> Retouradres Postbus 202012500 EE Den Haag
SNS REAAL N.V.
t.a.v. de heer Latenstein van Voorst
Croeselaan 1
Postbus 8444
3503 RK Utrecht
Ministerie van Financiën
Directie Financieringen
Korte Voorhout 7
2511 cw Den Haag
Postbus 20201
2500 EE Den Haag
Inlichtingen
dhr. v. Hüzelr
Date
Subject
28th of January 2013
Cushman & Wakefield Report
Ons kenmerk
FIN/ U
uw brief (kenmerk)
Bijlagen
Dear Mr Latenstein van Voorst,Dear Mr Lamp,
The Ministry of Finance has received your letter dated 13 January 2013 in which you state that bath the procedure foliowed and the methodology used in the creation of the Cushman & Wakefield (C&W) report are seriously flawed. We have relayed your arguments to our advisers C&W, Morgan Stanley and to professor Tom Berkhout, chairman ofthe Competence Center Real Estate ofthe Dutch Tax Authorities. This letter servestoshare their views and arguments with you and to enable you to reconsider your views.
In your letter you raise several concerns with respect to the use of information, appfied methOdology,calculated discount rate, potential double countlng of risk factors and the estimatlon of probability of default (PD) and loss given default (LGD) figures. We have critically assessed your points of attention and think it would prove insightful to give addltional information on the aforementioned concerns. This elaboratlon Is included as an annex to this letterand shows that C&W has made a thorough analysis of the Property Finance portfolio (PF portfolio). We hope that this explanation also ciarifJes your concerns.
Besides comments on the applied methodology by C&W, you also raise several concerns regarding the process ofthe C&W valuation in your letter. Therefore we believe that a clarification on the part of the MInistry of Finance could be useful. Due to the losses on SNS REAAL's PF portfoJJo, the capita! position of SNS BANK has increasingly become a concern. As a consequence the Ministry of Finance and DNB have investigated potentlal stabilizing solutions for SNS REAAL in theevent that the situation woufd continue to worsen. An important step In this process was to assess the impact ofthe potentlal PF losses on SNS REML's capita! position.
This has led the Ministry of Fina nee to hire C&W as a renowned and independent
expert on real estate loans and objects for an assessment of the Real Economie VaJue (REV) of the aforementloned portfolio in fine with the requirements of the European Com.mission. The aim of thls analysis was to make an independent assessment and nat to validate any (internal) review ofthe same portfolio by yourself,Ernst & Young (E&Y) or any other party. Such an undertaking would not
be meanlngful due to dffferences in scope and methodology applied by for example
the research done by C&W and E&Y; the farmer has focused on calculating a REV
value of the loan portfolio. By contrast, E&Y has focused on estimatlng the
potentlal (provisioning) shortfall whlch In ltself gives insufficient information on the potentlal transfer price that is warranted for such a portfolio.Understandably these discrepancles can lead to dlfferences between outcomes. Any dlfference in fees
paid to E&Y and C&W Is irrelevant in this respect as the scope of bath researches is different.
C&W informs us that lt has made a thorough analysis on the basis of the available Information provided by SNS REAAL, meetings with managementand your employees,and an expert meeting wlth E&Yon the interpretation of thelr research results. Partsof the dataset constructed by E&Y proved useful for the analysls of C&W, particularly the information on the collateral value of specific real estate loans. These parts are therefore also used In the C&W research.
Currently C&W is working on an update of the findingsin which the value of the PF portfolio will be calculated using the most recent data tape provided by SNS REAAL and the most recent forecasts for macro-economie parameters by DNB's macro economie department and other independent organlzations. It is worth mentlonlng that an extrapo lation of the applied discount rates and expected lossesin the PF portfolio to other outch market parties is nat valid; these parties each have
specific laan portfalias wlth different risk characteristics which makes a projection of implied losses on a larger scale meaningless.
