ECLI:DK:LYN:2024:BS0000000310 Sag om, hvorvidt sagsøger Aps har krav på erstatning fra sagsøgte revisionspartnerselskab efter køb af dansk forsikrings

RETTEN I LYNGBY DOM afsagt den 25. marts 2024 Sag BS-40901/2018-LYN Selskabet af 27. januar 2022 ApS (advokat Søren Lundsgaard og advokat Daniel Myhre Engell ) mod PwC Statsautoriseret Revisionspartnerselskab (advokat Jacob Skude Rasmussen) Denne afgørelse er truffet af Dommer 1, Dommer 2 og kst. Dommer 3, førstnævnte som rettens formand. Sagens baggrund og parternes påstande Retten har modtaget sagen den 25. oktober 2018. Sagen vedrører navnlig spørgsmålet om, hvorvidt PwC Statsautoriseret Revisionspartnerselskab har handlet ansvarspådragende i forbindelse med revisionen af årsrapporten for 2016 for Qudos Insurance A/S. Selskabet af 27. januar 2022 ApS har nedlagt påstand om, at PwC Statsautoriseret Revisionspartnerselskab til Selskabet af 27. januar 2022 ApS skal betale 118.483.705 kr. med tillæg af procesrente fra den 25. oktober 2018 til betaling sker. PwC Statsautoriseret Revisionspartnerselskab har nedlagt påstand om frifin-delse. 2 Påstandsbeløbet svarer til købesummen for QIC Holdings ApS – ejer af Qudos Insurance A/S – som blev fastsat på baggrund af årsrapporten for Qudos Insurance A/S for 2016. Selskabet af 27. januar 2022 ApS har, såfremt retten finder, at der skal ske hel eller delvis reduktion af kravet svarende til købesummen, tillige fremsat krav om dækning af to kapitalindskud i Qudos Insurance A/S på henholdsvis 22,5 mio. kr. og 44,67 mio. kr., dog ikke ud over påstandsbeløbet. Oplysningerne i sagen I februar 2016 overvejede Echelon Financial Holding Inc. – et i Canada børsnoteret selskab – at sælge det i 2011 stiftede forsikringsselskab Qudos Insurance A/S, som Echelon Financial Holding Inc. ejede gennem selskabet QiC Holding ApS. Den 22. februar 2016 afgav Milliman LLP, et internationalt aktuar- og konsulentfirma, en rapport: ”Qudos Insurance A/S Review of Outstanding Claim Liabilities and the Reserving Process as at 30 September 2015” . Det fremgår heraf bl.a.: ” Milliman LLP (“Milliman”) has been asked by Qudos Insurance A/S (“Qudos”) to: … Provide benchmark information on both the UK and Irish motor insurance markets … Provide benchmark claim development factors relevant for the UK and for the Irish motor insurance markets … Provide benchmark claim frequency and severity amounts for both the UK and Irish motor insurance markets, by the following subclasses: o Private passenger vehicles o Motorcycles o Taxis o Light commercial delivery vehicles o Motorhomes” Der viste sig en gennemsnitlig difference på 2,4 % i Millimans og Qudos’ estimater for ”the unpaid Claim Amount as at 30 september 2015” . Milliman anførte: ”Overall Qudos’s estimates are below our central estimates. However, as a general rule, for a mature book of reasonable size we would consider estimates of unpaid claim amounts +/-5% of our central estimate to be within a reasonable range of central esti-mates.” 3 Af Qudos’ årsrapport for 2015 fremgår bl.a., at årsrapportet var revideret af PwC Statsautoriseret Revisionspartnerselskab (herefter referet til som ”PwC”). Af referat fra bestyrelsesmøde i Qudos den 2. maj 2016 fremgår bl.a.: ”… Draft Q1 2016 financial Statements were circulated in advance of the meeting. There was a significant underwriting loss of 30.5 million DKK driven losses in UK/ Irish motor business. Person 1 advised that two largest underwriting losses were in Patrona (15.6 million DKK) and J&M (2.8 million DKK) programs. The Patrona losses were driven by changes in changes in case reserve levels and prior period adjustment (5.3 million DKK). In J&M, the losses are tied to the conversion of the Alpha taxi book which is showing increases in both severity and frequency. Gross premiums were less than budget but net premium was 52 million DKK above budget. Com-bined ratio was 108%. … Motor lines had an underwriting loss of 34.8 million DKK with a combined ratio of 112.1. Warranty lines reported an underwriting profit of 5.9 million DKK. Sickness and Accident lines had an underwriting profit of 3.9 million DKK.” Deloitte UK, der formidlede salget af Qudos, bød den 20. maj 2016 Aros Capital Partners velkommen som potentiel køber af Qudos og bad om et indikativt bud. Det fremgår af salgsmaterialet bl.a.: ”Qudos has already established the governance and control framework required to meet European regulatory requirements for Solvency II” Af Executive Summery fremgår således: “… • Qudos was established with the objective of Solvency II compliance in mind • This objective has been met with compliance across all 3 pillars of Solvency II • The significant growth of the business and the consequent operational risk component of the Solvency II requirement have placed pressure on the capital available within the business • As a consequence, actions continue to be taken to increase reinsurance protection and re-balance the underwriting • These actions are less significant than those facing many competitors with the consequence that market pricing is improving materially” 4 Aros afgav den 1. juni 2016 følgende indikative bud på Qudos: “Value Our non-binding offer to existing shareholders for a purchase of all or part of their shares is for a cash consideration of £20.4m being 0.7 x of April 2016 shareholder’s funds/ Net Asset Value and subject to full due diligence and regulatory approvals. Should Echelon wish to retain a sizeable minority stake of circa 20% for a period of 2 years our valuation would potentially be higher (circa 0.8x net asset value) on a proportional basis.” Det fremgår herudover bl.a., at “Full due diligence will be required, including but not limited to the following: … 4. Actuarial Reports …” Deloitte meddelte den 13. juni 2016, at Aros var inviteret til anden runde, hvor Aros ville få adgang til et virtuelt datarum. Den 10. juni 2016 afgav Milliman en rapport “on the Review of Outstanding Claim Liabilities as at 31 December 2015” . Rapporten lå i det virtuelle datarum, Aros fik adgang til. Af rapporten fremgik en difference på 5,9 % i Millimans og Qudos’ estimater for ”Total Outstanding Claims Liabilities” . Milliman anførte: ”Table 4.1 above indicates a potential deficit of 39.4 million DKK in the booked outstanding claim liabilities as at 31 December 2015 compared to our estimates on a gross of reinsurance basis. …. 8.1 In reserve estimation there are three primary components of uncertainty: 8.1.1 Model uncertainty:…. 8.1.2 Parameter uncertainty:…. 8.1.3 Process uncertainty… QUANTIFICATION OF UNCERTAINTY 8.9 We have undertaken analyses which provide some quantitative insight into the degree of uncertainty within the best estimates we have made. We must point out that actuarial and statistical science is not yet at a stage which allows full quantification of the uncer-tainty and so the following summary of our results should be taken as purely illustrative. 