to claim (due to i.e. wrongfully calculated outstanding debt and partial (or total) time-

barring of debt)

•

Overpayment

‒

Identical definition to over-collection, taking the customer's perspective

•

Redress

‒

Total compensation that Danske Bank is obliged to pay affected customers (adhering to remediation principles approved by Project Athens Steering Committee) that

have suffered a direct financial loss (over-collection). Total redress amount can include compensation due to direct financial loss, indirect loss and a time-value

component.

•

Offset

‒

If the customer's existing debt exceeds the prescribed redress amount, then Danske Bank will be entitled to offset the customer's claim against Danske Bank as

outlined in Plesner draft memos (13. December 2019 and 8. April 2020). However, this assessment relies on the assumption that the opposing claims are originating

from the same legal relationship (i.e. connexity)

•

Time-barred

‒

Means barred by the passage of time under a statute of limitations

Athens Remediation Analysis - July-2020 - Version 0.3

Project Athens

Agenda

Page

Executive summary

Evaluation of correction process

Scope of the remediation

Customer segmentation

Financial impact

Remediation plan

Appendix

Question

1-5

1,2,3 & 5

Athens Remediation Analysis - July-2020 - Version 0.3

Project Athens

Executive Summary

•

Project Athens analysed the identified data quality related issues in

Group Recovery and Debt Management (GRDM) that led to two ORIS

filings in the summer 2019

•

402,000 customers had been through the two systems

152,000 of these could safely be removed as they belonged to other

legal entities, where manual processes had mitigated the impact

Further 105,000 customers could be removed as they never paid

anything on their debt, and hence could not have overpaid

•

Calculation of the potential redress to the remaining customers in scope

(86,000) requires manual review. To reduce manual workload a

repayment ratio of at least 80% is implemented leaving 43,150

customers for manual handling. The repayment ratio is set based on

empirical data on overpayment ratios, including uncertainty factor and

will exclude customers extremely unlikely to be in scope for redress

•

A project team from EY have led the data analysis and delivered

conclusions in two previous reports in end December 2019 and in end

•

January 2020

The external law firm, Plesner, has performed legal assessments under

•

instruction of the Bank and the analysis in this report builds on the legal

basis provided by them

•

See complete breakdown of customer redress actions and non-actions

Additional 24,000 customers could safely be removed as they entered

illustrated in decision tree on p. 14

the systems and paid their debt before any interest could have time

barred (within 3 years)

This report answers the following five questions:

Finally 15,000 customers had entered the two systems through

validated channels and confirmed to be correct

This leaves 106,000 customers at risk of over-collection and hence a

redress for direct loss

Can the project assure that the root causes currently are

sufficiently prevented from causing additional and refundable losses

for customers?

What is the scope of the redress the Bank is going to pay back to

the impacted customers?

How can the project segment the customers into cohorts?

What is the estimate of the financial impact of the redress?

How does EY suggest to execute the remediation?

In these previous reports, EY concluded the following:

•

Four root causes had affected about 106,000 customers, of which

estimated 24,000 could be entitled to a redress as the impact of the

root causes could be offset in the remaining debt for the rest. Examples

•

of the four root causes can be found in appendix pp. 27-30

The root causes relate to improper and incorrect integration of existing

loans into the IT systems supporting the debt collection process and

dating back to when the systems were introduced (1979 and 2004)

•

It is from this group of customers that EY estimates that 24,000 (23%)

customers have suffered direct losses that exceed their remaining debt,

and hence need to be redressed

The redress should also include indirect, consequential and tort related

losses that both the four root causes, but also a series of other practices

in GRDM had caused for the customer (e.g. wrongful reporting to the

RKI, customers having to take out expensive loans to pay their debt or

settlements made on incorrect data)

•

Because of the root causes, the Bank has potentially (i) collected debt

(fees, interest and principal) that potentially was time-barred at the time

of collection (ii) collected debt based on wrongly calculated principals

and/or (iii) treated guarantors and co-debtors as primary debtors,

leading to potential over-collection. The data analysis considers all three

types of over-collection (i,ii,iii)

•

The root causes has been isolated to two systems,

Debt Collection

System

(DCS) and

Personlig Fordring

(PF), and impacted Denmark only

Redress approach is dependent on data availability:

as manual processes mitigated the impact in other countries (pending

line approval)

•

EY will apply a data model approach to identify customers in scope for

redress only where sufficient electronic transaction data exists (i.e. after

Insufficient governance set-up within GRDM and controls in the

2010). The model is applicable for 20,000 customers in DCS.

internally developed systems had failed to catch incorrect input entered

Calculation of the potential redress to the remaining customers in scope

into the systems at the time of implementation as well as in the following

(86,000) requires manual review

years

Athens Remediation Analysis - July-2020 - Version 0.3

These indirect losses cannot be calculated, but should be treated in a

new claims handling process.

Project Athens

Executive Summary (cont.)

Question 1: Root causes sufficiently mitigated?

•

EY s assessment of the red-

/ green mitigation process introduced in

July 2019, has shown there are at least three instances where the

customer still can be over-collected from. These are:

‒

•

The customer is following a payment plan settled prior to July 2019

The customer is making an unscheduled /un-agreed

payment

An estate handling or a bankruptcy is completed before the Bank

resubmits the correct claim

•

in the systems. Pending 2

line review

•

home

fees:

In facilitating voluntary deficit property deals (DK:

underskudshandler

on behalf of customers, established process

within GRDM has been to not negotiate the fee proposal from real-estate

agent

home,

while fee proposals from other real-estate agents should be

negotiated. This may have led to customers paying too high a fee to

home,

eventually leading to a higher remaining debt after the property

was sold off. Although investigations into the matter has shown, that a

basis for comparison of hat the right fee should ha e been is a er

difficult process, it is assessed that this issue should lead to a redress

for the customers in relevant cases. However, as the issue is unrelated

•

to the root causes, the exercise of potentially redressing these

customers should be separated from the main redress execution in

Project Athens

•

Debt collection agencies:

On the outsourcing to DCAs, the four largest

providers (19k out of 40k cases) have been contacted and a

reconciliation of their data been undertaken. Conclusion is that the

outsourced customers are included among the 160,000 customers

mentioned on previous page. Data flows have however been flawed

flowing back from the DCAs, leading to potential mis-matches in the

active- / passive status. Up to 950 customers could potentially have

paid on their debt, without it being properly registered in DCS and hence

appearing as non-payers in DCS. These are recommended to be added

to the redress portfolio

An overview of the investigated five side issues stated above is included

on p. 9

Further risks and issues, e.g. incorrect charged court fees, potential

conduct related issues and incorrect settlements are currently being

investigated. See pp. 10-11 for an overview of remaining items and

concerns

The assessment has however shown that enforcement of the mitigation

process has been strict and that cases are being manually reviewed

prior to e.g. court proceedings/foreclosures to ensure proper mitigation

of the known data quality issues

EY has reviewed 100 new cases and not been able to identify any

among these that have led to over-collection, but EY concludes that the

process lacks controls and 2

line review. An internal risk assessment

has been initiated on 3

March 2020, and conclusions are pending

EY recommends independent review from 2

and 3

line and tight

process control going forward, but do not recommend that this should

stall the redress execution as no evidence of over-collection have been

found

•

Question 2: Scoping

•

Plesner

concludes that the four root causes have led to the majority of

•

the issues identified; however, five additional significant side issues have

been identified and assessed. These are: Impact on other countries,

broker fees from

home,

GDPR concerns, outsourcing to debt collection

agencies (DCAs) and credit rating (RKI) reporting. Additional (10) issues

and concerns exist and is currently being analysed and qualified

GDPR concerns:

Processing of incorrect data is not adherent to the

GDPR regulation. This in turn may lead to indirect losses for the

customers. It is the conclusion, that from a legal perspective the

processing of the incorrect data is not a data breach as defined b

Question 3

–

Customer cohorts

GDPR, and hence the Bank is not required under the GDPR to

proactively communicate to affected customers. Potential direct losses

•

402,000 customers (private and commercial) have been handled in the

for customers, stemming from the

Bank s

processing of flawed data, is

two systems. Of these the Bank can safely exclude:

already covered by the redress approach. Customers substantiating an

additional indirect loss (e.g. tort, cf. the GDPR, sec. 82 ) could be entitled

‒

Non-DK portfolio (152,000)

–

pending 2

line review

to an additional redress. Estimated size of potential fine (if any) is in the

‒

Customers ho ha en t paid an thing –

non-payers (105,000)

range of 7.5 mDKK, based on the very limited precedent available

‒

Customers who have been active less than 3 years in DCS after

RKI reporting:

The Bank is under no regulatory requirement to report

2004 and customers who entered PF less than 3 years ago– i.e. less

customers that are sent for debt collection to credit reference agencies,

than the limitation period for fees and interest. An conservative cut-

such as RKI. Customers, who are unrightfully (or with incorrect data)

off period of 2.5 years is applied for DCS (24,000)

reported to RKI, can claim a loss and should be entitled to a redress. As

the likelihood of the entire principal being time-barred is limited, EY

concludes that the majority of the customers have been reported

correctly to RKI, and only interest might be time-barred leading to

additional redress, provided that the customer can substantiate a loss

‒

Customers correctly calculated and entered into DCS and PF

(15,000)

(continued on next page)

EY concludes that:

•

Country dimension:

The four root causes have not had an impact

beyond Denmark as manual processes have mitigated the incorrect data

Memorandum on root causes and obligation to compensate customers - 13.12.2019.docx

Athens Remediation Analysis - July-2020 - Version 0.3

Project Athens

Executive Summary (cont.)

•

This leaves 106,000 customers in scope for potential over-collection

These can be split into 80,000 customers in DCS and 26,000

customers in PF

20,000 of the customers in DCS can be re-calculated as original loan

data exists electronically.

The remaining 86,000 (60,000 in DCS and 26,000 in PF) have to be

manually reviewed in order to determine whether the Bank has over-

collected from them

through a claims process built for the redress. These claims can come

from beyond the customer group at risk of over-collection, as customers

could have suffered losses aside from direct over-collection

–

e.g. by

agreeing to settlements on time-barred debt, by keeping customers in

•

debt beyond the time-barring or by paying legal and default fees that

should have been covered by the Bank

•

Based on experience from previous remediation project, EY assesses

that the risk of a large number of customers filing claims for indirect

losses is low

•

approval (and binding ruling) on taxation of redress and adjustments in

tax filings

It should be noted that re-assessment of the tax assessments will

involve substantial burdens and challenges for the customers, the

Danish Tax Authorities and for Danske Bank

EY recommends to align the redress methodology with the BAU

correction approach to insure consistent treatment of customers

independently of whether they need redress or not (e.g. during offset)

