CRD IV Restoring order in the financial system

CRD IV Restoring order in the financial system Hearing of the Economic and Monetary Affairs Committee of the European Parliament Barbara Frohn May 3 , 2010

Basel III: The individual proposals are directionally correct … SANTANDER shares the goal of the BCBS and the EC to strengthen the foundations of the financial sector and supports the direction of the individual BCBS and EC proposals. However, additional focus could be directed towards the following: q Intrusive supervision has proven to be more effective than (over-) regulation; q In the crisis, weak corporate governance and lack of internal control were key contributing factors, and in some cases even root causes; q The aim should not just be to penalize and avoid bad practices, but equally to give full recognition to, and thus reward, good practices; q Systemic risk cannot be measured by sheer size. 2

…But the Combined Impacts may lead to Unintended Consequences q Possible disruption of the interbank and equity markets i. e. Large Exposures rules, correlations, liquidity buffers, capital conservation standards: will (bank) investor and issuer interests prove reconcilable? q Concentration risk may prove unavoidable; herd behaviour may lead to systemic risk i. e. ´Good Quality´ liquidity & capital restrictions, stringent rules for securitisation & CDS q The CRD IV objective of level playing field cannot be served in a single market where full harmonization has not yet been achieved and where grandfathering may impact differently on banks. i. e. Deductions from Common Equity, Leverage ratio, Grandfathering provisions q By whom will the growth of the real economy be serviced in future and will we then be better off? i. e. Maturity transformation and intermediation role, Leverage Ratio 3

Forward Looking Provisioning: What is the purpose? • Economic activity is subject to periodic cycles: 4 where good years are followed by downturns or even strong recessions. • Banks should, as the ant in the Aesop fable, store up provisions for the “winter” during the “summer”. (i. e. forward looking provisioning). But the current model (“incurred loss” approach, the grasshopper view) does not address this need. • The Spanish model, which uses generic provisions for that purpose, has proved its usefulness to mitigate the impact of the recession on the P&L. • The EC and Santander proposals on this subject are forward looking provisioning based on the Spanish model but taking advantage of the risk data already available for Basel II, and using the concept of through-the-cycle expected loss.

Provisioning: Multiple Objectives, One Solution ? FASB: ´deliver right information on IASB: ´solution to ¨too little, too late¨, but only for closed portfolios´ true financial condition of the firm & incentives to investors´; ´improve incurred loss model by earlier recognition of credit losses´ Forward Looking Provisioning BCBS: ´Capital Stability & Answer to Procyclicality´ FSB: ´better reflection in accounts of underlying economics of lending activities´ EC: ´avoid underpricing & excessive credit growth´ G 20: ´loan loss provisioning to incorporate a broader range of credit information´ 5

Prudential proposals - Strengths and Weaknesses + Combines high provision in good times with drawdown in bad Bank Of Spain: Dynamic Provisioning regime times + Transparent & Simple + Maintains the possibility to move over time to a system in which internal models may be used - Not very risk sensitive: only 6 loan groups - Not exactly an Expected Loss model (statistical model based on historic data) - Requires existence of rich historic data on each loan group EC CRD IV : Expected Loss + provisioning relating Expected Loss amounts at + the beginning of the year to Net Specific Provisions + during the year Simple to implement, and also offers solution to small & medium sized banks with less sophisticated systems Makes optimal use of internally & externally validated Through. The-Cycle parameters already in use for regulatory capital calculations of IRB banks Serves shareholders & depositors interests - May not be fully compatible with expectations of accounting standard setters SANTANDER: Variation of the EC proposal with inclusion of a ´rho´ adjustment factor + Has the same advantages as the EC CRD IV proposal and, in addition: + The ´rho´ factor adjusting the allowance presents a good compromise between the countercyclical smoothing of provisions in the P&L and the need to preserve a minimum sensitivity of the P&L and risk management to the changes in 6

