The European Commission accepts that EU law requires accounts to provide a “true and fair view” of assets, liabilities, financial position and profit or loss as specific indicators of financial health of companies, including banks. That is something that that IFRS accounting rules do not do, leaving banking stress tests to try to do that job instead. President Draghi accepts the legitimacy of concerns that a new accounting standard may threaten financial stability.
Central banks now rely on the stress testing of large banks as the primary tool of bank health assessment. We, and other commentators, have previously expressed concerns about the tendency to skew results in favour of positive outcomes, the methodology, the choice of hypothetical adverse economic scenario, the pass/fail mark. However, perhaps the most understated concern is the actual data used, the numbers transposed into the stress test models from the accounts and books of banks.
Two recent news items go to the root of the controversy about Europe-wide proposed accounting rule changes. The background to both items is the implementation of a controversial new accounting standard called IFRS 9. Of course, we accept that accounting rules are not as exciting as revelations such as the “Panama papers”, but given the importance of this subject to the recovery or otherwise of our banking system, the media coverage was unimpressive. The news received headlines only in technical, professional media, not in the mainstream financial press.
The two items of news are:
1) A reply to a written request to the European Commissioner for Financial Stability, Lord Hill, which a British MEP has recently received. More on this later.
2) ECB President Draghi responded in writing to a letter from the chair of the European Parliament’s Committee on Economic and Monetary Affairs (ECON), expressing concerns that IFRS 9 might negatively affect the stability of large banks. President Draghi’s letter accepted the legitimacy of these concerns. He promised to initiate a new work stream to examine whether the “broader effects of fair value accounting" undermine the reliability of matters such as loan impairment reporting. He anticipated that he might be in a position to respond further to ECON sometime during 2017.
The background to this is that tensions are now detectable between two groups of members of ECON. A recent report by the rapporteur to the ECON committee, Theodor Stolojan, former Romanian Prime Minister and now an MEP, appeared to downplay the concerns about accounting standards, as well as criticisms of the governance of the private sector standard setter, the IFRS Foundation:
“The International Accounting Standards Board (IASB) based in London, is developing a set of high quality globally applicable accounting standards, and is therefore creating a common language for global financial reporting. The rapporteur acknowledges the importance of such globally applicable standards for improving competition and removing investment barriers on a global scale.”
Noted IFRS sceptic, German Green MEP Sven Gielgold, called for comments on this report, a call willingly taken up by the Chairman of one of the UK’s leading public sector pension fund representatives, LAPFF. This organisation has recently criticised not only the IFRS accounting rules, but also its governance. LAPFF, together with prominent UK shareholder representative the Pensions Investment Research Council, has led the criticism of both IFRS 9 and the standard being replaced, IAS 39. The problems with the old IAS 39 standard are well documented, the most prominent of which were:
a) it was used by banks to substantially reduce their loan loss provisions. Certain clauses and examples within the standard were deemed to justify booking a provision only when actual evidence of loan impairment was available, such as a missed payment. This resulted in gross under provisioning for likely future loan losses.
b) the standard also was interpreted as requiring banks to mark to market their own liabilities. This led to the bizarre result of banks reporting profits when market actors sold bonds as confidence in the creditworthiness of such banks declined.
IFRS 9 mandates the booking of loan losses expected in the next twelve months. Clearly this is more realistic than IAS 39, but banks have generally chosen to recognise one year forward looking losses because this is a requirement of the Basel rules on capital. Critics of IFRS 9 point out that by ignoring post-12 months forward losses then reported assets and capital are still going to be overstated.
According to LAPFF and the legal opinion of George Bompas QC, a Deputy Judge of the UK’s High Court, the implications of these and other deficiencies of IFRS are very significant for banks. In a December paper they stated:
“it is difficult to assert that accounts which fail to enable a determination of what is or is not available for distribution by reference to amounts stated in them, can give a true and fair view of a company’s assets and liabilities, financial position and profits or losses”.