•

As the majority of these customers have not fully paid their debt and

since the corrections performed by the BAU team stay below 20% of the

principal, it is safe to assume that repayment rates below 80% of the

principal would allow over-collection to be offset in the remaining debt

•

Hence manual handling can be reduced to 43,150 customers

Customers with repayment ratio below 80% would be treated as a part

•

of the BAU, but would not be entitled to a redress

The cohorts for redress are therefore 20.000 customers that can be re-

calculated and 43,150 customers which cases have to be manually

•

reviewed

24,000 customers are estimated to be entitled to redress. 7,500 of

these are identified

and the remaining customers are expected to be

identified by manual handling

•

Question 4

–

Financial impact

•

The distribution of the over-collection is a skewed normal distribution

. Across both systems the total

with an average of 1,600 DKK

estimated over-collection (excl. interest compensation) sums up to 49.6

mDKK

–

see p. 18 for an overview of redress composition

•

To this should be added an unknown amount of indirect, consequential

and tort related losses that customers have suffered and reported

Plesner concludes in their memorandum Set-off right in relation to debt

collection on an unjustified basis , that

Danske Bank is legally obliged to

Question 5: Remediation plan

compensate customers according to the Danish Interest Act section 5

•

As the customer cohorts have been identified and the analysis signed off

by the Steering Committee, EY propose to initiate the execution of the

Customers that are eligible to receive interest compensation as part of

remediation

their redress either have no current debt, or no outstanding debt after

offset is performed. See p. 17

•

Due to the limited scope of customers (24,000 est.) and size of over-

collection (DKK 1,600 on average) a direct communication to the

The Tax impact of the redress paid to the customer has to be agreed

impacted customers is recommended

with the Tax Authority (probably in a binding ruling). Current assessment

is however that if the Bank redresses beyond the legal obligation of 10

‒

A series of training material and IT tools should be developed to

years, the redress will be taxable for the customers

ensure proper support for the first line organization to qualify the

dialogue with customers. The tool for the claims process would have

The incorrect tax filings on behalf of the customers have led to

to be developed, tested and deployed quickly and prior to Day Zero

customers, potentially having had to high deductions in their tax income

statements, which the Bank should repay to the Tax Authority on behalf

•

In the letters to the entitled customers they will be asked to substantiate

of the customers. This is however deemed a minor issue as correct,

indirect losses within a time limit (suggested 4-6 weeks). Further, the

internal

ta marking DK

Restance markering

ithin the DCS and PF

letter will contain information about the expected compensation amount

systems, has led to limited incorrect interests filings to the Tax

Authorities

•

A revised programme governance structure has been put in place to

effectively realize the ambition of completing the remediation of the two

Hence, with the level of uncertainty currently present, it is impossible to

customer groups not entitled to a compensation ((i) & (ii)) by end of Q2

estimate the total financial impact for the Bank of the potentially

2021

incorrect tax filings

Root cause specific e amples of customers ta handling and proposed

redress methodology are being presented to Tax Authority to achieve

Athens Remediation Analysis - July-2020 - Version 0.3

•

Although, customers

entering DCS in 2004 have to be manually reviewed regardless of repayment ratio due to root cause 3 and 4 |

See p. 12 for further description of estimate I

Moving number as further calculations are performed, further additional redress is possible due to indirect losses

Project Athens -

Question 1

Ensuring that no new over-collection cases are created is a prerequisite for executing a

remediation. EY review has identified a manual working process

•

Evaluating the current correction process* put in place since 16

of July 2019, revealed a manual but working processes that enforce control and avoid generation of

new over-collection cases

Process risks

EY has assessed associated risks and concludes the following:

The process has few controls and has not been subject to an internal risk assessment yet

•

Has the tap been turned off?

EY has not been able to find cases that

ould increase the olume of redress, but…

Assessment has shown that there is a risk that the current mitigation process

could generate new cases of over-collection

EY has reviewed 100 new cases and not been able to identify any amongst

these that have led to over-collection, but EY concludes that the process lacks

controls and 2nd line review

EY recommends independent review from 2nd and 3rd line and tight process

control going forward, but do not recommend that this should stall the redress

execution as no evidence of over-collection have been found

Local customer teams can immaturely mark Light cases green or use non-standardized operating

procedures compared to the procedures applied in the Correction Team

Resource constraints prevent the Correction Team from handling cases proactively which increases

the backlog of in-active cases which could expire down the line if not attended to

Seemingly correct green DCS cases could be wrong since quality control is conducted once a month

Customers could repay the debt based on information previously taken from Netbank or annual tax

return (rare but could happen) which could lead to over-collection

•

Customers could have been granted debt relief, which had been judged before 17

of June 2019, and

fulfills his/her obligations leaving Danske Bank with more money than they should

Untimely re-submission of estate and bankruptcy cases could cause the executor to process the case

based on incorrect claims

Customer

defaults and

enters Soft

Collection

Customer moved

from FEBOS to

DCS

DCS case is

automatically

marked red

Activate, review,

assign case and

notify customer

If the case is Light;

Local customer team

corrects case, from

backlog

Mark corrected

case green and

mail Hard

Collection

Receive correction

mail and try to

contact customer

Normal DCS flow

and debt

collection

GRDM

•

Soft collection

•

Hard collection

Termination of

engagement

•

Customer

•

Customer

representatives

(lawyers, family

member, advisor)

Customer

Payment

Receives notification

and may act and

cause a trigger

Trigger

examples

Phone / mail

Settlement request

Correction

Team

•

Appointed

Corrections Team

Backlog / non-active

portfolio of red-

marked cases

If the case is

complex; Correction

corrects case from

backlog

Mark corrected

case green and

mail Hard

Collection

Continuous quality

control of green-

marked cases

*See appendix pp.31-32 for process description and further details | **Process steps that poses the biggest risks have been marked with red in the process flow illustrated above

Athens Remediation Analysis - July-2020 - Version 0.3

Project Athens -

Question 2

Four root causes have led to the majority of the issues identified, however five side issues

have been identified and investigated

Description

•

home

fees

Documented practice within GRDM

never to negotiate fees with

home

when approving property deals

Undocumented practice when

approving property deals in insolvent

estates where GRDM have

recommended

home

to the estate

handler, despite the existence of

lower offers from competitors

•

GDPR concerns

Processing of incorrect data is not

adherent to the GDPR regulation and

a fine is plausible (2.5-7.5 mDKK)

Some customers could be entitled to

redress (10-20 tDKK), if they can

justify a loss i.e. tort

Processing of incorrect data does

not require the Bank to proactively

communicate to all customers

Country dimensions

•

The four root causes have not had

an impact beyond Denmark as

manual processes have mitigated

the incorrect data in DCS

The other side issues investigated in

Denmark can as result also be ruled

out

•

RKI reporting

Customers who are unrightfully (or

with incorrect data) reported to RKI,

can claim a loss and should be

entitled to a redress

The Bank is under no regulatory

requirement to report customers

that are sent for debt collection to

RKI

•

Court cases / DCAs

Court cases:

Legal proceedings

conducted prior to June 17th 2019

are potentially based on incorrect

documentation

DCAs:

Flaws in the data flows

between Danske Bank and DCAs

can lead to discrepancies in the

customer s pa ment status

•

Analysis Phases

•

Status

–

28-Feb

SteerCo decision to resolve

home

issue within a separate project,

coined Project Agency Fees

Quantitative analysis of property

deals in insolvent estates is not

started

Plesner is reviewing internal policies

for managing conflict of interest

•

Status

–

28-Feb

Approved remediation principles

entail treatment of GDPR losses as

indirect losses, where a process to

handle customer claims will be built

A panel will treat these customer

inquiries and asses documentation

of their indirect losses

•

Status

–

28-Feb

EY data team and local teams have

performed data sample checks

which found no cases eligible for

redress

All local teams have documented

their analysis, confirming there is no

issue and was signed off by

EY has compiled documentation on

SharePoint for 2

line review*

SteerCo decision to close the issue

regarding root causes in DCS

pending positive 2

line review

Once approved, Project Athens

(Regulatory Affairs) should notify

local FSAs

This will reduce remediation scope

to Denmark and allow discharge of

country representatives in Operating

Committee

•

Status

–

28-Feb

Approved remediation principles

entail treatment of losses related to

RKI reporting as indirect losses,

where a process to handle customer

claims will be built

A panel will treat these customer

inquiries and asses documentation

of their indirect losses

•

Status

–

28-Feb

Court cases:

Quantitative

assessment of closed cases with

legal action since 2004 is initiated

to estimate potential redress

amount

DCAs:

Data received from 2 DCAs.

See appendix p. 33 for initial

analysis and current findings

•

Comments

•

Plesner assessed that a case-by-

case approach is required to

determine whether Danske Bank

has incurred liability towards the

affected customers using

home

EY has built a financial model to

narrow the scope of potential

redress worthy property deals

(based on voluntary deals only)

Compensation can either be based

on individual assessment or model

•

Plesner has not covered the

potential consequences related to

personal data transferred to 3rd

parties

Danske Bank Group Deputy DPO

and Regulatory Affairs to liaise with

Plesner and coordinate

communication to the DPA and/or

DFSA to get sign-off on approach

•

EY concludes that the likelihood that

a customer has been reported to RKI

with a time-barred principal is very

low, and hence that the error is

confined to the root causes

•

Adjacent to Triage stream, three EY

resources have been onboarded to

uncover:

•

# of outsourced cases

•

# of cases with mismatch

•

Average deviation as % of

principal

EY recommends to include the

mismatched customer cohort in the

portfolio for redress, and investigate

further

Further concerns include incorrect charged legal fees, conduct related losses and incorrect settlements which will be further investigated. See next page

Athens Remediation Analysis - July-2020 - Version 0.3

Project Athens

- Question 2

Additional potential side issues analysed prior to Day Zero

•

EY has applied the same procedure utilized to investigate the Country Dimension issue, which overall involves

Planning, Analysing

and

Execution

Phases

‒

Planning: Mobilize dedicated team, draft action plan, conduct initial interviews and gather necessary data for analysis

‒

Analysing: Analyse and evaluate collected data within dedicated team and with business SMEs

‒

Execution: Finalize report and supporting audit trail and utilize report as a guideline for (potential) compensation and/or implementation according to remediation principles

•

Responsibility and completion of the last phase (Execution) is handed over to internal Danske Bank project managers

Investigation Topic

Legal fees allocated towards court cases

Interest rate on defaulted mortgage loan

Outsourcing to DCAs

Court cases involving other creditors

Legal fees incorrectly merged with principal

Treatment of ill customers

Description

–

based on output from work sessions conducted during Q4 2018 to Q3 2019 with representatives from GRDM, Group Non-Financial Risk and Compliance

•

Clarification needed on the process for how legal fees and allocated towards court cases and who is responsible (Danske Bank or customer) for the fees

Current understanding is that fees are paid by customers although they should be paid by Danske Bank

If the customer has obtained the mortgage from Realkredit Danmark (RD) through referral from local Danske Bank branch, then an agreement in place between RD and the bank stipulates

that the bank has to provide a 20% guarantee towards RD in case the mortgage defaults. The interest rate on the debt owed to RD and the bank will often differ

Issues of concern requiring further investigation into case handling, customer agreements, payment visibility, DCA financial reporting, proper reconciliation of data between the bank and

DCAs and data flow between DCAs and Danske Bank

Issue regarding cases which have gone to court with incorrect data and where decisions made on allocation of estates have impacted other creditors

Issue regarding cases where court fees have incorrectly merged with the principal amount (similar to RC1)

How do we process customers who have gone into default, then paid towards their debt after which they become ill rendering them not able to pay? Are these customers allowed to go into

default a second time?