8.10 Of the three sources of uncertainty, explained above in paragraph 8.1, model uncertainty can be partially illustrated by setting out the estimates, and the differences between those estimates, us-ing a range of approaches and underlying assumptions. In 5 Appendix C of this Report we have set out the results of the differ-ent projection methods for each line of business. Had we consist-ently chosen, for each accident quarter and for each line of busi-ness, the largest projected value then our estimate of the unpaid claim amount would have been 94 million DKK larger than that se-lected; similarly, had we consistently chosen, for each accident quarter and for each line of business, the smallest projected value then our estimate of the unpaid claim amount would have been 294 million DKK less than that selected. Not all of these projected val-ues would have been credible but this illustrates the model uncer-tainty in our estimates. ….“ Den 28. juni 2016 afgav Milliman en rapport “on the Review of Outstanding Claim Liabilities as at 31 March 2016” . Rapporten lå i det virtuelle datarum, Aros fik adgang til. Af rapporten fremgår bl.a.: “Table 2.1 above indicates a potential deficit of 0.9 million DKK between Milliman and Qudos in the estimated outstanding claim liabilities, gross of reinsurance, as at 31 March 2016. This is much reduced from the difference that we noted as at 31 December 2015. However, we note that we have not been provided with claims data in respect of the Frida account and therefore have not been able to assess the outstanding liabilities for Frida as at 31 March 2016.” Den 1. juli 2016 afgav Aros et betinget købstilbud på ”Qudos Insurance A/S and its affiliates and holding company” med bl.a. følgende indhold: “

1. a)Valuation and terms Our offer in cash for 100% of the company is 0.65x Net Asset Value, as at 30 June 2016 which corresponds to a valuation of £20m based on 31 May 2016 balance sheet, subject to later movement in value. Assuming the Net Asset Value is equal to £30m, as stated in the Deloitte Project Odin Process letter dated 13 June 2016, our bid would equal £19.5m. This represents a slight revision in the valuation since our first indicative offer to reflect current uncertainties surrounding co-insurance availability and current market volatility. We are however prepared to agree an additional payment based on the performance of existing business over the coming three underwriting years (see
2. b)below). Our offer is contingent on additional actions concerning the current prudency of reserves. 6 Given the 8% deterioration in reserves during the first quarter of 2016 (as noted in the Milliman review of claims liabilities at 31 March 2016), our offer is conditional on obtaining a guarantee from the vendor / current shareholders of the Company to cover any fur-ther adverse development in the reserves of existing business. Should such a guarantee not be available, we would require, at the Company’s expense prior to closing any transaction, the placement of an acceptable Adverse Development Cover (ADC) reinsurance on prior– year’s claims reserves. In the event that neither a parental guarantee nor ADC be offered, a further adjustment to the transac-tion consideration would be required.” Det fremgår videre af tilbuddet bl.a.: ”Suggested Process: We understand the sensitivity and confidentiality relating to our Proposal. To minimise any disruption to the business and provide the Odin Board with the maximum execution certainty, we ask Odin’s Board to initially grant Aros a four-week period of exclusiv-ity to negotiate a final and comprehensive Share Purchase Agree-ment on market terms….” Willis Tower Watson foretog en beregning af Qudos claimes liabilities pr. 30. juni 2016. Det tal, Willis Tower Watson nåede frem til, indebar at der skulle ske en styrkelse af hensættelserne med 204.487 tDKK. Den 22. juli 2016 afgav Aros et bud ”Subject to Contract, Financing & Confirmatory Due Diligence” : “Valuation and terms We wish to submit our offer of 0.57x Net Asset Value correspond-ing to DKK 142,500,000 based on an estimated Net Asset Value of DKK 250,000,000 as of 30 June 2016 provided by Deloitte in e-mail dated 22 July 2016. The final Net Asset Value being used as the ba-sis of the transaction value shall be validated on the transaction date pending proper reserve evaluation as of that date. As an amendment to our previous offer letter we have decided to withdraw our requirement for a parent guarantee on reserves. Instead we shall in corporation with Qudos management, look to put in place an adequate reinsurance solution. The costs associated with such solution shall be borne by us in addition to our consider-ation, with the sale price adjusted at the closing of the transaction 7 to reflect the cost of reinsurance and to provide for the basis of the Price expressed as a multiple of Net Asset Value.” Vidne 1, Partner og CIO for Aros, e-mailede den 31. juli 2016 til Person 2, Echelon Financial Holding Inc., og anførte bl.a.: “As part of the signing of the SPA we need to have complete access to all books, lines of business, reinsurance discussions and employ-ees. Also, we need all the data from the dataroom in a format where we can work and model it. We will naturally keep it confi-dential …” Person 2 svarede 1. august 2016 bl.a.: “Vidne 1 I think it's ok but we need to make sure we don't distract management too much. Can we limit it to a certain time period? Can we leave it to Deloitte UK and DK to codify?” Samme dag skrev Person 3 fra Deloitte UK til Vidne 1: ”… In principle understood that you will need access to certain information and people but I am mindful that (
3. a)we need to keep a complete record of all information provided, management meet-ings requested etc for disclosure and regulatory purposes and (
4. b)that as Person 2 says this doesn’t distract the business. Vidne 2, Deloitte DK] - Perhaps you and I can discuss what is required and make arrangements? …” Den 4. august 2016 blev indgået aftale, SPA, mellem Echelon Financial Holding Inc. og New Nordic Advisors Ltd. om salg af aktierne i Qudos. Det fremgår af aftalen bl.a.: “Amount The Purchase Price is 57 per cent of the Estimated Net Asset Value (which as of 30 June 2016 was DKK 142,500,000) subject to adjustment in accordance with clause 3.3” Om begrænsning af sælgers ansvar fremgår: “9. Limitations on Seller's Liability 9.1 Clauses 9.2 and 9.3 and the provisions of paragraph 3.1 of Schedule 9 apply to Impose limitations on the Seller's liability un-der this Agreement. 9.2 The Seller will not be liable in respect of a claim made by the Buyer in respect of paragraph 4 (Compliance with Laws) of Schedule 6 ("Compliance With Laws Warranty Claim") unless the Buyer has 8 (