The Ministry of Finance has been as transparent as possible during the processin termsof the applied methodology by C&W and the alm of the conducted research. C&W has given a presentetien on the methodology and also on the outoornes of thelr report and ca refully assessed your concerns. In addition you have also recelved a copy for your own assessment as did the European Commission on the same day. The report is prepared for the Ministry of Flnance and we see no added value of distributing the report to ather parties than the ones to whom the report already has been dlsclosed.
We hope that thls letter has given you additionallnsight in the background of the foliowed pro dures and outco mes of the C&W report.
wURaab
Directer Finandng Directorate
Directie Financleringen
Onskenmerk
FIN/ U
Please flnd in this addendum further explanatlon and background to the points of concern raised in your letter dated 13 January 2013. This addendum deals with five topics:
1. Use of information
2. Applied methodology
3. Calculated discount rate
4. Potentlal double cou nting of risk factors
5. Estimation of PD and LGD figures
C&W reviewed information provided by E&Y as well as documents that were made available on the Merrill Data Site, maklng extensive use of this information, in partlcular the collateral data reports. The data provlded formed part of the inputs for C&W's financial models. The E&Y information was helpftll and provided insights that C&W would norma lly have sought to extract from credit and collateral files. This has been an important reason why the start of the C&W trajectory was timed to take place shortly after E&Y had prepared thelr analysis.
E&Y provided C&W access to collateral information that it had put into Excel format for their analysis which made the processing of this material more time efficlent. However C&W raised saveral written questions with E&Y and SNS REAAL that remained unanswered o r not met in a timely manner.
C&W was invited to provide an assessment of the REV-value of SNS REAAL's PF
portfolio under a base and an adverse scenario. Our advlsor was nat engaged to provide a loall loss provision analysis or to validate the work of another party. we understand that E&Y was engaged to provide advlee on loss provistons (shortfall) for the top relationship complexes only. In your letter it is inferred that E&Y has valued the laan portfolio and that therefore lts results might be camparabie to those produced by C&W. The objective of E&Y's reports was to estimate the potentlal shortfall of the selected RCs. The comparison of results of E&Y's analysis with those of C&W's analysis is therefore flawed. With raferenee to the macroeconomie estimates used by C&W it is worth mentlening that C&W based their projectlans on EC foracasts where they are available and a consensus view where they are not.
Directie Flnanciérlngen
Ons kenmerk
FIN/ U
C&W based thelr estimatlon of the discount rate on their extensive professional
experience in the valwation of bank loans. They used a discount rate which Is derived from saveral components to reflect the specific risk which is attrlbutable to individual loans (examples of such specific risk components are the riskfree rate and the locatlon of the real estate collateral). These estimates are based on observable market parameters and C&W's professional experience. The different
riskprofiles of the loans result in an a ppraprlate average discount rate of /o).
C&W would have calculated a lower discount ra te if the portfolio contained mainly good quallty real estate investment loans and less development loans. The collateral of the portfolio is however of relatively poor quality; an analysls of c&W shows that the percentage of development ioans is very substantial and the average laan to value (LTV) ofthe loans is high. In addition the portfolio contains a
,_
large amount of non-performing commercial real estate loans (approximately Directie Flnaocleringen
./o).
Dns kenmerk
C&W calculated a discount rate whlch Is used in both the base and adverse case. The mentioned scenarios however lead to different expected cashflows. C&W prevented double countlng of risk factors by differentlating between the estimation of the discount ra te (which is unaffeeted by expected cashflows), and the
projection of cashflows (which are based on different macro-economie
assumptions).
LGD is calculated for the entire portfolio based on the individual and thorough analysis of the non-performing loans. Our adviser derived the PD-rates from a qualitative and quantitative analysis of the available data In relation toa representative sample of loans from each bucket. These PO figures have been estimated on the basis of interest coverage ratlos (IeRs) and debt service coverage ratios. C&W applied the same expected default rates in both the base and adverse case. The LGDs however are substantially higher in the adverse case due to the economie scenario.