Issue regarding cases where the correct procedures have not been followed in order to adhere to limitation rules, various products, effect on cases of not following happy-flow patterns, effect

on customer agreements if interests have affected principal, rates and market changes

Item of interest related to (1) supplementary booking process and (2) information flow between group finance and central bookkeeping. More clarity required on the topic of information flow

between the two

Identification of additional stakeholders (Group Finance). Further work required to understand full impact on Group Finance

Potential issue linked to process of how we manage markers: if markers are not in place and we provide credit to customers where we should not; or if markers remain for longer than

necessary and customers stay in bankruptcy for longer than required

Potential issue as Tableau data is used for various decision making processes (i.e. data feeds into risk assessments). Due to further lack of analytical competencies and GRDM expertise in

those who are reviewing such data, potential risk lies with what conclusions we are drawing out of the tableau data (in particular related to risk assessments)

Interest application in DCS & PF

Central bookkeeping and Group Finance

•

Risk markers practice

#10

Evaluation and use of Tableau data

•

Customer Concern

Technical Concern

Process Concern

Athens Remediation Analysis - July-2020 - Version 0.3

Project Athens -

Question 2

Status on remaining potential side issues

Phases

Overall status

–

(and path to green)

Planning

Analysis

Execution

Analysis

status

•

Planning phase complete Anal sis phase nearl done hile considering each issue s potential impact on redress

Confirmation that side issues will not impact Day Zero as pilot customers have been adjusted accordingly

Handover of non-Athens side issues with no impact on Day Zero or redress initiated

Impact on



Day Zero/Redress

Closed /

Open

Closing

Open

Closing

Side issue

#1 Legal fees allocated towards court cases

#2 Interest rate on defaulted mortgage loan

#3 Outsourcing to DCAs

#4 Court cases involving other creditors

#5 Legal fees incorrectly merged with principal

Comments

Sample check indicates that higher than allowed legal fees has been charged to some

customers. Implicated customers removed from pilot. No issue for PF

Review interest rate practice in PF. Handed over to

High impact on redress as the correct legal

fee requires case investigation

No impact



EY separate DCA report is finalized

Potential liability towards other creditors. Plesner has provided legal opinion on how to

address issue and potential liability towards third parties

Confirmed no systemic issue for DCS and PF

No legal conflict. Issue is related to conduct. A Conduct Risk policy draft is due year-end,

that will reference protection of vulnerable customers.

No systematic errors in interest application in DCS and PF. However, interest rate

practice in PF needs review (same outcome as side issue #2‘)

Root causes in DCS have potentially resulted in inaccurate reporting in the Bank’s

central bookkeeping system (KRS). PF needs further investigation

No impact on pilot payout, but will impact

redress in subsequent batches

No impact on Day Zero but redress of other

creditors is likely

No impact on Day Zero or redress

#6 Treatment of ill customers

#7 Interest application in DCS & PF

#8 Central bookkeeping and Group Finance

#9 Risk markers practice)

No impact

No impact on Day Zero or redress

Closing

Open

Closing



•

The Bank received an injunction based on its manual risk marker practice some years

ago. New SOPs and policies have been implemented to mitigate risks

#10 Evaluation and use of Tableau data

Known issue since 2018 and mitigating actions has been in place since. No formal risk

assessment seems to be conducted based on GRDM data exhibited in Tableau.

Separate track is already working to resolve data feed issues

No impact

Closing

•

On track

•

Requires Attention

Not on track, requires attention



•

= Impact on Day Zero

Completed

= Impact on redress

Athens Remediation Analysis - July-2020 - Version 0.3

Project Athens

- Question 3

Segmenting the customers into cohorts and splitting between direct and indirect losses

•

On top of direct losses from over-collection indirect losses should be redressed as well, which creates an

dependency upon a loss committee

•

Of all the customers (402k) who have been in the two systems, 106k have been affected by the four root

causes to such an extent that they have potentially been subject to over-collection

•

42.85k customers of these have paid less than 80% of their principal and impact from root causes are

assumed to be offset in their remaining debt

. These are therefore sent to the line organization for data

cleansing, but out of scope for redress

•

Of the remaining 63k customers, 20k of these can be recalculated as loan data exist in electronical

format

•

The last 43,15k have to be manually reviewed to determine whether they are entitled to a redress

•

These customers should be redressed based on the direct loss that the Bank have caused them,

however indirect losses exist beyond this customer group

•

Pp. 10-11 contain a summary of all concerns that potentially could have caused losses to the customers

(home fees, GDPR concerns, settlements, etc. are examples hereof)

•

EY has created a model that will identify the over-collected customers and calculate the direct loss that

the customer is entitled to in redress

•

This should form the basis for the redress, but all customers can potentially substantiate any indirect

loss they have suffered. As such, it is necessary to inform all customers that are identified with incorrect

debt registered in DCS and PF so the customer is able to state a claim of indirect loss

Dashed line

=Cohort is not entitled to

redress

Possible origin of customer claims

based on valid indirect loss leading

to additional redress

Recalculation

possible

20,000

<3 years

24,000

Repay < 80%

See p. 34 in

appendix

(2005-2020)

Repay > 80%

or entered DCS 2004

Send to BAU

42,850

Manual handling

43,150

At risk of over-collection

106,000

Entering DCS/PF

correctly

15,000

Scope for

redress cohort

Valid

Non-payers

Excluded as over-

collection is not

possible

Depending on claim

type and evaluation

Offset possible

Excluded as impact

from root causes can be

set off in the remaining

debt

Depending on claim

type and evaluation

In scope for

redress

Identified through EY

model and entitled to

redress

Depending on claim

type and evaluation

Time-barred

Non-DK

Portfolio

152,000

Non-payers

106,000

Direct loss

Indirect loss

See

Is included

Depending on claim

type and evaluation

All customers

in both systems

402,000

Athens Remediation Analysis - July-2020 - Version 0.3

p. 34 in appendix I

p. 35 in appendix for further description of the filter removing customers active less than 3 years in the systems

Project Athens

- Question 3

Customer segmentation

–

Top-down logic for identifying customers in scope and

at risk of customer impact

•

Before calculating potential impact, the customer population can be narrowed down to customers that are in scope, where calculations are feasible, and where monetary

corrections are needed

–

resulting in 24,000 customers that are at risk of an impact

Waterfall of customers potentially in scope

Number of customers in scope

450

402

Waterfall of customers ultimately at risk of impact

Number of customers at risk of over-collection

120

106

Customers in Customers

scope

where

sufficient data

is not

available and

repayment

ratio < 80%

DCS

Customers Customers

Possibly

with data not entitled to affected DCS

available or

redress

and PF

repayment

customers

ratio > 80%

Customers (thousands)

400

350

300

250

200

150

100

Total

Non-DK

Non-payers Customer Customers

Paid

Customers Customers

customers customers customers

who has active less principal

entering in need of

paid

than 3

with

DCS/PF monetary

possible

years,

correctly corrections

having

limitated

entered

fees and

DCS/PF

interest

after 2004

DCS

Both DCS and PF

181

105

145

108

121

106

333

152

250

100

38,5

24,0

Both DCS and PF

The waterfall above shows how the total population of customers in DCS and PF can be narrowed down to the

customers that are potentially in scope. Customers are filtered out if they are registered in DCS in other Group

countries (Non-DK customers),

ha en t paid an thing or ha e been acti e less than

ears in DCS after

2004

and entered PF less than 3 years ago, as well as customers where no issue was identified. The resulting

106,000 customers are considered in scope.

The waterfall above seeks to further reduce the number of customers to

focus on customers that are likely to receive a compensation. Estimation of

impact will be performed on these, taking potential offset in remaining debt

into account. Please note that customers entering DCS in 2004 are part of

the 63k cohort illustrated above regardless of their repayment ratio which is

due to the potential impact from root cause 3 and 4.

Athens Remediation Analysis - July-2020 - Version 0.3

Project Athens

- Question 3

It is now becoming possible to indicate the volumes and means to redress relevant

customers. Further analysis including manual review required to provide precise picture

Breakdown into customer redress actions or non-actions

Entire customer portfolio

–

no 10 years limitation

In scope

customers

Infliction point logic

Account for ability to apply advanced

modelling for analysis in DCS system

Apply empirical data about overpayment

ratios, including uncertainty factor, to

exclude customers extremely unlikely to

be in scope for redress

Apply detailed data modelling where

transaction data quality enable accurate

answers

Apply advanced data modelling to identify

customers relevant for redress

Apply empirical data about overpayment

ratios, including uncertainty factor, to

exclude customers extremely unlikely to

be in scope for redress

Analyse debt creation dates to identify

root cause 3 and 4 issues

For RC3, all previous BG cases

are

identifiable and are potentially exposed to

RC3

–

this amounts to 5,000

RC4 can not be readily identified in the data

but exists in the form of physical case

documents. These amounts to 6,000 cases

where 50% have been checked resulting in

10 cases in scope for redress

Athens Remediation Analysis - July-2020 - Version 0.3

106,000

PF system

DCS

26,000

80,000

Sufficient data

Insufficient data

Sufficient data

(after 2010)

Insufficient data

(before 2010)

25,000

1,000

20,000

60,000

Repayment >

80%

Repayment <

80%

Unconfirmed

Confirmed

redress

Certain of no

redress

No redress

identified

Repayment >=

80%

Repayment <

80%

5,149

Manual

handling for

redress

19,851

Not in scope

Handover to BAU

1,000

Manual

handling for

redress

7,500

4,000

Not in scope

Handover to BAU

8,500

Further adv.

data modelling

28,500

Manual

handling for

redress

31,500

Debt creation

dates

Pay out in pilot

5,149

1,000

28,500

31,500

Out of scope for

redress

~23-26,500

Total #customers for manual

handling: 43,149

PF (5,149 + 1,000) = 6,149

DCS (28,500 + 8,500) = 37,000

Possibility

of redress

Certain of

redress

Lessons learned from Louisiana:

Analysis will shift customers to the

two other groups

Root cause 3 &

~5-8,500

Further data

analysis

ii.

Uncertain

of redress

Not in scope

Handover to BAU

8,500

Legacy cases

from BG Bank. BG Bank became part of Danske Bank Group in 2001 as part of the merger with Realkredit Danmark. In 2007 BG bank was discontinued

Project Athens -

Question 4

Financial impact analysis - Bottom-up estimation of over-collection in DCS where re-

calculations are possible

For customers in DCS with

available FEBOS data 20,000

customers are at risk of over-

collection

Over-collection estimate: 11.6 mDKK

The estimate of over-collection amounts to approximately

11.6 mDKK excluding interest compensation (for time value of

money), but accounting for set-off

The individual over-collection is at an customer level and log-

normally distributed as shown below

Cohort characteristics:

The cohort consists of ~18,800

private customers and ~1,200

corporate

Out of these ~7,500 customers are

estimated to be entitled to redress

3.000

Distribution of estimated over-collection

DCS customers with available FEBOS data (7,383 customers)

2.845

2.346



= 2,126 DKK



= 11,982 DKK

Number of customers

2.500

2.000

1.500

1.000

500

0-150

150-

500

DCS is affected by

root cause 1-4

For the customers in scope where

transaction data is electronically

available, the customers entering DCS

earlier are more probable to be affected

by over-collection

1.372

230

500-

1,000

1,000-

3,000

291

100

6,000-

10,000

116

3,000-

6,000

10,000- 30,000- 50,000 <

30,000 50,000

Over-collection in DKK

Athens Remediation Analysis - July-2020 - Version 0.3

Project Athens -

Question 4

Financial impact analysis*

–

Estimated over-collection from customers in scope for

manual handling

For customers where manual handling is required, the total number of customers at risk

of over-collection is equal to 43,150

Over-collection estimate: 38 mDKK

DCS: 25 mDKK

PF: 13 mDKK

Cohort characteristics:

These customers have repaid at least 80% of their principal (except customers entering

DCS in 2004 that are potentially impacted by root cause 3 and 4)

37,000

customers are in DCS. However, assuming similar characteristics as cases with

available FEBOS data (previous page)

~11,500

customers are estimated for redress

6,150

are in PF and likely to be entitled to redress

DCS

4.000

Distribution of estimated over-collection**

Customers with repayment ratio > 80% estimated to be entitled to redress

(11,541 customers)

3.490

2.998



= 2,793 DKK



= 17,638 DKK

1200

Distribution of estimated over-collection

Customers with repayment ratio > 80% estimated to be entitled to

redress (5,149 customers across 3,743 cases

)