5. a)served a written notice on the Seller on or within three months of the Completion Date which sets out those details of the Compliance With Laws Warranty Claim that the Buyer has, including the Buyer’s then best estimate of the amount of the claim; and (
6. b)properly issued and served proceedings in respect of the Compliance With Laws Warranty Claim within six months of serving the written notice described in clause 9.2(
7. a)on the Seller. 9.3 The total amount of the Seller's liability in respect of all Compliance With Laws Warranty Claims (Including elements relating to interest and costs) is limited to DKK 2,500,000.” Det fremgår videre om køberens retsmidler: ”22.3 Buyer's only remedy is damages for breach of contract under this Agreement (
8. a)The Buyer has no claim or remedy in respect of a warranty, statement, misrepresentation (whether negligent or innocent) or undertaking made to it by or on behalf of the Seller in connection with or relating to a Transaction which is not expressly included in a Transaction Document. (
9. b)The only claim or remedy that is available to the Buyer in re-spect of a warranty, statement, misrepresentation (whether negli-gent or innocent) or undertaking made to it by or on behalf of the Seller is a claim for damages for breach of this Agreement and not termination, rescission, claim of proportionate reduction of Pur-chase Price (in Danish: forholdsmæssigt afslag), damages in tort, damages under statute or any other remedy. 22.4 Clause does not apply in the event of fraud Nothing in the above clause 22.3 or elsewhere in this Agreement limits or excludes liability arising as a result of fraud or wilful misrepresentation.” Qudos ved Vidne 3, chefaktuar i Qudos, afgav en Reserve report as at 31 may 2016 med update august 2016. Det fremgår heraf bl.a.: “The figures presented in this report are on a best estimate basis. This means that there is no intentional bias for prudence or opti-mism, and there should be a roughly equal chance of the actual outcome being higher or lower. However there is significant poten-tial for volatility, and in section 5 I have highlighted the key sources of uncertainty that could cause the outcome to differ from this best estimate. 9 … At 165.0m DKK, the total booked gross IBNR for classes in the scope of my review is 19.1m lower than my best estimate … 6.3 Tail selection / case reserve adequacy We have no historical evidence on which to base our assumptions on the degree of tail development that we are likely to experience. Different companies have a wide range of tail experience – positive and negative – depending mainly on the strength of their case reserves. In this work I have generally assumed neutral tail development – i.e. that case reserves are, in the aggregate, adequate to cover the final claim costs, by around three years after the incident date of the claim. This is the single most significant assumption that we have made, and it is made in the unavoidable absence of sup-porting evidence. Sensitivity: If the actual tail factor required were 10 percentage points higher than the selected factor, in aggregate across the mod-elled portfolios, then the final claims outcome would be around 150m DKK higher than the selected ultimate cost. … Development before the tail: in general When modelling the development of claims within the first three years after the incident date, we have generally derived a development pattern from the historical data for the portfolio in question. However we have sometimes made adjustments to the derived average pattern, for one of two main reasons: - To exclude what appears to be random variation from the pattern we would expect - To adjust for a systematic change in pattern due to, for example, a change in the strength of the case reserving basis. … … Patrona …Development ratios represent the change in the reported in-curred cost of claims from a particular accident quarter, as time passes and they get closer to settlement. There are a number of different possible interpretations of this observation. However, the interpretation that I have chosen, sup-ported by other evidence (see below) is that during the period April to September 2015 there was a strengthening of case reserves for bodily injury claims, bringing them to a level where we should ex-pect less unfavourable development thereafter. 10 The other evidence that supports this interpretation is: - Email and anecdotal evidence about a change in case reserving philosophy in 2015, following the appointment of a new claims director in Patrona - Information gathered through face to face discussions with Patrona’s claims director about the basis for case reserves, and how it varies from the past - Generally lower values in the three diagonals following the adjustment period … Sensitivity: if instead I had used the average development pattern from the past, the IBNR would have been around €4m higher. …” Person 4, CEO i Echelon, skrev den 22. september 2016 til Person 5, bestyrelsesformand i Echelon, om emnet ”Qudos…capital injection” : “Hi Person 5, As you know the financial situation at Qudos is deteriorating rapidly primarily to due to case reserve deterioration on the Patrona account. Qudos management are now asking us for a capi-tal injection of 20M DKK but we are now seeking approval for a 25m DKK injection I order to bring the SCR in the 115% range. With the expected reserve strengthening the capital ratio (SCR) is expected to at 100% or just below and Qudos management and New Nordic believe that the FSA will be hesitant to approve the transaction with the SCR around 100% and may prolong the approval process. Person 2, spoke to New Nordic and they would also like the capital ratio to be above 100% in order to expedite the approval process and investor perception. We, however do not want to inject capital into Qudos to receive 57% back. New Nordic has therefore agreed that if we inject 25 M DKK, they will exclude this amount from the book value and will repay us 100% of the injec-tion, …” Vidne 4, CFO i Qudos, skrev den 29. september 2016 til Person 6, Deloitte UK, med materiale efter en forespørgsel fra Person 6 da-gen før: ”Please find below the list of request from Deloitte and the status: From Vidne 3: - Solvency II line of business per agent/binder (submitted to Deloitte 28/9-2016) - Details of model assumptions (submitted to Deloitte 28/9-2016) - Detailed modelling and results (submitted to Deloitte 28/9-2016) - Run-off-profiles for claims (submitted to Deloitte 28/9-2016) 11 - Documentation of the different types of claims modelling (submitted to Deloitte 28/9-2016) - Details of the quota share reinsurance on the different LoBs (submitted to Deloitte 28/9-2016) - Details of which LoB have coinsurance (submitted to Deloitte 28/92016) From Vidne 4: - Premium patterns (submitted to Deloitte 28/9-2016) - Latest SCR calculations (Expected to be submitted to Deloitte 29/92016, around 12 noon) - SCR of assets (submitted to Deloitte 29/9-2016, morning) - Model for calculating the SCR risks for receivables (Expected to be submitted to Deloitte 30/9-2016) - Quota share counterparties on the reinsurance (submitted to Deloitte 28/9-2016) Additional information: - Description and calculation of tax assets (agreed with Deloitte to be submitted after the close of the Quarter)” Den 30. september 2016 skrev Person 1, CEO i Qudos, til Vidne 5, CUO i Qudos, bl.a.: “Given the recent development on this account Vidne 3 is coming up with a very large volatility in the selection of the IBNR (i.e. the range of the IBNR is between 12 to 60 million). Although this seems very large, I briefly discuss with Vidne 3 yesterday and logical argu-ments could be made for selecting either of those two numbers. Therefore, we need a good operational description of what is creat-ing the current run-off that we have seen so far this year, otherwise I am afraid that Vidne 6 at Pwc Denmark and any of the actuaries at Deloitte Denmark could also be expected to react the same way. Consequently, I would prefer that when we deliver the 30/.09/2016 we also provide a succinct description of our selection of the IBNR and the rationale for it with a good description of the underlying operational issues taking place in each of those accounts. Vidne 3 is working today and Monday to complete his assessment of the IBNR and estimated ultimate losses for each program but his first cut would likely indicate a net IBNR increase of 40 million DKK which he is well aware that such number would have very significant implications. However, we have to respect that the four of us ultimately need to agree on a realistic assessment of the cur-rent and past business and right now the emerging experience indi-cates that it is much worse that we estimated before and 12 unfortunately we don’t have a good grip on the changes taking place in the reserve philosophy taking place at our agents. … As I mentioned to Vidne 3, we might not be able to eat the elephant all in one sitting but we probably have to take a very good bite. …” Den 6. oktober 2016 skrev Person 1 til bl.a. Person 4, Person 2 og Vidne 4 bl.a.: ”… All in all, Vidne 3's best estimate currently increases the IBNR by an additional 14 million DKK net pretax or 11.0 million after tax … …I asked Vidne 4 to prepare draft financial statements without any adjustment to the current IBNR that we keep at the end of August …” Af referatet af møde i Audit and Risk Committee for Qudos den 12. oktober 2016 fremgår bl.a.: “3. Reserve Report a. Vidne 3 presented a report on reserves as of August 31, 2016 (Appen-dix 4). b. Vidne 3 described the work that had gone into the reserve review in the quarter, and he specifically directed the Committee to the Reserving Committee minutes of October 5, 2016. c. Vidne 3 reported that he was estimating a 26.3 million increase in ulti-mate claims cost for claims up to May 31, 2016. This increase was driven by a large reserve increase in one J&M claim and a signifi-cant adverse development in Patrona open claims from 2014 and early 2015. To some extent the Patrona is victim of Irish inflationary factors, the costs of having been increasing significantly the longer the claims are left open. On the other hand Staveley Head had bet-ter than expected development. The reasons for the over estimate on Staveley Head were discussed; Vidne 3 attributed largely to more and better data. d. Person 1 described the claims environment in Ireland. There was a consideration of how to approach Patrona about claims handling. Management will approach Patrona about the need to better esti-mate the claims costs taking into account the anticipated inflation-ary impacts of not settling claims sooner. Also management will conduct a claims audit within the next 30 days. e. Qudos has booked 476.6 million for claims Iiabilities (net of reinsurance) which is 11.4 million less than Vidne 3 best estimate. Given 13 the various uncertainties described, both Vidne 3 and Person 1 are comforta-ble with an increase of IBNR of 13.2 million. …” Den 12. oktober 2016 skrev Person 1 til Vidne 1 og Person 7, Aros Capital Partners LLP, bl.a.: “… In addition, you should note that the ARC supports my recommendation to terminate our binding authority with J&M effective November 1, 2016 … … In the spirit of openness and cooperation in completing your purchase of Qudos I trust that you concur with this decision. …” Af referat fra bestyrelsesmøde i Qudos den 14. oktober 2016 fremgår bl.a.: ”… Q3 Business Update 3.1 The CEO report (Appendix A) was reviewed. Person 1 confirmed that there is approximately 11 million DKK difference between what is held in reserves and what Vidne 3 is recommending. The 11 million has not being recognized in full at this time as there is on-going further review of case reserves. …” Person 1 skrev den 25. oktober 2016 til Person 8, Patrona: “Person 8: If Vidne 3 keeps puting more IBNR over your case reserves I am gonna be the one lamping him at night or broad day light” . Barrian Claims Limited afgav den 26-28. oktober 2016 et ”Patrona Audit” hvoraf bl.a. fremgår: “Patrona had recently been audited by the Central Bank of Ireland and given a clean bill of health by the Regulator. Audit therefore concentrated on file by file review looking at large losses and random sample of files selected by Qudos. Please see Patrona File Audit – Appendix 1. This file comprises 2 tabs with our comments added. All files reviewed evidenced proper investigation with settlements made in line with liabilities and all payments validated. Review evidenced Patrona’s commitment to minimising the impact of claims to Qudos. 