3.1 The link between monetary policy objectives and central banks’ financial strength

The debate regarding central banks’ financial strength and its implications for the successful implementation of monetary policy is not a new development.^[7] However, the recent financial crisis has renewed interest in this topic, since many central bank balance sheets expanded significantly as a result of the various non-standard operations conducted (see Chart 1).

Chart 1

Balance sheet expansion of major central banks.

Central banks can issue base money and earn seigniorage income^[8] at negligible cost. Therefore, under normal circumstances, their long-term profitability and financial strength should be assured and thus be of little concern.^[9] However, empirical evidence has shown that central banks’ financial positions can deteriorate for various reasons associated with the conduct of monetary policy. First, exchange rate stabilisation and revaluation of foreign reserves can also lead to large-scale losses in the longer term.^[10] Second, crisis intervention can be risky. This is because the frequently observed expansion of central bank balance sheets in times of crisis, due to measures undertaken to stabilise financial markets such as the provision of emergency liquidity to the banking sector or large-scale securities purchases, exposes the central bank to credit, interest rate and exchange rate risk. At the same time, provision of capital by the government is often not automatically triggered by the respective legal framework, and even if it were, the government might refrain from recapitalising the central bank because of a precarious fiscal situation or a preference for increased inflation over using public resources or increasing taxes to fund the losses incurred by the central bank.^[11]

If the financial health of a central bank deteriorates, it might seek recapitalisation from the government, which then might try to influence the central bank’s decisions in return for committing public money.^[12] Alternatively, the central bank might try to avoid situations where recapitalisation is required, for fear of government intervention or political pressure. Therefore, the central bank might find itself in a situation where monetary policy objectives are not the only goal pursued and, as a result, a sub-optimal monetary policy might be implemented (e.g. an excessively inflationary policy aimed at generating more revenue for the central bank).^[13] Even if the central bank were to refrain from conducting a sub-optimal monetary policy, this might not be correctly reflected by market expectations, and thus its credibility might be affected.^[14]

Several empirical studies provide evidence that central bank finances affect monetary policy. In a study of 87 central banks, Ize (2006) finds that, in countries where the central bank has negative structural profits, gross domestic product (GDP) per capita tends to be lower and inflation tends to be higher compared with countries where the central bank is financially strong. A negative correlation between central bank financial strength and inflation is also found by Klüh and Stella (2008) in an econometric evaluation of 15 countries in Latin America and the Caribbean. Adler, Castro, and Tovar (2012) investigate the interest rate rules of 41 countries, both emerging and advanced economies, and find that variations in central bank financial strength can explain large negative deviations from optimal interest rate levels.

Clearly, financial strength per se is no guarantee that an effective monetary policy will be conducted. Similarly, financial weakness does not necessarily imply that central banks may not be able to pursue monetary policy goals effectively. Some central banks have performed well in the past even though they were financially weak: well-documented examples are the central banks of Chile, the Czech Republic, Israel and Mexico.^[15]

In a nutshell, the empirical evidence seems to confirm that a weak financial position is not an optimal situation for a central bank because it could constrain policy actions or the transmission of monetary policy under certain conditions. In contrast, a sound capital base increases financial strength and the trust of market participants in a central bank’s policies.^[16]

For the reasons mentioned above, a number of central banks show a tendency to be concerned about their financial position. The ECB has also explicitly linked financial independence with the conduct of an effective monetary policy and central bank credibility.^[17] The ECB has further stated that any situation where the net equity of a national central bank (NCB) of the Eurosystem is below its statutory capital for a prolonged period of time may negatively impact that NCB’s ability to perform its tasks and perhaps affect the credibility of the Eurosystem’s monetary policy. Therefore, an NCB should always be adequately capitalised such that its financial independence is not endangered.^[18] The paper now considers how accounting rules can affect the financial situation of a central bank, making use of a simulation study of the ECB’s financial results.