965

672

515

228



= 7,042 DKK



= 11,422 DKK

Number of customers

3.500

3.000

2.500

2.000

1.500

1.000

500

0-150

3.398

1000

914

800

600

935

333

150-

500

500-

1,000

1,000-

3,000

3,000-

6,000

240

400

200

111

146

173

6,000- 10,000- 30,000- 50,000 <

10,000 30,000 50,000

0 - 150

Over-collection in DKK

150 -

500

500 -

1.000

1.000 -

3.000

3.000 -

6.000

6.000 - 10.000 - 20.000 - 50.000 >

10.000 20.000 50.000

Over-collection in DKK

Athens Remediation Analysis - July-2020 - Version 0.3

*All over-collection estimates on this page is based on statistical estimates I

**The estimated over-collection is based on the assumption that the cases follow the average characteristics identified by the cases with available electronic FEBOS data

Project Athens -

Question 4

Costumers will be compensated for time-value of money by adding a interest

compensation to their redress

Plesner concludes that Danske Bank is legally obliged to compensate customers

according to the Danish Interest Act

section 5

Customers included in pilot will receive interest compensation accordingly

Interest compensation is calculated as simple interest (i.e. not compound)

The rate will equal the official lending rate (stated semi-annually) plus a surcharge of either 7% or 8% p.a. dependent on the

vintage of the customer’s claim (see below)

The lending rate varies significantly over time

–

se development illustrated in graph to the right

–

and have been equal to

0,5% p.a. since July 2015

See appendix p. 36 for more background on legal basis and implementation of interest compensation principles and

calculation of sample case on pp. 37-38

Lending rate

March 1 2013

7% + lending rate

Before

Customers with claims arising

before

the 1st of

March 2013 will receive official lending rate plus a

surcharge of

p.a. as interest compensation

8% + lending rate

After

Customers with claims arising

after

the 1st of March

2013 will receive official lending rate plus a surcharge

p.a. as interest compensation

Athens Remediation Analysis - July-2020 - Version 0.3

Customers that are eligible to receive interest compensation as part of their redress either have no current debt, or no outstanding debt after offset is performed

Project Athens

–

Question 4

Financial impact analysis

–

Redress composition

Over-collection

Over-collection is just one of the

parts of the redress amount for a

customer. Over-collection stems

from the four root causes identified

Interest compensation

Interest compensation is paid to

eligible customers as a time

compensation for the period that

Danske Bank had collected the

debt on an incorrect basis.

The estimate provided below

uses the minimum of when the

debt entered DCS or PF and 10

years.

Indirect losses

Indirect losses refer to any loss

incurred as a consequence of

identified root causes

Indirect losses will be gathered

through the claims process that the

programme will establish prior to

Day Zero

No statistical modelling has been

made to estimate the amount of

these indirect losses. While EY

expects that the number of claims

will be limited the claims that do

arise are, however, expected to be

significant. EY estimates the indirect

losses to 50 mDKK - although a lot

of uncertainty is tied to this number

For customers with electronically

available FEBOS data over-

collection is estimated to 11.6

mDKK

Customers in scope for manual

handling in PF, over-collection is

estimated to 13 mDKK

Customers in scope for manual

handling in DCS, over-collection is

estimated to 25 mDKK

See appendix p. 36 for more

background on legal basis and

implementation of interest

compensation principles.

Appendix pp. 37-38 contain

walkthrough of a sample

customer case from the pilot

batch case H incl. the interest

compensation calculation

Estimated interest compensation

for over-collection:

In total, estimated over-collection

amounts to:

49.6 mDKK

42.7 mDKK

Adding a conservative time

compensation of 50 mDKK, puts the

combined indirect losses at

approximately:

100 mDKK

Athens Remediation Analysis - July-2020 - Version 0.3

Project Athens

–

Question 4

Financial impact analysis

–

Tax impact

Tax impact (high level)

Direct tax effect

Wrongfully collected debt which according to Danske

Bank/Plesner was time-barred

Condictio

Indebiti

►

Post 31 December 2009: Claiming refund of

previously paid debt (interest, fees etc.), which is

time-barred, may be reclaimed by the customer

without taxation, assuming that the refund occurs

less than 10 years from when the payment was

made

►

Prior to 31 December 2009: Repayment by the

Bank to the customer of paid debt (interest, fees,

etc.), which is time-barred, is taxable for the

customer, if the refund occurs more than 10

years from when the payment was made. Such

taxation will be based on a

gift-like

consideration

Interests

►

Any interest (or time compensation) paid by

Danske Bank to debtor on redress amount should

be taxable as interest

Examples

Tax reporting

►

Applicable to all customers in DCS and PF incl.

debtors that have not paid on debt

►

Based on examples, the wrong tax reporting

appears in following cases:

•

Too large principal amount

Wrong debtor status

Interests amount

No arrears mark, etc.

When transferred to debt collection, the default date and

not the relevant due date was used as the origination date

of the debt. Danske Bank wrongfully collected the debt

Condictio Indebiti

•

Condictio Indebiti -claim is not taxable

•

Interests on redress are taxable

•

The tax reporting is flawed and should be re-assessed

Indirect tax effect

Interest and fee deduction

►

Interests are generally deductible if the interest

amount for previous years have been paid. Fees

are as main rule not deductible but special rules

may apply

►

In some cases of wrongfully collected debt

interest deducted is reversed when the Danish

Tax Authorities re-assess within three years

period (ordinary re-assessment deadline) when

there is no arrears mark Restance markering

►

In cases where arrears mark has been made the

debtor has not received deduction of interests.

Information shows arrears mark occurred in 51%

- 68% of the cases implying that a large part of

debtors have not received interest deduction and

therefore minimizing indirect tax effect

►

Other indirect tax effects (if any) are still under

investigation

RC1

Principal, interest and fees collapsed

Too high principal amount (incl. fees and interests) was

transferred into DCS. The derived effect is that the accrued

interest amount is too high. The debt may be time-barred

and hence wrongfully collected

•

Condictio Indebiti -claim is not taxable

•

Interests on redress are taxable

•

The tax reporting is flawed and should be re-assessed

RC2

Incorrect debt origination date

►

Currently being analyzed how to correct and how

far back in respect of data and tax reporting rules

RC3

Guarantors and co-debtors treated alike

•

Currently pending as data search is complex

Recommendation

•

Danske Bank to consider time period of redress amount given direct tax effects for customers prior to 2009

Dialogue with the Danish Tax Authorities regarding direct and indirect customer tax effects and reporting correction approach

Potential VAT issues should be analyzed further

Binding ruling of the customer taxation of redress amount to be considered by Danske Bank

Further work on root cause 1-4 examples to assess tax effect and additional root causes (if any)

RC4

Co-debtors charged full principal

In DCS the debtor and co-debtor both were registered as

main debtor. The debt was redeemed twice. Tax reporting

of interests are non-existing but debt reliefs were agreed in

2016

•

Condictio Indebiti -claim is not taxable

•

Interests on redress are taxable.

•

The tax reporting is flawed and should be re-assessed

Athens Remediation Analysis - July-2020 - Version 0.3

Project Athens

–

Question 4

Financial impact analysis

–

Overview of Tax consideration process

RC1

Principal, interest and fees collapsed

RC2

Incorrect debt origination date

RC3

Guarantors and co-debtors treated alike

RC4

Co-debtors charged full principal

Accrued interest and fees are collapsed into the principal amount

Incorrect collection on limited interest / fees due to difference in

limitation between principal and interest

When transferred to debt collection, the default date not the due

date is used a origination date of the debt

Collection could happen on expired interest and principal

A guarantor would then appear as fully liable for the debt, and the

bank would incorrectly treat both parties equally

Full principal amount assigned to each co-debtor

Potentially full debt would then be collected from multiple debtors

Has time-barred

debt been charged?

(New vs. old rule)

Over Payment

(Paid interest is wrong)

Has time-barred

debt been charged?

(New vs. old rule)

Has too much debt

been charged?

(Taxation of guarantors)

Errouneous tax

reporting?

Doube payment

Errouneous tax

reporting?

Yes

Start

Yes

Start

Yes

Start

Yes

Start

Yes

Start

Yes

Start

Yes

Start

Combined proces flow

Compensation

Elements:

Principal amount

(Non-taxable)

Condictio Indebiti

(Index demand)

Is the demand time-barred?

Interests

(Taxable)

Correction of taxation

(Determined by law)

Yes

Deducted taxation

(Historical taxation)

Re-assessment of taxation

(includes default interest and surcharge(pct.))

Further

compensation

to be paid

Yes

Potential redress

(to be considered)

No arrears mark

(Deductables recieved)

Wrong principal

amount reported?

Athens Remediation Analysis - July-2020 - Version 0.3

Project Athens

–

Question 5

The project will move into execution phase

–

Based from EY redress methodology

Identification

Identification of

issue (Project

Texas)

Initiation and methodology development

Findings from QA and

analysis can lead to

changes in previous steps

Data analysis and redress calculation

Learning from

outpayment pilot to

include in step 3

Execution plan (compensation & communication)

Pilot

Execution

Closure

Investigate

root cause

Define

Methodology

Quality Assurance

Define &

identify

customer

population

Gather

and

analyze

data

Build

analysis

platform

Perform

redress

calculations

Customer

redress

decision

Outpayment

process

Customer

communi-

cation

Outpayment

pilot

Outpayment

in batches

Quality Assurance

Engage with regulators

Program

closure:

Insight &

Business

improvement

Preparations before Day Zero

Day Zero

Ongoing customer communication and reporting after Day Zero

Main steps going forward

Current Focus

Calculation:

•

Agree on indirect / consequential losses

•

Estimate impact of DCA data

Communication:

•

Coherent communication strategy with

Authorities

•

Secure data for key messages in public

communication

Calculation, Communication, Compensation

Considerations

Execution Phase

Compensation:

•

Agree on date for Day Zero

(proposed April 14 2020)

•

Secure resources for execution,

DOCS, Advisors,

Implementation etc.