14 Please also see large loss reports produced by Patrona dated 21 June 2016 – Appendix 2 and dated 23 September 2016 – Appendix 3. In our view, this report should be provided quarterly to Qudos. Large losses reserved in line with liabilities and reserves adjusted immediately on receipt of new information. As per our Audit Report May 2015, Patrona run a very professional and commercially minded claims operation. Claims reserves continue to be prudent and care should be exer-cised by Qudos when allocating IBNR as there may well be a situa-tion of double counting.” Det fremgår af referat fra ekstraordinært bestyrelsesmøde i Qudos den 27. okto-ber 2016 bl.a.: ”If the SCR falls down to around 100% FSA will want to see the plan to bring the SCR to 125%. FSA indicated that a managed sale to New Nordic will suffice as a satisfactory plan.” Af Qudos Insurance A/S’ “Reserve report as at 31 August 2016” , der blev gennemgået på det ekstraordinære bestyrelsesmøde og er udarbejdet af Vidne 3, fremgår bl.a.: ”… … Change since previous valuation In the previous valuation, we only assessed ultimate costs of claims gross of reinsurance, and so it is only possible to compare against last time on a gross basis. The following table breaks down the movements in best estimate ultimate cost of claims by incident year, or in the case of 2016 by incident quarter. Note that the move-ment for 2016Q2 compares the estimate as at 31 May 2016 against the updated estimate at 31 August for the same partial quarter, 1 April to 31 May. … The total movement in ultimate cost compared to the previous valuation is 25.5m DKK, gross of reinsurance. The largest individual contributors to this are: - J&M: movement in the reserve for a large loss from 2014. … - Staveley Head: significant reduction in ultimate cost across all years, … - Patrona: ultimate cost has increased by €3.4m. … … Tail selection / case reserve adequacy 15 … This is the single most significant assumption that we have made, and it is made in the unavoidable absence of supporting evi-dence. … … Patrona… … My hypothesis is that the extent of increased reserve strength can be gauged from the extent of increase in reserves applied to claims on day 1. Between claims occurring in early 2015 and those occur-ring in early 2016, I have observed around a 33% increase in initial reserves, from around €15k to around €20k. As the ultimate aver-age cost is around 40-45k, then we might expect around 20% less future development over the life of the claim (say +20k average in-crease vs +25k before). … Sensitivity: if instead I had used the average development pattern from the past, the IBNR would have been around €4m higher.” Vidne 3 skrev den 1. november 2016 en e-mail til sig selv med teksten ”po-tential priorities as a group” : ” Bring out your dead – IBNR gaps, UPR gaps, SCR gaps, etc. Confront where we are, even if only between us Get to the bottom of the hole …” Den 21. november 2016 sendte Deloitte DK til New Nordic Capital en ”Model support document for the Qudos model ver. 1.9.6” Det fremgår heraf bl.a.: “This document gives a brief introduction to the Qudos model. It gives a description of the structure of the model, and how data flows through the model. Furthermore, it also touches upon the key assumptions and important modelling points. … SCR non-life The SCR non-life module consists of the premium and reserve risk and the non-life cat risk. The premium and reserve risk is modelled in detail, while the non-life cat risk is modelled at a higher level. The calculation of the premium and reserve risk is based on the expected premiums and claims provisions for each line of business calculated within the model. We use the forward-looking formula for the premium volumes, which uses the following 12 months pre-mium for each of the modelled lines of business.” 16 Vidne 7, PwC, skrev 24. november 2016 til Vidne 4 om Qudos bl.a.: ”Hej Vidne 4, Jeg vedhæfter som aftalt udkast til materiale til din kommentering til genemgang på mødet i morgen, som følger: - Audit Plan - Aftalebrev - Oversigt over ikke fakturerede arbejder på nuværende tidspunkt Jeg hører gerne dine kommentarer. I relation til den drøftelse, vi havde i går omkring anvendelse af je-res budgettal for præmier fremfor de historiske, kan jeg oplyse, at der ikke os bekendt er eksplicitte krav til, at dette skal godkendes af Finanstilsynet. I skal dog naturligvis kunne dokumetere det. Vi vil dog anbefale, at I orienterer Finanstilsynet om ændringen med en kort beskrivelse af grundlaget for ændringen i form af de væsentligste binders og andre aftaler, der repræsenterer reduktio-nen i præmier. Anbefalingen skal ses i lyset af at undgå, at der efterfølgende fremkommer korrektioner på Finanstilsynets foranledning, særligt i lyset af, at Finanstilsynet i særlige tilfælde har ret til at give jer et tillæg til SCR.” Person 1 skrev den 24. november 2016 til Person 9, New Nordic advicors.com: ”Person 9: See attached the forecast we developed based on the discussion we had yesterday morning in your office where we assumed that at 31/12/2016 Qudos would close the year with a Capital Solvency Ratio of 95%. Assuming that with the change of control an additional 35 million DKK will be injected that will bring the closing ratio to roughly 110% which will be less than the targeted level by the DFSA. However as most of the larger motor portfolios that we terminated in 2016 are now entering their run-off phase (i.e. we no longer underwrite new or renewal business) then the volume will start dropping more rapidly which would start bringing a better balance between our volume of gross written premiums and the capital re-quired to support that business. …” 17 Den 30. november 2016 skrev Vidne 5 til Person 1 bl.a.: ”… Secondly I do get a bit scared when Person 10 starts talking about strengthening of reserves, both in general and on some specific cases (think you know about the AF ones). He could become the second Vidne 3, and with a Person 11 in there as well, we might just close down …” Person 2 skrev den 1. december 2016 til Person 1, Vidne 4, Person 4, Person 12, general council Echelon, og Person 11, be-styrelsesnæstformand i Qudos, bl.a.: “We had a couple of constructive days with Vidne 1 and Vidne 2, Deloitte DK] and also with the DKFSA[Finan-stilsynet]. We managed to agree a mechanism for payment from NN to echelon and will document that with the lawyers over the next few days. As you know we will defer the proceeds in the form of a 6 month bridge facility. Our meeting with the Dkfsa was also interesting- they seemed supportive of the transaction but couldn't promise approval in the next 2/3 weeks. We will likely get approval in mid January. One thing they did seem concerned about was the solvency ratio. I think with the derisking of the bond portfolio we should be above 100% for November. However if we dip below 100% at the end of the year, I think it's important we think about if there will be any unintended consequences- for example issues with PwC or with the reinsurers or MGAs. Can I ask you to think about this and we can discuss at the next board meeting, so we are fully prepared.” Person 1 svarede samme dag til Vidne 5: “I heard from Vidne 1 that Person 4 just went in the meeting with the DFSA and threaten the regulator that he will put