Returning to the first news item. The British MEP is Syed Kamall. He recently obtained a written confirmation from Commissioner Lord Hill, that the process of endorsing accounting standards must respect EU law; accounts so produced must provide creditors with a “true and fair view” of the “undertaking’s assets, liabilities, financial position and profit and loss”.
The EU is presently ratifying IFRS 9 based on advice from EFRAG. EFRAG do not appear to have accepted to date that the “true and fair” legal constraint applies to the specific accounting numbers generated by IFRS standards, preferring a looser interpretation of the test, which they have defended in reams of submissions. Furthermore, EFRAG has not addressed the requirement of EU law that the primary purpose of accounts is the assessment of solvency.
Following Commissioner Lord Hill’s letter to Syed Kamall, LAPFF wrote to the Commissioner urging the suspension of the endorsement of IFRS9. Their letter makes clear how high are the stakes:
“We cannot exaggerate how serious defective accounts are…they enable value destroying business models to pass off as profitable whilst in truth they are secretly consuming capital.”
The ECB confirms that non-performing bank loans will be allowed as collateral for refinancing purposes. Collateral valuation rules for “held to maturity” bank loans mean that nominal values, not market values, of assets are used. Is the ECB planning in this way to bail out certain banks or even national systems?
In which side’s favour is the above dispute over the process of accounting standard setting likely to be resolved? We have no doubt that every effort will be expended to fudge the issue, for three reasons:
1) There will be widespread political resistance to a significantly harsher loan loss recognition standard than the presently drafted IFRS9;
2) The ECB and related bank supervisory institutions will be concerned that the reductions to bank capital levels necessitated by the publication of more realistic expectations of future loan losses would lower confidence in several banks;
3) The ECB recently confirmed its willingness to refinance certain banks’ non-performing loans. We think the price they will use will be closer to par than market values, but the ECB efforts to support such banks will be undermined if stricter accounting rules are applied. We now explain.
We reported in February that the Italian government had obtained European Commission approval to help support its banks by selling them government guarantees covering their non-performing loans. We suspected that these loans would be refinanced by either Italy’s national central bank or even by the ECB. Now it looks like the ECB will fund them.
When asked whether the ECB would “buy” non-performing loans under its asset purchase programme, President Draghi denied that the ECB would buy loans, but would allow the loans as collateral for refinancing purposes. How might this work? It is established practice that banks can obtain credit by pledging “acceptable” collateral to the ECB via the intermediation of their national central banks (NCBs). (For brevity we will hereinafter ignore the NCB intermediation).
The key issue is the valuation assigned to this collateral by the ECB. The rules establishing this were introduced in 2011 by President Trichet, the predecessor of President Draghi, presumably prompted by the realisation of the scale of the problems of Greek banks.
“Securities classified as held-to-maturity shall be treated as separate holdings, valued at amortised costs and subject to impairment. The same treatment shall apply to non-marketable securities.”
We discussed impairment in the context of IFRS rules above. Under the rules, if a troubled borrower has failed to pay any of — say — a recently missed 5% annual interest payment on its €100 loan, then the bank can minimise its “impairment” of the loan, and maximise its collateral value for refinancing purposes by writing down the loan only to the extent of the interest payment due in nearly one year’s time. The loan could be presented for refinancing at an impaired amortised cost of €95. After adjustment for the “risk premium” of say 10%, the ECB would provide refinancing credit of €85.5 against an asset with a much lower market value.
It therefore appears that banks may be able, under the recently announced expansion of the ECB’s asset purchase programme, to refinance non-performing loans at close to par value, even though they are worth a lot less. This creates a tension between the interests of creditors, who want the accounts of banks to report realistic “true and fair” numbers, and the interests of the ECB, NCBs and politicians, who believe that the expanded asset purchase programme can deal with the problems of banks. Lord Hill is surely now working hard on the fudge.