•

Decisions on thresholds for

calculations

•

Remediation principles must be approved before execution

•

Certainty that no new over-collection cases are being generated during execution

•

Agreement on scope setting parameters

–

triage stream execution closed

•

Legal advise from Plesner on remediation and compensation principles

•

Organisation and governance for execution in place

Athens Remediation Analysis - July-2020 - Version 0.3

Project Athens

–

Question 5

Communication planning* towards Day Zero to customers receiving a compensation

PRE-PLANNING:



28 FEB

(press and authorities)

PLANNING:

28 FEB

–

Day Zero

EXECUTION:

Day Zero

Day Zero:

June 18 2020

Orientation to authorities

FOLLOW UP AND

OUTPAYMENT: Day 1

EMPLOYEES CUSTOMERS EXTERNAL

28. Feb:

EY report

Ongoing dialogue with

authorities and orientation of

plans

Ongoing dialogue with

authorities and update on plans

Monitoring of customer

reactions, SoMe, politicians and

press

Update on Dansk Bank front

page

–

see appendix p. 41

Development of communications

plans and key messaging

Finalising communications plans

and key messages with final

data

**Letter is sent to customers

included in pilot batch

–

see

appendix p. 42

Letter to customers that have

submitted a claim based on

indirect losses incl. conclusion

on claim

Notification to selected

customers with incorrect debt

incl. how it has been rectified

Ongoing communication on

outpayment process

Development of materials for

Day Zero

Website regarding indirect

claims with submission form

–

see appendix p. 41

Headlines

Material for Danske

Direkte/advisors + DCM

Look-up tool for advisors (WOID)

Development of materials for

advisors and debt collection

management

Development of look-up tool for

advisors

*See appendix pp. 39-40 for additional best practice methodology leading up to execution of Day Zero **e-boks is delivered within 1-2 days. Physical letters is delivered within 2-5 days

Athens Remediation Analysis - July-2020 - Version 0.3

Project Athens

–

Question 5

Remediation of customers that are not eligible to receive compensation on the basis of

identified four root causes

A workstream dedicated to the remediation of customer groups (i) and (ii) will work towards identifying and rectifying customers with incorrect debt

registered in DCS and PF

Remediation of customer groups (i) and (ii):

•

Programme structure has been revised and simplified to ensure a clear

remediation focus and speed of execution

A programme plan has been outlined with expected completion Q3

2021, subject to changes. See timeline on next page

Specificall , the purpose of the stream Portfolio trimming & clean-up

is to rectify incorrect data on customers debt with expected completion

end of Q2 2021, pending further investigation

The programme is committed to ensure a suitable communication is

delivered to the impacted customers:

Portfolio

trimming &

clean-up

Reference

Group

Programme

Manager

PMO

Programme governance structure (June 2020):

SteerCo

•

Business Owner

Retail Banking DK

Non-Financial risk

Compliance

•

Regulatory Affairs

Comms & Public Affairs

Insolvency

Group Legal

GRDM Manager

CIO Area

•

Remediation

& data

analysis

Outpayment

execution

DCS & PF

Compliance

(Horizon)

Mitigate DCA

issues

Analysis of

side issues

Customer group (i)

•

Customers that have debt

incorrectly registered, but not

repaid any debt will be notified

about the correction hereof

Customer group (ii)

•

Customers subject to over-

collection, but have sufficient

outstanding debt to allow offset

will also receive adequate

communication about correction

of their case as they are

identified and recalculated

GRDM SME s

Corporate Functions (Finance, Legal, Risk, Compliance & Communications, etc.)

Internal and external resource suppliers

(IT, Integrators, Platform providers, Management Consultants, etc.)

Athens Remediation Analysis - July-2020 - Version 0.3

Project Athens

–

Question 5

Programme Athens timeline

Q2-2020

Q3-2020

Project scope & plan

Subject to changes following dialogue with regulatory

bodies related to e.g. remediation principles, Tax

treatment and further calculations etc.

Q4-2020

Q1-2021

Q2-2021

Q3-2021

Portfolio trimming & clean-up

Develop portfolio strategy

Proof-of-Concept

Portfolio trimming & clean-up execution

Specify automation

tool

Develop & test

automation tool

Remediation & data analysis

Cohort segmentation, analysis, ready for out-payment

Out-payment pilot

Customer batches prepared for out-payment

Customer batch out-payments

Outpayment execution

DCS & PF compliance (Horizon)

TBD

Court cases corrected & resubmitted to courts

Mitigate DCA issues

Assess Vendor set-up

Plan and implement new Vendor set-up

Align with GRDM SME and Remediation workstream

Analysis of issues

Analysis of side issues

Hand-over to line for execution

: Decision milestones

Athens Remediation Analysis - July-2020 - Version 0.3

Appendix

Project Athens

Appendix overview

Page

Examples of the four root causes

Evaluation of correction process

Scope of the remediation

Customer segmentation

Financial impact

Remediation plan

27-30

Question

Athens Remediation Analysis - July-2020 - Version 0.3

Project Athens

Root Cause 1: Principal, interest and fees collapsed - Example

Situation description

When Customer A is transferred into DCS, the principal amount of 97,000

DKK is not mo ed into DCS s data field –

Principal. Rather, 102,000

DKK (interests plus principal) is incorrectly transferred to principal (field 3)

As a direct result, the interests of 10% PA is now being calculated on an incorrect principal amount > 100,000 DKK resulting in an even higher

monetary amount left for Customer A to pay

Interest has a limitation of 3 years, and Principal has a limitation of 10 years. Danske Bank would ordinarily have 10 years to collect 97,000 DKK

and 3 years to collect 5,000 DKK. Now that the combined amount is residing in Principal, Danske Bank runs the risk of loosing out of the interest

amount of 5,000 DKK, for which they have no claim to after 3 years

Danske Bank tries to

contact Customer A and

sends reminders (fee*) each

month. After the 4

reminder, Customer A is

transferred into DCS

DCS Data fields

Customer A and Danske Bank

agrees to a loan of 100,000 DKK

with an interest rate of 10% PA,

with quarterly accrued interests,

starting the 1

of Jan 2018

Payment terms are that

Customer A is to pay 1,000 DKK

and the end of each month

Customer A diligently pays 1,000

DKK each month until Sep 2018,

after which Danske Bank does

not receive any payments

Legal/court

costs

10% PA

Interests

2018

Monthly

payment

Interests

Principal

amount

(primo)

Balance

(ultimo)

Jan

1,000

Feb

1,000

Mar

1,000

2,500

Apr

1,000

May

1,000

Jun

1,000

2,500

Jul

1,000

Aug

1,000

Sep

payment

2,500

Oct

payment

Nov

payment

Dec

payment

2,500

102,000

Principal

100,000

99,000

98,000

97,500

97,000

Acknowledge

Interests

99,000

98,000

99,500

98,500

97,500

99,000

98,000

97,000

99,500

102,000

Accrued

Interests

* For the purpose of this example fees are not included.

Each reminder is

DKK and ill e entually be put into DCS’s Interest data field

Athens Remediation Analysis - July-2020 - Version 0.3

Project Athens

Root Cause 2: Incorrect debt origination date - Example

Situation description

Interest has a limitation of 3 years, and Principal has a limitation of 10 years

The incorrect origination date in DCS will make the debt appear 90 days younger than it actually is

The due date is wrong and shifted in the customers favor so the outdated interests will appear to have a 3 years and 3 months expiring period

instead of the legal 3 years

Collection post expiration is therefore a large risk

Danske Bank now has Customer

A in DCS, but with the wrong

origination date along with the

wrong principal amount

consisting of interests, fees* and

correct principal amount

DCS Data fields

Danske Bank starts to send

Customer A reminders as no

payment was received in Sep

2018

After having sent the 4

reminder

to Customer A, Danske Bank

transfers Customer A to DCS

(maximum 90 days after first

failure of payment)

DCS logs Customer A and his

owed amount from the due date

of the 4

reminder and not from

the default date where Customer

A first missed the payment

29-Dec

Legal/court

costs

Limitation: 3 year

29-Dec

Interests

•

In the example, Customer A misses out on

his instalment on 30-Sep-2018

90 days passes, were 4 reminders are sent

On 29-Dec-2018, Customer A defaults on

his loan

The advisor terminates the product and

sends Customer A to debt collection

In the transfer to DCS the default date and

not the due date is used as origination date

This leads to incorrect limitation

30-Sep

Due date

4 reminders

29-Dec

Default date

Limitation: 10 year

Principal

2018

Monthly

payment

Interests

Aug

1,000

Sep

Oct

Nov

Dec

Acknowledge

Interests

No payment No payment No payment No payment

2,500

Accrued

Interests

* For the purpose of this example fees are not included.

Each reminder is

DKK and ill e entually be put into DCS’s Interest data field

Athens Remediation Analysis - July-2020 - Version 0.3

Project Athens

Root Cause 3: Guarantors and co-debtors treated alike - Example

Situation description

During the conversion into DCS in 2004 there were no differentiation between co-debtors and guarantors, as both (technically) needed an

account in the system and no flag for marking them as guarantors existed

Hence the Bank was not able to distinguish the two from each other and could potentially be pursuing the full debt from both individuals, despite

that the guarantor could by liable for a smaller amount

As a guarantee has to be called in order to be valid, the incorrect registration have led to incorrect collection

Danske Bank introduces the DCS

system July 2004, leaving them

with one consolidated system

used for debt collection

Guarantors, such as Customer A,

is treated as co-debtor and is

incorrectly created in DCS as

such

Legally the Bank cannot collect

from Customer A till Customer B

has defaulted, and the Bank will

have to call the guarantee before

collecting from Customer A

DCS Data fields

Guarantor

Customer B is liable for

200,000 DKK

Customer A is guarantor

for 50,000 DKK

Customer B signs a loan of

200,000 DKK for her business

and has Customer A as guarantor

for 50,000. This is done before

the year 2004

DCS Data fields

Legal/court

costs

Legal/court

costs

Interests

Customer B s debt is correctl

migrated into DCS system

during the conversion

Guarantee is incorrectly

migrated as debt of 50,000 into

DCS, and he now appears as

debtor

Interests

200,000 DKK

Principal

50,000 DKK

Principal

Acknowledge

Interests

Acknowledge

Interests

Accrued

Interests

Only Customer B owes Danske Bank

200,000 DKK

Accrued

Interests

Athens Remediation Analysis - July-2020 - Version 0.3

Project Athens

Root Cause 4: Co-debtors charged full principal - Example

Situation description

When data was converted into DCS in 2004, the link between co-debtors was in some instances left out

Hence the Bank was unable to correctly track the collection of the shared debt, and would treat the payments separately

There have been instances where settlements have been agreed with each part of the shared debt, leading to a combined overpayment

Potentially the collection team

will now pursue both Customer

B and Customer A for the full

amount, even despite one of

them might have paid it all

DCS Data fields

Customer B and Customer A

collectively signs a loan from

Danske Bank as co-debtors for

200,000 sometime before the

DCS Data fields

year 2004

Danske Bank introduces the DCS

system July 2004, leaving them

with one consolidated system

used for debt collection

Co-debtors

Customer B is liable for

200,000 DKK

Co-debtors, such as Customer B

and Customer A, were created in

DCS each with the full loan

amount and with not linking

between the cases

Customer A is liable for

200,000 DKK

Legal/court

costs

Legal/court

costs

Interests

When migrated to DCS,

Customer A and Customer B are

created as two separate cases

without linkage

Hence the debt collection team

will potentially collect the full

amount from both Customer A

and Customer B (double

payment)

Interests

200,000 DKK

Principal

200,000 DKK

Principal

Acknowledge

Interests

Acknowledge

Interests

Accrued

Interests

The co-debtors collectively owes Danske

Bank 200,000 DKK

Accrued

Interests

Athens Remediation Analysis - July-2020 - Version 0.3

Project Athens -

Question 1

In order to ensure that no new over-collection cases are created, it is necessary to review

GRDM s correction process including the implemented red/green case markings

•

When Danske Bank became aware of the issues in DCS, GRDM implemented a manual failsafe process in order to minimize the risk of creating new issues

In general, all Danish cases registered in DCS are considered incorrect and need correction before reaching out to the customer

Each case transferred to DCS will (since 16

of July 2019) automatically be marked red, stating that it has not been reviewed by the Corrections Team

Only cases that have been treated / corrected by the Corrections Team can be viewed as correct

–

exemplified by a green tick-marker post-correction

Current correction process* since 16

of July 2019 looks as follows:

GRDM

•

Soft collection

•

Hard collection

Customer

defaults and

enters Soft

Collection

Termination of

engagement

Customer moved

from FEBOS to

DCS

DCS case is

automatically

marked red

Activate, review,

assign case and

notify customer

If the case is Light;

Local customer team

corrects case, from

backlog

Mark corrected

case green and

mail Hard

Collection

Receive correction

mail and try to

contact customer

Normal DCS flow

and debt

collection

•

Customer

•

Customer

representatives

(lawyers, family

member, advisor)

Customer

Payment

Receives notification

and may act and

cause a trigger

Trigger

examples

Phone / mail

Settlement request

Correction

Team

•

Appointed

Corrections Team

Backlog / non-active

portfolio of red-

marked cases

If the case is

complex; Correction

corrects case from

backlog

Mark corrected

case green and

mail Hard

Collection

Continuous quality

control of green-

marked cases

Flow description

An customer in FEBOS suddenly defaults on his loan e.g. missing payment and is

moved to Soft Collection which initiates the Dunning Process

Dunning Process is unsuccessful and the engagement is terminated (4

reminder)

10 days after termination the (customer) case is moved from FEBOS to DCS

When the case is registered in DCS it is automatically marked red

Hard Collection (Lithuania) activates the latent case in DCS, reviews the case, assigns

the appropriate customer team if the case is deemed Light or to the Correction Team

Correction if the case is more comple the case is no in the backlog depositor

The notified customer (or customer representative) could activate a trigger* (e.g. make

communicative contact, make a payment or send a settlement request)

10.