the com-pany in runoff if he did not get their approval for change of control by year-end which obviously backfired” Og senere samme dag til samme: “Pwc also indicated to Vidne 3 that they have a number of question about the size of our IBNR for Frida Vidne 3 mentioned that he spoke to you but as you know we have essentially no room to sig-nificantly increase the IBNR” 18 Den 1. december 2016 fremsendte Deloitte (Danmark) på vegne New Nordic Insurance ApS ”Application for Change of Control” til Finanstilsynet med ansøgning om tilladelse til køb af holdingselskabet QIC Holding ApS. Af an-søgningen fremgår bl.a.: “As discussed at previous meetings with the Danish Financial Supervisory Authority where Person 13 and Person 14 participated, New Nordic has entered into a sales and purchase agreement (SPA) with Echelon on the acquisition of Qudos Insur-ance. As part of this agreement, Qudos has prepared quarterly fi-nancial statements as at 30 September 2016 and reported YTD Octo-ber 2016 to the FSA, which will form a basis for Danish Financial Supervisory’s approval of this acquisition. During the preparation of the reporting ending 30 September 2016 and 31. October 2016, Qudos has updated the calculations of the company’s provisions and reassessed the company’s other balance sheet items. As agreed on the meeting with the FSA on 30.11.2016 we have included the October YTD performance as reported to the FSA as well as the projected performance of the company for a full year 2016 as a basis for the FSA to assess the SCR level etc. going forward in 2017. Fur-ther to this, as agreed, we have excluded any new business consid-ered by New Nordic in the business model of Qudos attached in our appendix 25 “Qudos Model ver 1.9.6). Given the current SCR level of Qudos, no new business will be started, before the SCR level is well above 125%. … Qudos represents both a strategic and opportunistic transaction for New Nordic. The acquisition would further our ambitions and accelerate our immediate strategy to establish an underwriting hub in Denmark. Our valuation of the business is conservative given the overexpansion of the business, which has been focused on pre-mium volume above underwriting profits (particularly in UK mo-tor business) and failure of the previous management team with re-gard to underwriting discipline, operational matters and asset management. A new management team has been put in place under Person 1, an experienced actuary from Canada for whom we have professional respect and a good working relationship. We believe we were successful as the bidder given our ability to act in a hands on manner, assist with Danish regulatory dialogue, en-gaging collaboratively with management in the development of a joint business plan for the business and assume certain responsibili-ties for capital and investment management as a partner as well as an investor. 19 We have completed due diligence and actuarial due diligence in order to arrive at our bid, which conservatively implies no value for new business and a significant discount on the legacy portfolio with a price paid for the business below 0.57 x Net Asset Value. … … Following this and within the overall short-term goal of improving the SCR ratio we would take the following actions which management has signaled that it would support: Re-brand Qudos Support new management team Strategic separation of legacy portfolio from new business Review (where required renegotiate or exit) all binder relation-ships Accelerate operation efficiencies – invest in and revitalize IT, reporting and claims systems Facilitate new business in Denmark Restructure reinsurance program to target risk and capital in a smarter way with a view to reduce brokerage costs … … 4.3 Economics We believe Qudos has substantial investment potential. The historically lacklustre returns can be enhanced with a revised underwrit-ing policy (excluding loss making or marginal businesses, particu-larly in the UK motor market), greater expense discipline and en-hanced investment yield on the asset portfolio. Whilst the business has been placed under some stress from his-toric management failings and overexpansion, it is well reserved and is Solvency II compliant, albeit with reserves which have shown deterioration since year-end 2015. We would seek to protect the company from any further reserve deterioration, through the purchase of reinsurance, which the management is already evaluat-ing. This will be viewed in the context of the forward looking strat-egy and the risk that the company seeks to retain.” Der blev afholdt bestyrelsesmøde i Echelon Financial Holdings Inc. den 2. december 2016. Bestyrelsesmødet omhandlede salg af Qudos. Det fremgår af refe-rat fra mødet bl.a.: ”In response to a question, Person 2 advised that Qudos will be reflected in the EFH financial statements as “held for sale” basis on the December 31, 2016 financial statements. The difference between assets and liabilities will be reflected as one line. PWC has indi-cated that they are comfortable with that approach. Since the assets are being held for sale the Appointed Actuary does not need to 20 review the Qudos reserves. Person 15 [Willis Towers Watson] will not opine on the Qudos reserves. Qudos reserves will be deter-mined by Qudos actuary and audited by PWC (Denmark). Person 15 has indicated his agreement with this approach.” Vidne 2, Deloitte DK, skrev den 8. december 2016 til Person 4 og cc. Person 2; Vidne 1, Person 16, Deloitte UK, og Person 3, Deloitte UK, bl.a.: “… I had a good meeting with Person 13, Person 14 and Person 17 today at the FSA. It was a well prepared Person 13 that had been reading the entire CoC application and he had a various of detailed questions that I answered. A few of which we will provide him further details on tonight. Person 13 mentioned some important points – not as much in re-gards to the CoC but to Qudos in general. 