Accounting proposals - Strengths and Weaknesses + Presents an improvement compared to the incurred loss model to the IASB: Expected Cashflow Model (ECF) extent it recognizes real loan costs in the accounts - Primary focus on shareholder interests - Operationally complex and costly: co-mingling of interest rate and credit risk management in one measure - Present Value of changes in ECF must be recognised immediately and thus may add to procyclicality FASB: Modified Incurred Loss approach + Clear and sole focus on shareholder interests - It is questionable whether extending the provision event criteria may substantially remedy the shortcomings of the current model EBF: ´Expected Loss + over the Life of the Portfolio´ model (ELLP) * Adapts well to existing accounting principles; responds to accounting requirement of not exceeding current maturities Remains as yet ambiguous on the calculation method to be used for the calculation of the expected losses over the life of the portfolio; Unclear whether Point-in-Time or Through-The-Cycle parameters are to be used. To avoid divergent implementations, clear instructions should be formulated - Estimating EL over a long horizon (lifetime) is complex, and may still need frequent re-adjustments and thus end up being procyclical after all * This is a mixed model aiming to satisfy both regulatory and accounting demands 7

Main advantages of Expected Loss Provisioning q In times of economic growth a cumulative provision buffer is being created: Ø prevents overheating of the economy Ø excessive credit growth is avoided Ø ´real´ loan costs are being recognized from inception q In periods of recession the accumulated provision is being used: Ø less chance of major capital constraints on banks Ø no sudden lending freeze q As a consequence, Ø depositors that have entrusted their money with a bank can be sufficiently at ease about the solvency situation of the bank in question Ø financial stability will be preserved Ø the negative growth of the real economy is not further aggravated by bank behaviour. 8

on of Minimum Requisites forthe Forward Looking Span Provisioning ish syste m, unif q Uniformity: Aiming at reducing discretionality in the ormi calculation of provisions; applicability to all firms; ty, q Simplicity: Using existing, externally and internally validated, simp parameters which were duly tested during the crisis; keeping licity the horizon oversee-able which reduces the risk of multiple and re-adjustments; tran spar q Transparency: Shareholders´ and depositors´ interests will ency not be served by implementing a dual system; their are investment analysis will furthermore be accommodated by key Pillar 3 disclosures a. o. on risk parameters and other factors supporting the EL estimations. 9

10 Implications of a dual system BASEL IFRS

…and therefore Convergence is the best way forward 11 q Ideally, one unique forward looking provisioning regime should be in place responding to both accounting and prudential concerns; q On a similar note, a ´European only´ solution must be discouraged; q Decisions on a EL provisioning regime should not be made in isolation, but considering other procyclicality measures and capital buffers; q Increasing the provision pool in ´good´ years must also imply being able to consume it in ´bad´ years; provisions must go directly through P&L; q Contradictory incentives: it is evident that EL provisions leading to DTAs should not be deducted from capital; q If ultimately regulatory demands still exceed accounting solutions, the excess should be introduced as an economic cycle provision

So can EL Provisions deal with procyclicality? 12 q Accounting standard setters insist that they I. Cannot accept provisions beyond existing portfolios and maturities and therefore do not promote ´Through-The-Cycle´ provisions II. Aim to provide a true reflection of the firm´s current situation to investors and therefore dislike adjusting the value of a loan for its EL creating a ´day one´ loss q Therefore, only if: Ø Accounting standard setters are willing to compromise and to implement a system which does not lead to continuous fair value re-adjustments flowing through P&L Ø Regulators accept that the average lifetime of the portfolio is not too far off A joint regime may be created resulting in a less marked the objective of a full cycle procyclicality effect, even though the capital regime may still have to be reinforced by mandatory ´Through-The-Cycle´ rating parameters to become less procyclical q However, given the dynamic nature of lending practices and product development procyclicality can never be eliminated in full.

Conclusions and Message If a constructive public-private sector dialogue on the new Basel proposals ensues, (cumulative) impacts are measured extensively and rules are tested before implementation, Then a new homogeneous regulatory framework could emerge that satisfies the expectations of politicians, governments, supervisors and accounting bodies, prevents (most) crises from happening whilst still allowing banks to remain private ventures in a level playing field context that allows the market to properly differentiate between good and bad practices. But for this to happen a silo´ed approach to regulation and standards must be avoided. Only then the financial sector stability can be restored and preserved. 13

Annex A new model for provisioning (I) § No ´in advance´ recognition of credit losses in good years (mispricing) § Strong fluctuation of provisions § Volatility of results and possible loss of market confidence in the institution § Procyclicality: possible excessive credit growth in good years and credit squeeze in bad years

A new model for provisioning (II) § Recognition in advance of credit losses in good years § Use of generic fund in bad years § Flat Provisions (for a stable portfolio) § Stability of Results § Mitigation of procyclicality

A new model for provisioning (III) § Recognition in advance of credit losses in good years § Use of generic fund in bad years § More stable provisions but not flat § Results less volatile § Mitigation of procyclicality