11.

If a trigger is activated it prompts either the local customer team (Light) or the

Correction Team to locate, review and correct the case from the backlog of cases

Once the case is corrected it is marked green followed by a mail to Hard Collection that

the particular case has been corrected

After receiving the mail, Hard Collection seeks to contact the customer in order to

address the specific trigger

Regardless if customer contact happens or not, the green-marked case is now

following normal flow in DCS

As a quality measure, a list of green-marked cases are reviewed by GRDM each month

in order to double check that the case has been handled and corrected the right way

Athens Remediation Analysis - July-2020 - Version 0.3

* There are more triggers (externally and internally) but it is important to note that any handling of a red-case starts with a trigger

Project Athens -

Question 1

Is it ensured that over-collection does not still occur despite implementing red/green

mitigation process?

•

Current correction process* since 16

of July 2019 looks as follows:

Get an overview of the case

Correction of the case

Immediate post-

correction activities

Apply mitigating

red/green process

•

Locate a red-marked / non-

corrected case

Check basis information in

DCS

Determine expiration / statute

of limitation for each account

Assess which type of account

needs correction (promissory

notes, overdraft, overdraft

facility etc.)

•

Review bank statements

Conduct correction(s) by

utilizing appropriate

correction table

•

Create notification date for

each stakeholder s CPR

number pertaining the case

Create playtast

–

predefined

table for case comments

•

When the case has been

corrected, notification date

and playtast has been

correct, apply green tick in

the overview image

Once a month a quality control of corrected cases marked green is conducted

If one or several elements are missing (e.g.

notification date and/or playtest), the Corrections Team is alerted in order to rectify the administrative mistake

* Vejledende løsning til tilretning – Sager efter

. 9.

Athens Remediation Analysis - July-2020 - Version 0.3

Project Athens -

Question 2

Outsourcing to DCAs has not significantly impacted data quality

Background for DCA stream

A task team within the project has been working to analyze and understand the data on customers who were

outsourced to DCAs. The purpose is to determine whether or not the data quality is acceptable to proceed with the

remediation project in its current form

Active DK customers in DCS identified as

potentially outsourced:

40,500

DK customers in data received

from the four largest DCAs:

Review of the outsourcing to Debt Collection Agencies (DCAs) shows that flaws are minor

The total number of potentially outsourced and active DK customers in DCS is estimated to be

40,500.

EY has

received data from the four largest DCAs

, which contained

21,700

unique costumers. The received data was then

matched with data from DCS resulting in

20,750

matched costumers.

96%

of the matched customers had the

correct status in DCS (i.e. active / closed). EY therefore concludes that the data quality issues with the DCAs

should not postpone execution of remediation

~21,700

DCS data matched

with data from four

largest DCAs (sample

used for analysis on

case status):

How the initial analysis was carried out

For this analysis EY built a total overview of the data in DCS. From here EY:

Isolated potentially outsourced data for the DK organization

Identified and focused on data associated with the 4 largest DCAs

Matched outsourced DCS data with data received from the 4 largest DCAs

Compared the case status between the matched DCS and DCA data

Throughout this process EY has validated relevant data and aligned across all EY data teams on the DCS project.

Furthermore, EY has sparred with Danske Bank SMEs

20,750

Case status

agreement

rate:

96%

Bias in data

Initial analysis of sample data has shown a modest error rate.

However, EY needs to understand the rest of the data foundation and

although the error rate is likely to resemble the sample error rate,

further analysis is required to ensure that this is the case

Athens Remediation Analysis - July-2020 - Version 0.3

Next steps

The next steps are to:

Deep dive analysis of the exhaustive data foundation (a large sample was investigated for this analysis)

Analyze the data flow between Danske Bank (DCS and PF) and the DCAs

Project Athens

- Question 3

The 106k customers in scope can be split between DCS (80k) and PF (26k)

A repayment ratio of 80% will yield 43,150 customers for manual handling

•

26,000 customers in PF system are in scope for redress

•

A repayment ratio (percentage of principal that has been repaid) of 80%

will yield 5,149 customers to be manually reviewed

•

This is under the assumption that lower repayment ratios will ensure

that possible corrections can be offset in the remaining debt

•

Additionally, 1,000 customers have insufficient data and as a result will

have to be manually reviewed leaving a total of 6,149 customers for

manual review

120.000

100.000

80.000

60.000

40.000

26.000

20.000

Customers in scope

Customers in DCS

Customers in PF

(incl. overlap)

80.000

20.000

15.000

10.000

5.000

PF customers in scope Remove customers with

repayment ratio less

than 80%

Customers to be

manually reviewed

19.851

6.149

106.000

30.000

25.000

PF - 80% repayment ratio*

26.000

DCS

•

80,000 customers in DCS are in scope for redress

•

20,000 of these can be modelled by re-calculating the debt with data

from after 2010

•

60,000 customers can not be modelled due to lack of data

•

By reviewing the corrections made since June 2019, data shows that

corrections remain below 20% of principal. Corrections are (for 98.5%)

maximum 20% of principal

•

Assuming that corrections can be offset in remaining debt, will allow for

discarding customers with repayment ratios below 80% (28,500)

•

However, all customers entering DCS in 2004 (8,500) have to be

manually reviewed due to their potential exposure to root cause 3 and 4

•

In total, 37,000 customers in DCS have to be manually reviewed

120.000

100.000

80.000

26.000

60.000

40.000

20.000

Customers in Customers in Customers in Customers

scope

DCS

that can be

(incl. overlap)

modelled

(after 2010)

Customer

with

insufficient

data

Athens Remediation Analysis - July-2020 - Version 0.3

20.000

60.000

106.000

80% repayment ratio

43,150 customers for

manual handling

* These are all customers that are potentially eligible, the actual number after further investigation is expected to change

Project Athens -

Question 3

24k customers that have entered and left the two systems after 2004 and been there for

less than 3 years. Hence they can not have been impacted by the four root causes

Characteristics of the cohort

•

All of these customers have been in the systems less than 3 years*

•

For

DCS,

customers who has been active less than 2.5 years in DCS are removed by this

filter

•

For

PF,

customers entered in the system less than 3 years ago are removed as no debt is

at risk of being time-barred (arrears date will always be correct for PF)

•

That principal, fees and accumulated interest might have collapsed does not lead to

incorrect interest calculation as the interest is consistent across debt types

•

Hence root causes 1 and 2 have not impacted the 24k customers:

‒

RC1

–

Collapsing of fees, interest and principal

As interest rates are consistent across all debt types, collapsing of debt has no impact

‒

RC2

–

Incorrect arrears date

Not an issue for PF customers as independent review ensures correct input date. Not an

issue for DCS, as long as soft collection process is less than 180 days

‒

RC3 - Guarantors and co-debtors

Only applicable for DCS, but none are among the 24k as they are all after 2004. The root

cause only occurred during the migration into DCS

‒

RC4 - Co-debtors charged full principal

Only applicable for DCS, but none are among the 24k as they are all after 2004. The root

cause only occurred during the migration into DCS

Breakdown of the cohort across DCS and PF

24k

23k

DCS

< 1k

Both systems

Areas of risk are non-systemic operational errors as e.g.:

Court fees being incorrectly collapsed when entering DCS. Interest should not

be calculated on these fees. The issue is currently under investigation

Duration of soft collection exceeds 180 days (standard duration is 90 days)

* For DCS a conservative cut-off

equal to .

foreclosure process

ears is applied as a rolling indo

throughout the period. For PF an e act cut-off

equal to 3 years is applied going back to 2017, as the arrears date is correctly registered due to independent review in a

Athens Remediation Analysis - July-2020 - Version 0.3

Project Athens -

Question 4

How interest compensation will be implemented in redress calculations

Legal basis for interest compensation

•

Plesner has provided their legal assessment

of Danske Bank s set-off right and hen the bank is obliged to compensate customers and add interest to the customer s compensation

Plesner concludes that Danske Bank is legally obliged to compensate according to the Interest Act section 5, for customers with no current debt and when set-off in existing debt will not fulfil the

costumer s restitution claim

Such customers should recei e compensation

ith attributed interest corresponding to the rele ant process rate in Danish

morarenten

at the time of each o erpa ment made b the customer

Customers that have been subject to over-collection, but have sufficient existing debt to allow for set-off should not receive interest compensation

Implementation of interest compensation principles

How to apply the interest rates on over-collection?

•

From a legal perspective, Danske Bank is required to;

•

Option 1a

- Add interest to over-collections incrementally as they occur (i.e. add interest to

the individual over-collection from the time of the respective collection)

•

Which time period to apply?

•

From a legal perspective, Danske Bank is required to;

•

Option 1b

–

Add interest from the point in time of each individual payment by the customer

that was paid on an incorrect basis (i.e. time-barred or in excess of actual correct debt)

Implementing such logic is impractical and instead EY suggest to;

•

Option 2a

- Add interest on the entire over-collection amount starting from the time that

the first over-collection event took place

Implementing such logic is impractical and instead EY suggest to;

•

Option 2b

- Add interest from the earliest point in time that the customer theoretically risk

overpayment

–

i.e. the first payment by the customer after 2.5 years in debt collection

•

Option 2a is chosen

as it will (always) result in a benefit for the customer as opposed to option 1a.

•

Although, this might lead to unintended taxation as the interest compensation will fall beyond

Danske Bank s legal

obligation

Option 2b is chosen

as it will (always) result in a benefit for the customer as opposed to option 1b.

Although, this might lead to unintended taxation as the interest compensation will fall beyond

Danske Bank s

legal obligation

Tax effect of such interest compensation should be clarified

with Danish Tax Authorities (pending)

Further complexities such as interest compensation to customers with time-barred claims (i.e. older

than 10 years), will have to addressed by programme as fulfilment progress

Memorandum prepared by Plesner the 8

of April 2020

– SET-OFF RIGHT IN RELATION TO DEBT COLLECTION ON A UNJUSTIFIED BASIS

Athens Remediation Analysis - July-2020 - Version 0.3

Project Athens -

Question 4

Sample customer in pilot batch

case H

H have been subject to over-collection as

he has repaid;

H obtains an uncollateralized

retail loan (Danske Lån) at

Danske Bank

Feb

2011

(i)

Dunning #1 is

sent without a fee



fees that were limited (DKK 300)

interest on limited fees (DKK 37.6)

Dunning #2

100 DKK

fee



Dunning #3

100 DKK

fee



Dunning #4

with DC warning

100 DKK

fee

(ii)

2008



He starts paying his

monthly installments

H stops repaying his

monthly installments

Soft collection starts

Dunning proces in FEBOS

FEBOS

Hard collection starts

Registration in DCS

DCS

Apr

2011

Dunning fees are

incorrectly collapsed with

principal when migrated to

DCS

2016

H has repaid his debt in full

making him eligible for redress

GRDM contacts H and convinces him to sign

settlement agreement

Case is

closed in

DCS

Derivatively, incorrect reporting

is made to DTA

in 2016

However, fees (300 DKK) have now limited

incl. interest on them (37,6 DKK)

Day

Jun 18

2020

Zero



Unpaid

‘ordinary’ interest + fees

and penalty interest

Limitation = 3 years

Principal

Limitation = 10 years

H will receive a compensation equal to 476,2 DKK

476,2 DKK = 337,6 DKK (over-collection) +138,6

DKK (interest compensation

–

See next page)



Danish

Apr

2015



H’s total debt of

35,782 DKK

registered in DCS

–

see entry balance

Actual registration

DCS

Data field

DKK

929

34,853

Dunning fees (300 DKK) are

incorrectly registered as principal debt

Total

35,782

H will then have 2 years to

substantiate any indirect losses….