1. He think it’s imperative that Person 1 stays on and continues the work and turn around 2. Qudos will continue to be under intense supervisory watch for the next 6 months due to several issues being: a. The FSA is contemplating to require Qudos to report on a monthly basis, technical results pr. Binder including, run-off, insur-ance liabilities etc. b. In general they were unsatisfied with the monthly management reporting c. That they would require He mentioned – even though not promising anything – that 29m. was better than 0m. and a run-off and that the new buyer seemed professional and commitment to get this done.” PwC fremlagde i december 2016 en Audit plan 2016 til brug for Qudos bestyrelsesmøde 13. december 2016. Det fremgår heraf bl.a., at det forventede honorar udgjorde 408.500 kr. Af planen fremgår bl.a. følgende figur: 21 Af referat fra bestyrelsesmøde i Echelon Financial Holdings Inc., der blev af-holdt 14. december 2016, fremgår bl.a.: ” 5.2 Person 18 [bestyrelsesmedlem i Echelon] provided an update on the Qudos transaction. The approval process is moving forward and the parties are working towards a January 20 closing. The terms of a bridge financing loan to assist closing of transaction was still under negotiation. Approximately 70 million Euro funding is required but as best that can be determined New Nordic has raised about 8 million Euro including a planned 1 million Euro contribution by Qudos management. There was a discussion about the security for the loan to New Nordic.” Person 1 skrev i e-mail af 3. januar 2017 til Vidne 4 bl.a.: ” I assume you have been reviewing the month of November all together but the month was overall poor… I am attaching my summary review of the month but from my calculation we would need to look for a reduction in IBNR in the range of 14 mDKK gross of proportional reinsurance to keep the month revenue neutral which would be a tall order …” Vidne 4 svarede i e-mail af 3. januar 2017 under emnet “Subject: RE: Loss Reserve Review” til Person 1 bl.a.: "We are not yet finished with the figures for December. 22 My preliminary and very high level calculations before IBNR adjustments and Sliding Scale adjustments, etc. shows an UW loss of approx. DKK 9m for the month. Meaning that we need to release IBNRs of approx. DKK 20m. As I see it we only have a few positive contributors this month (Patrona, Frida DK and Adrian Flux). All other agents are contributing with negative results. I had an one hour long meeting with the actuaries from PwC today. They have raised concerns about the IBNR allocated to mainly Frida, Patrona and J&M but also raised the question of whether we need to establish a provision for unexpired risk end of 2016. My main concerns are · Frida (there are a lot of focus on Frida these days) · Patrona (largest binder) · J&M – Run-off · Moorhouse – Run-off · Staveley Head (data issues and Largest UK motor Binder) · Provision for unexpired risk” Vidne 3 skrev den 4. januar 2017 en e-mail til Person 1 med emnet ”Reserve review - high-level findings (provisional!)” , hvoraf fremgår bl.a.: “… Just sending this to you for now; here is a view of the overall position. I think we need a top up of 31.4m DKK net. … · 11.4m that we already knew about at Q3 · 5.4m that’s been taken out of IBNR since then · 14.0m for classes that I’ve reviewed for the first time (notably Trent and Frida Norway) · 0.5m of other movements on classes I’d already reviewed Not what any of us wants to hear, but this is how I see it. I haven’t yet taken into account Person 10 warnings that J&M and Flux large loss reserves are likely to need to increase. …” Person 1 svarede samme dag bl.a.: “Ok thanks I guess we need to be prepared for 2 long days That amount will likely put us in run-off so will need shortly there-after a cash flow assessment for each program …” 23 Person 1 skrev i e-mail af 7. januar 2017 til Vidne 5 bl.a.: ”Vidne 5: Hopefully Vidne 4 briefed you up on the call with EFH but essentially they think our solvency ratio should close at 1 or above and equity no less than 220 mDkk.…” Vidne 5, svarede i e-mail af samme dag til Person 1 bl.a.: ”Yes Vidne 4 did inform me about the “constructive” call you had with EFH. No help from them but I do understand some of their points. Please do, and I totally agree with you. You will hopefully get some clarification on his assessments, and be able to challenge him on his view on the IBNR. It is very important that he changes his numbers and not us, not following them. It will be very hard to get PWC to approve that (apart from Patrona where we have done it before and where there is an improvement, which support our view), and perhaps also for the DFSA. … I am a bit disappointed, and must admit that I thought Vidne 3 was more aligned with us, and we have to survive the next couple of months. In theory he is willing to go in runoff, instead of seeing the bigger picture.” Person 1 skrev senere samme dag til Vidne 4 og Vidne 5 bl.a.: ”My cursory review of the reported numbers was leading to the same conclusion that Vidne 3 did take considerations of develop-ment volatility … However I might press Vidne 3 to shuffle more IBNR gross to this account from other programs if we find ways to cut in the other programs.” Der foreligger flere mails i dagene herefter, hvor Person 1 og Vidne 4 skriver om forsøg på at reducere IBNR, ligesom der er opmærksomhed på at holde SCR over 100. Det fremgår videre af e-mails af 10. januar 2016, at Person 2, Echelon Insurance, var orienteret herom. Person 2 anfører således i følgende mailstreng: 24 [ Udeladt mailkorrespondance ] [ Udeladt mailkorrespondance ] [ Udeladt mailkorrespondance ] Person 1 e-mailede den 12. januar 2017 til Person 4, Person 2, Person 12, Person 11, Vidne 4 og Vidne 3 om Qudos Year-End Closing bl.a.: “This afternoon Person 11, Vidne 3 and me met with the Pwc actuarial team supporting Pwc external audit as Vidne 4 was still tied up with the calculation of the SCR. The purpose of the meeting was to explain the basis of our methodology and assumptions supporting the valuation of the claims liabilities. Generally they were satisfied with the level of information and documentation we had supporting our valuation as their valuation seems to be more driven by a straigth selection of development fac-tors from the loss emergence history. Not surprisingly strict reliance on observed development factors derived from a small volatile claims history is driving them to forecast a higher IBNR provision than ours. …” Vidne 2 skrev den 13. januar 2017 til Person 1 bl.a.: “… 25 Further, I would like to discuss the loss absortion in the SCR calculation since our actuary team might have a different view of this in the context of CoC that we would like to have included in the SCR filing by year end. …” Vidne 8, aktuar hos PwC, e-mailede den 13. januar 2017 til Vidne 3 og skrev bl.a.: “I certainly don't expect that you do deliberate cherry-pick. How-ever, I don't agree that the high low is objective – at least not in combination with a volume choice shorter than "all" for triangles where some of the excluded factors by the high-low method are obvious errors in data.” Vidne 3 e-mailede den 13. januar 2017 materiale til Vidne 8 og sendte senere samme dag bl.a. følgende oplysninger: ” As requested yesterday, I’m providing some more detail about the apparent reducing trend in average claim severity over time. … The factors that contribute to this include: - Appointment of in-house surveyors o Frida hired one person in 2015, and two more in 2016 o The in-house surveyors are empowered to negotiate settlements with claimants, and have been able to close claims more quickly, and more economically as a result …” Person 1 e-mailede den 15. januar 2017 til Person 4, Person 2, Vidne 4, Vidne 3 og Vidne 5 bl.a.: “Person 4 Could we arrange for either a Qudos Board or acknowledgement by EFH at 4pm London time to review and approve the statement we plan to file with the DFSA on Tuesday. Following meetings with Deloitte Denmark on loss absorption assumptions in the calculation of the SCR and further meeting with Pwc on their concerns with our valuation of our estimates I think we have most of the information that we need to file with the DFSA on Tuesday morning. Pwc still have a bit of an issue with our estimate for Patrona and would like to see us adding a bit more to them but they also suggested that we could take a more aggressive position on loss 26 absorption in the SCR calculation such that we might ends up increasing our IBNR provision that would further increase our net operating loss by another 3 to 3.5 mDkk but by taking a more lib-eral assessment of the SCR loss absorption assumption we could keep the capital adequacy ratio close to 100%. …” Vidne 7, PwC, fremsendte den 15. januar 2017 PwC’s “questions and request for documentation in respect of the Patrona IBNR.” Person 1 e-mailede den 16. januar 2017 til Vidne 7, Vidne 4, Vidne 3, Vidne 8 og Vidne 6 og an-førte bl.a.: ” Vidne 7: In reply to questions that you and your audit team have raised regarding our valuation of Qudos Claims Liabilities at 31/12/2016, please find enclosed the requested information specific to the portfolio of Irish motor business managed by Patrona Underwriting Limited (“Patrona”). Question 4: We note that Qudos has not included any tail factors for development after 36 months observations. We ask that data and analysis, including any relevant market data or experience, that can document this assumption be presented to us We have included a tail factor; it is a negative one, because we assume that case reserves are in surplus. The discussion above covers the reasons why we think this assumption is valid. … We also have considered external factors such as a study of loss development factors developed by Milliman, an independent actuar-ial consulting firm, on the overall Irish Motor business suggesting the following maturing process for all coverage combined: 27 …” Person 1 e-mailede senere den 16. januar 2017 til Person 4, Person 2, Person 12, Person 16, Vidne 4 og Vidne 3 og anførte bl.a.: “… However, Pwc also indicated that we could be a bit more aggres-sive on the loss absorption calculation which will contribute to re-ducing further the SCR and we would therefore look at the follow-ing adjustment in the calculation of our net capital adequacy ratio: … The recognition of the additional loss absorption within the calculation of the SCR is consistent with the discussions we held yesterday with Deloitte (Denmark). Their issue was that they fell we could be more aggressive on recognizing future income but once we re-viewed with them the current size of the deferred tax asset and the shrinking portfolio that will be in effect from 2017 onward they un-derstood that we have only limited room to improve the SCR be-fore the assumptions start running too far away from the current performance of the business. …” Af Qudos Insurance A/S’ udkast til “Reserve report as at 30 November 2016” , der er udarbejdet i januar 2017 af Vidne 3 fremgår bl.a.: “… 28 In aggregate, my estimate of the gross ultimate claim cost for peri-ods covered by the previous reserve review as at 31 August 2016 has reduced by 12m DKK. The material components of this overall favourable movement are reductions for Patrona (-33m) and Adrian Flux (-8m), … However PwC, in their review of our reserves as part of their over-all yearend audit of Qudos’ financial statements, were unable to agree that there is sufficient quantitative evidence to support this assumption, and therefore Qudos’ Reserving Committee decided to book an additional reserve margin (€3m on a gross 100% basis) for Patrona to provide additional comfort. …” Vidne 8 afgav den 26. januar 2017 et arbejdsnotat vedrørende Qudos. Det fremgår heraf bl.a.: ”As part of the audit of Qudos Insurance A/S (henceforth Qudos), we have reviewed the actuarial reserve reports of Qudos. Our work has been undertaken in close cooperation with the financial audit team. During our review we have discussed our preliminary find-ings with Vidne 4 and Vidne 3 of Qudos. Vidne 8, qualified actuary, CFA has undertaken the actuarial review with support from Vidne 6 (both qualified actuar-ies). We have reviewed • Reserve report and IBNR reserves as of 31. December 2016. … Conclusion Our performed review finds Qudos’ IBNR reserves within, but in the low end, our estimated range of estimates possible using stand-ard actuarial techniques for all the reviewed lines of business. The differences between our estimates and Qudos’ are mostly caused by the actuarial choices on exclusions. This is a field where the ac-tuarial calculations is more an art than a science and opinions on which methodology to use differ. We prefer a methodology with few justified exclusions instead of a mechanical application of a high-low exclusion methodology – especially when the data is lim-ited and volatile due to changes to the underlying processes. We believe that if used, the mechanical high-low exclusion methodol-ogy should be applied after the data has been cleaned i.e. by re-moving high or low factors caused by errors in data or diagonal ef-fects caused by changes to case reserving standards, clearing of 29 back log etc. This is not how Qudos applies the methodology and we urge Qudos to look into this practice going forward. For Patrona and Frida our estimates are much higher than Qudos and for these lines of business we find Qudos reserves in the very low end (or on the lower boundary even) of possible reserve esti-mates resulting from the use of standard actuarial techniques. We reach the same conclusion for J&M TPI: We find the IBNR estimate in the extremely low end (or on the low boundary) of the reserves obtainable from using standard actuarial techniques. …” Vedrørende agenten Frida anføres i notatet bl.a.: “We have recalculated reserves using a standard chain ladder method. We are able to reproduce Qudos’ numbers when applying the same volume choices and exclusions. However, the outliers and the mechanical high-low exclusion methodology applied leaves many exclusions unexplained. In this regard, when looking at the data there are several development factors that appear to be due to data errors (e.g. sep-12