Fulfillment complete

Tax Authorities

Athens Remediation Analysis - July-2020 - Version 0.3

Project Athens -

Question 4

Interest compensation -

case H

•

H’s

debt was registered in DCS in 2011 and his first repayment was made on April 23

2015

•

Thus,

H’s

claim against Danske Bank fell after March 2013

•

For simplicity, it is assumed that the entire over-collection amount occurred on this

date to the benefit of the customer

•

Interest compensation is calculated as simple interest (i.e. not compound) from date of

over-collection until fulfilment date

•

For H, the time period is April 23 2015 to June 18 2020 (Day Zero)

Calculation of

H’s

interest compensation

Total over-collection amount x Interest compensation rate (per day) x Number of days

= Interest compensation amount

DKK 337.6 x (8.20% / 365 days) x 68 days + 337.6 x (8.05% / 365 days) x 1,815 days

= DKK 138,6

March 1 2013

Interest compensation rate

7% surcharge

Before

8% surcharge

After

Lending rate

8.20%

(April 23 2015

–

June 30 2015)

8.05%

(July 1 2015

–

June 18 2020)

Athens Remediation Analysis - July-2020 - Version 0.3

Project Athens

–

Question 5

Based on best-practice redress methodology, EY suggests to plan towards a Day Zero,

with compensation commencing immediately after the press release

BEFORE

DAY 0

TBD: Potential outreach to

selected stakeholders

PILOT & OUTPAYMENT PERIOD

AFTER

COMPLETION

Comms

Press/public affairs

Dialogue with the Danish FSA

Customer comm.

Populate database for

individual customer

look-up in WOID

Customer letter 1:

Information about the

issue identified

Customer letter 2:

Information about how much in compensation they will receive. If the

tax matter is clarified, this will be included

Advisors

Advisor tools

development

Advisor training

Ensure proper

documentation

Taxation

Settle taxation on each re-dress strategy

Build scenarios for

Straight to Pay (STP)

outpayment flow

Benchmark (BM)

and calculation

approach

StC sign-off for

BM and calc.

approach

Calc’s

Scenarios

Calculation

Approach &

methodology

Build scenarios

for manual flow

Review

learnings

Outpayment

Prepare

payment file

Design QA

and controls

Code systems

with rules for

STP annual flow

Train staff on

scenarios for

manual flow

Run test pilot with

letter, advisor lookup

and outpayment

Close program

Pilot

Execution

outpayment

Scope payment phases

(current expectation)

Execute

payment file

phase 1

Execute

payment file

phase 2

Execute

payment file

phase 3

Execute

payment file

phase 4

Calculation, Communication, Compensation

Athens Remediation Analysis - July-2020 - Version 0.3

Project Athens

–

Question 5

EY recommends to use best-practice communication approach and platform from

previous remediation projects

Communications Day Zero

Decision for SteerCo to agree on

Tel.: XX XX XX XX

•

EY suggests to use the familiarities of the

communication approach, tools, platforms

and training approach from previous

remediation projects

Headlines

On at 9.00am

Press release

Published at 9.00am

Website

www.danskebank.dk/

Athens

Letter

Sent to affected customers with

hotline campaign number

Material for advisers

WOID

The Regulations about in estments

item, Q&A, background info, copies of

letters to customers, calculation look-up

Customer and Advisor experience

Messages for customers

Available Monday morning -

Overview of affected customers,

possibility of proactive contact

Opslagsværk

Find out whether a specific

customer is affected and how.

Find it on WOID

Athens Remediation Analysis - July-2020 - Version 0.3

Project Athens

–

Question 5

Announcement on the front page of Danske Bank s site links to a site with details on the

issue; incl. Q&A and a form to submit claim based on indirect losses

Danske Bank front page,

Aktuelt

www.danskebank.dk/privat/kunde

Athens site with details and form

www.danskebank.dk/tilbagebetaling

Indirect losses form

[indirect loss form]

Application form will be available

to all site visitors, but a review

process is designed to filter away

applicants without legitimate

claims

Tilbagebetaling af tidligere inddrevet gæld

Ved en fejl har vi opkrævet nogle kunder i vores

inkassoafdeling for meget på deres gæld. Er du berørt, får

du nu penge tilbage.

Information to public will be

announced under

Aktuelt

on the

front page of Danske Bank s

homepage

Ne updates under

Aktuelt

hen

critical outpayment batches are

due, e.g. when outpayments

amounts are large

Detailed information on the issue

are outlined with further details in

Q&A setup

Link is also provided in the letter

sent to customers in scope for

outpayment

Athens Remediation Analysis - July-2020 - Version 0.3

Project Athens

–

Question 5

The redress process is designed to ensure customers are compensated as soon as possible

In addition to swift payout the process is designed to ensure proper consideration of customers indirect losses

Danske Bank

customer

Compensation for direct

losses paid into

NemKonto immediately

Hotline number provided

Customer have

further questions

Customer

letter sent to

inform about

issue and

compensation

Including time compensation

if applicable

14 days

If customers have no indirect

loss claims and no further

questions the redress is

considered fulfilled

Running 3 years after

last letter is sent

Webpage provided in

letter

Compensation for

direct losses paid

into chosen account

Customer wish to

claim in-direct

losses

Non-Danske

Bank customer

Customer request

payout to non-

Nemkonto account

Customer reports

claim in webpage form

Debt collection

experts

Senior

management

Indirect claim is

recognised by

board

Customer informed

about conclusion

Indirect claim is not

recognised by

board

Customer wish to dispute

conclusion

Compensation for

indirect losses paid

into chosen account

If customer have no disputes the

customer is considered redressed

in full

Legal

Indirect loss compensation board

assess the claim for indirect losses

Opportunity to dispute conclusions

on indirect loss claims will be

provided in information letter and

managed on ad hoc basis

Athens Remediation Analysis - July-2020 - Version 0.3

BILAG 2.5

–

YDERLIGERE PROBLEMSTILLINGER

Danske Bank er på nuværende tidspunkt ved at undersøge en række problemstillinger, der kan have

været af betydning for vores kunder, og som er blevet opdaget i forbindelse med vores interne

undersøgelse af inkassosagen.

Undersøgelserne af disse problemstillinger er i mange tilfælde på et tidligt stadie, men vi synes, at det

er vigtigt at give en opsummering af dem. Konsekvensen af den enkelte problemstilling kan være

forskellig for vores kunder og for os. Oversigten er ikke udtømmende, idet den igangværende

undersøgelse kan bringe yderligere problemstillinger frem i lyset.

Afsluttede retssager om inkasso

Resumé:

Fejl i data i DCS og PF kan have påvirket en række afsluttede retssager, der involverer vores

kunder. Dette kan have betydet, at udfaldet har påvirket disse kunder negativt, og i nogle tilfælde kan

det også have haft betydning for andre tredjeparter, der har været involveret i den enkelte sag.

Rykkergebyrer

Resumé:

Når en kunde modtager et rykkerbrev, opkræves der et gebyr og renter pålægges

rykkergebyret og opkræves hos kunden. Der er i banken nedsat en arbejdsgruppe med henblik på at

identificere og dække eventuelle tab som følge af uretmæssigt tilskrevne renter på rykkergebyrer.

Modregning af kunders gæld mellem koncernselskaber

Resumé:

Risiko for mulig uberettiget modregning mellem koncernenheder. Denne praksis er nu

suspenderet og afventer resultatet af yderligere undersøgelser.

Data vedrørende kunders kreditvurdering i DCS og PF

Resumé:

DCS- og PF-systemerne tildeler automatisk en kunde en D4-rating (den lavest mulige rating),

når kunden misligholder sine afdragsbetalinger og derfor overføres til systemerne. Når en kunde har

misligholdt et lån, vil kunden være ratet D4 i 5 år efter det tidspunkt, hvor gælden er fuldt indfriet. Der

er indledt yderligere undersøgelser af, hvorvidt kunder uretmæssigt har været ratet D4 på grund af

systemiske datafejl.

Inddrivelse af gæld hos sårbare kunder

Resumé:

Danske Banks metode til gældsinddrivelse hos sårbare kunder er ved at blive undersøgt,

således at det kan fastslås, hvorvidt der tages tilstrækkeligt hensyn til disse kunders situation i

inkassoprocessen. En koncernpolitik er under udarbejdelse for at sikre en retfærdig og hensynsfuld

behandling af sårbare kunder på tværs af Danske Bank.

Renter på misligholdte lån

Resumé:

Problemstillingen vedrører situationer, hvor en kunde misligholder sin betalingsforpligtelse

på et lån, og hvor der opkræves en forkert rente i misligholdelsesperioden. Indikativt kan oplyses, at

den rente, som blev anvendt over for kunderne, var betydeligt lavere end standardrenten på

misligholdte lån.

Tableau-data på grundlag af DCS og PF

Resumé:

Den dårlige datakvalitet i systemerne blev indberettet i Tableau, som er bankens interne

data ase til opfølg i g på KPI’er. I d eret i gerne

fra Tableau var derfor unøjagtige, hvilket medførte

et mindre præcist grundlag for driftsmæssige beslutninger. Problemstillingen er løst af Customer

Service Delivery. Der er ikke udarbejdet nogen risikovurdering på grundlag af de tidligere forkerte data.

Advokatsalærer i retssager

Resumé:

Foreløbige undersøgelser har indikeret, at der var fejl i allokeringen af advokatsalærer i

specifikke sager før overgangen til DCS. Retspraksis er at allokere salærer tilkendt af retten og ikke de

faktiske salærer. Dette forhold er undersøgt yderligere, og der foretages stikprøvekontrol. Midtvejs i

forløbet er der ingen tegn på fejl.

Advokatsalærer indeholdt i hovedstolen

Resumé:

I sjældne tilfælde kan der i bedragerisager være risiko for, at advokatsalærer fejlagtigt er

medregnet i hovedstolen. Dette er imidlertid kun forkert, hvis de allokerede advokatsalærer er de

faktiske advokatsalærer og ikke de af retten tilkendte advokatsalærer. Denne mulige problemstilling

eksisterer kun, såfremt ovennævnte stikprøvekontrol mod forventning viser tegn på forkert allokering

af advokatsalærer.

Home-problemstilling

–

mæglersalærer

Resumé:

Dette er et område, hvor der kan have været en interessekonflikt vedrørende mæglersalærer

i for i delse ed ”volu tary defi it sales”. Pro le stilli ge vedrører ale e

home,

og proceduren er

rettet i juli 2019, så der tages højde for eventuelle problemstillinger. Det udpegede team har

identificeret problemstillingens karakter og konkluderet, at ejendommenes tilstand ikke blev taget i

betragtning i den oprindelige analyse. Projektet har analyseret hele porteføljen af kunder med

”volu tary defi it sales” og

udarbejdet en grundig analyse af de godkendte salærer for

salgsporteføljen.

Unøjagtig skatteindberetning

Resumé:

Det er muligt, at Danske Bank som følge af hovedårsag 1 har indberettet et forkert rentebeløb

til skattemyndighederne, hvilket i så fald har medført et forkert beløb for fradragsberettigede renter

for nogle kunder. Kunder, der har lidt et tab som følge af forkert skatteindberetning fra bankens side,

vil modtage fuld godtgørelse. Danske Bank har taget kontakt til skattemyndighederne vedrørende

dette forhold.

GDPR

Resumé:

Det er muligt, at der er visse dele af persondataforordningen, som Danske Bank ikke har

overholdt som følge af fejlbehæftede data i DCS og PF, og Danske Bank er i kontakt med Datatilsynet

vedrørende dette forhold.

Inkassobureauer

Resumé:

Danske Bank outsourcer inkassosager til en række inkassobureauer, der handler på vegne af

banken og bruger data, som leveres af banken. Banken er i gang med at vurdere og tage forbyggende

tiltag i forhold til potentielle yderligere problemstillinger.

Asset Finance

–

Nordania Finans

Resumé:

Der kan være privatkundeaftaler, hvor flere rykkergebyrer er blevet opkrævet flere gange,

end det er tilladt. På baggrund af Danske Banks undersøgelse af beregningen af renter på gebyrer har

Asset Finance fastslået, at der kan være en udfordring i forhold til opkrævning af morarenter på

rykkergebyrer i forbindelse med bankens erhvervsaftaler.

Project Data Quality

Conference call

October 2019

Introduction

Impact of system and control failures

Facts informed by Danske Bank

•

Approx. 600,000 cases (both active and historic)

in the DCS

•

80,000

–

130,000 active cases

•

Incorrect calculation of loans, interests, costs

and fees

•

Around 55 % of the active cases are estimated

to be flawed (based on actually recalculated by

the collection team today)

•

The financial size of customer claims cannot be

determined

•

Interviews with employees supports view as to

severity of the facts

Risks

•

Reputation

•

TAX

•

FSA

•

GDPR

•

Regulatory and Criminal Liability

•

Financial risk

•

Prudent Practice and Conduct

Remediation

•

Maintaining the reputation of Danske Bank,

•

Reimbursement of suffered losses, and

•

Develop a compliant debt collection framework

High impact flaws

Examples of high impact flaws (pre and post 2004)

•

Improper integration of existing loans into DCS in 2004 resulting in interest and fees being merged and added to a single field/entry in the DCS system, which

uses the statute of limitation for the principle amount

•

The time limitation for principal and interest is generally counted from the time the debt is entered into DCS rather than from the time the debt was established

•

Statute-barred interest, costs and fee payments have incorrectly been included in settlement agreements entered with customers, and also in successful claims

made against customers in courts, estates etc.

The key impact of the system flaws

•

Customers in risk of overpaying

•

Customers have overpaid

•

Customers have paid debt that was statute-barred

•

Danske Bank has collected debt that was not owed

•

Danske Bank has submitted inaccurate tax reports to the Tax Agency

•

Inability to accept payments from debtors who wants to make total repayments on their outstanding debt

➢

It is not possible to "auto-correct" the incorrect debt calculations

➢

An update to the DFSA should be provided

Tax

Tax reporting obligations

•

By law banks are required to report the interest of the customers on a

quarterly and annual basis to the Danish Tax Authorities. These reports will

affect the customers' ongoing tax returns.

–

The reporting deadline for Q3 2019 is on October 31, 2019.

•

Danske Bank will not be able to submit an accurate report for all

customers due to the systems flaws, and subsequent improper calculation

of interest etc.

•

It is a criminal offense to knowingly make inaccurate, misleading or

incomplete reports.

•

Historic tax assessments (10 years back) for the customers reporting

might have to be re-opened and re-calculated due to inaccurate reporting.

•

Pragmatic solution;

–

Engage with the Danish Tax Authorities og the Danish FSA to seek a

preliminary plan of actions.

Skatteindberetningsloven

§ 58

Den indberetningspligtige efter denne lov bortset fra § 41, der forsætligt

afgiver urigtige, vildledende eller ufuldstændige oplysninger, der resulterer i en

for lav skatteansættelse, straffes med bøde eller fængsel i indtil 1 år og 6

måneder, medmindre højere straf er forskyldt efter straffelovens § 289.

Stk. 2. Begås overtrædelsen i stk. 1 groft uagtsomt, er straffen bøde.

§ 59

Med bøde straffes den indberetningspligtige, der forsætligt eller groft

uagtsomt undlader at opfylde sine pligter efter § 23, stk. 1 eller 2, overtræder §

52, stk. 2, undlader at foretage indberetning efter §§ 1, 5, 6, 8, 11 a-19, 24-36,

37-40 eller 42, § 43, stk. 1, eller §§ 44-46 inden for fristerne i medfør af § 15,

stk. 2, § 39, § 42, stk. 3, eller § 54, stk. 1-4, eller § 4, stk. 1, i lov om et

indkomstregister, undlader at foretage genindberetning inden for den frist, som

told- og skatteforvaltningen har meddelt efter § 54, stk. 5, eller undlader at

opfylde pligten til at opbevare eller indsende materiale efter reglerne i § 56.

Stk. 2. I forskrifter, der udfærdiges i medfør af loven, kan der fastsættes straf af

bøde for den, der forsætligt eller groft uagtsomt overtræder bestemmelser i

forskrifterne.

§ 60

Der kan pålægges selskaber m.v. (juridiske personer) strafansvar efter

reglerne i straffelovens 5. kapitel.

Regulatory and criminal investigations (prior "page 12")

The key tasks for Plesner in connection with the project is;

•

To describe the main flaws that have been identified in the system and associated processes,

To analyse and describe the main regulatory legal risks resulting from the identified flaws, and

To assist Danske Bank in identifying possible ways of solving the issues resulting from the system flaws

At present we have not made final assessments of whether or not Danske Bank is likely to be found in breach of the following (and some of

the points are outside our scope of work), however, at this stage it may be too premature to rule any of them out:

•

Market abuse

•

Governance failure

System and control failures

Fit and proper / accountability of senior management

Failure to manage conflicts of interest between Danske Bank and the customers

•

Long term failure to address known regulatory non-compliance in the collection systems

Potential Criminal charges (economic crime/fraud)

Governance failures

•

The management body of a bank defines, oversees and is accountable for

the implementation of the governance arrangements to ensure effective

and prudent management of the bank.

•

Danske Bank is expected to take reasonable steps to ensure continuity and

regularity in the performance of their services, and employ appropriate

and proportionate systems, resources and procedures to achieve this.

–

To our knowledge; Senior management at Danske Bank has been

made aware of various system flaws pervasive in the debt collection

department.

Furthermore, at implementation of the DCS the system was known to

inhibit system flaws.

•

To our knowledge neither compliance or internal audit have been made

aware of, nor have they themselves identified the system flaws

•

Interviews with employees supports the above. Furthermore, it is our

understanding that senior management has been made aware of the

ongoing faulty collection of debt.

Fit and proper

•

Senior management must at all times have sufficient knowledge,

professional competence and experience to perform their job concerned.

•

Failure to address known system flaws might represent a long-term and

repeat failure to address known regulatory non-compliance in the

collection systems, which in turn might demonstrate a lack of adequate

and timely remediation.

•

The DFSA will likely initiate an investigation as to who knew what and

when with particular focus on senior management.

System and controls

•

Danske Bank is required to maintain robust systems and controls, as well

as sound administrative and accounting procedures.

•

System flaws (both pre- and post implementation of DCS in 2004) has

resulted in inaccurate calculation of loans, interests, costs and fees.

–

Interests, costs and fees incorrectly forms part of the principal

outstanding amount for the purposes of calculating new interests.

Statute-barred interest, costs and fees

–

and potential principal

amounts - have incorrectly been included in settlement agreements,

court decisions, notifications in various estates etc.

•

There is a risk that Danske Bank may be found to have failed to properly

adhere to organisational requirements under the Danish Financial Business

Act.

Potential Criminal Charges

Specific obligations by law is imposed on the management and non-compliance or violation of such might lead to civil or criminal liability

•

At present it cannot be ruled out that members of the management/and relevant decision-makers (historic and present) may become

subject to criminal liability (fines and potentially imprisonment) due to having accepted continuing collecting of debt that were known to

possibly be incorrect

Danske Bank has made severe violations of proper debt collection management

Danske Bank has reported incorrect data to the Danish Tax Agency

•

A more detailed description and assessment will require a greater insight in the individual cases, the underlying documentation,

correspondence etc.

Key non-financial risks (analysis not completed)

Key non-financial risks

Reputational damage

Comments

•

Failure of a bank to perform 'basic' services such as calculating principal amounts, interest and fees on loans correctly

Ongoing awareness of system issues poses a risk that this will be viewed as "yet another" example of bad culture in Danske Bank

Sanctions from the authorities

Stress and pressure

Backlog of 1,500 customers who wants to make repayments, but are awaiting manual re-calculation of their debt

Public disclosure risks causing huge increase in customer requests

Increased work-load and possible bad press will increase the risk of resignation by employees in the debt collection team, and in

particular in the correction team

The authorities will likely initiate an investigation into accountability, with initial focus on the management and possibly also a

subsequent investigation into individual employees

The DFSA will likely demand a timetable for resolving the issues including ongoing updates, primarily in respect of customers that

have paid too much, but also in respect of a future state solution

Employees leaving

•

Accountability of management and employees

Pressure on organisation to fix issue asap

•

Key financial risks (analysis not completed)

Key financial risks

Repayments to the customers

Depreciation of debt that is actually time barred

Tax related liabilities

Sanctions from Tax Agency/SØIK

Sanctions from DFSA/SØIK

Comments

•

The obligation to repay customers will result in a net loss

Debt registered in DCS as 'owed', but which is actually statute-barred will need to be depreciated

There is a risk that re-opening customers' annual tax assessments will result in "late payments" of the relevant taxes. Such a cost will

likely be borne by Danske Bank

Risk of being fined for having reported incorrect data to the Tax Agency (Skattestyrelsen)

Breach of conduct of business rules (god skik) for collecting debt that was statute-barred, or otherwise not owed by the customer

Governance failures for not resolving the issues despite escalation efforts made by employees and the continuation of debt collection

whilst knowing that the calculations might not be correct

Governance failures for not having sufficient systems in place

Customers and other third parties may claim against Danske Bank for damages for the consequences of incorrect debt collection

The DCS contains incorrect data on customers' debt and other personal data in breach of GDPR.

Due to the number of customers impacted by the issue, the costs in fixing the issue will likely be sizeable. Work on re-calculating debt

correctly will take a lot of effort and resource. The need to resolve the impact on customers' tax returns may be costly. Finding a new

solution for debt collection will naturally also be costly (we understand this aspect is already ongoing)

Damages for losses suffered by customers and

other third parties

Fine for GDPR-non compliance

Work to fix issues may be costly

•

Q&A

RLESIMER

Project overview

Governance structure

To respond to the new requests from DFSA the

programme is now organized to handle both

remediation and BaU activities

Executive

Steerco

ELT

SteerCo

Authorities

Communications

Programme

Manager

PMO

Group Recovery &

Debt Management

Insolvency

Remediation &

data analysis

Mitigate side

issues

Banking DK

GRDM SME’s

Corporate Functions (Finance, Legal, Risk, Compliance & Communications, etc.)

Internal and external resource suppliers

(IT, Integrators, Platform providers, Management Consultants, etc.)