3.2 The link between accounting and central banks’ financial strength: a simulation of the ECB’s financial results

Accounting frameworks, profit distribution and loss coverage rules for central banks differ around the globe. Therefore, international comparability is difficult and one should always take into consideration the legal and economic environment in which a central bank operates as well as its assigned tasks. Nevertheless, some accounting frameworks may be more conducive to maintaining financial strength than others.

A comprehensive analysis of accounting frameworks implemented by central banks is provided by Archer and Moser-Boehm (2013). In the foreword to this study, Jaime Caruana, General Manager of the Bank for International Settlements, proposes ideas on how to structure accounting rules that could help build trust in the independence and effectiveness of central banks. He recommends that unrealised revaluation gains should not be distributed; however, rather than being hidden by accounting policies, they should be ring-fenced from distribution in a transparent manner. If the distribution framework cannot be adapted in such a way, as would certainly be the case for the Eurosystem NCBs and the ECB, two accounting policies could make sense for central banks, namely the use of revaluation accounts and general risk provisions; both of which are incorporated in the Eurosystem accounting rules.^[19] The following analysis provides empirical evidence of whether this was conducive to the building up of financial buffers with regard to the ECB. But first there is a brief description of the basic principles of the Eurosystem accounting framework as well as of the ECB’s profit distribution and its loss coverage mechanism.^[20]

3.2.3 Simulation results for the profit of the ECB

Table 1 displays the means and volatilities (standard deviations) of the ECB’s profit in the period from 1999 to 2013 under the Eurosystem accounting rules and IFRS for all three scenarios described above. The first row of scenario (i) refers to the actual reported profit of the ECB. One can observe that IFRS would inevitably have led to a higher mean profit, but also to more volatile net results. What stands out immediately is the extent to which the different accounting rules affect the net results. The mean profit over the sample period is about twice as high under IFRS for each scenario. This higher profitability comes at the expense of an increased dispersion – volatility increases nearly fourfold. The different profit distribution and loss coverage scenarios seem to have little influence on the mean profit and profit volatility, as there are only moderate differences between scenario (i) and scenarios (ii) and (iii).

Table 1

Simulation results for the profit of the ECB.

Scenario	Accounting framework	Mean (in EUR billion)	Volatility
(i)	Eurosystem	0.64	1.07
	IFRS	1.26	3.95
(ii)	Eurosystem	0.62	1.02
	IFRS	1.17	3.94
(iii)	Eurosystem	0.62	1.06
	IFRS	1.10	3.93

^[27]

Chart 2 illustrates the differences in profit under the Eurosystem accounting rules and IFRS in the “existing rules” scenario (i) during the period 1999–2013. Scenarios (ii) and (iii) show similar patterns. The IFRS profit is large, both in a positive and in a negative direction, whereas the profit generated under the Eurosystem rules is of a much lower magnitude and does not vary as much.^[27] In other words, the Eurosystem rules lead to a smoothing of annual results; a natural result of the asymmetric treatment of unrealised gains and losses and the build-up of risk provisions.

Chart 2

Simulation results for the profit of the ECB under scenario (i) – “existing rules” (in EUR billion).

3.2.4 Simulation results for the profit distribution of the ECB

Table 2 shows the simulation results for the profit distribution of the ECB. The higher and more volatile profit under IFRS carries over to a higher and more volatile profit distribution. One can see that different profit distribution and loss coverage rules do have some impact on the distributable amounts, but not as much compared with the impact of a different accounting framework. It is clear that the distributed profit under the Eurosystem accounting rules, while lower on average, is generally smoother and thus future shareholder profits are more predictable.

Table 2

Simulation results for the profit distribution of the ECB.

Scenario	Accounting framework	Mean (in EUR billion)	Volatility
(i)	Eurosystem	0.74	0.77
	IFRS	1.86	2.16
(ii)	Eurosystem	0.62	0.72
	IFRS	1.16	1.77
(iii)	Eurosystem	0.62	0.66
	IFRS	1.79	2.11

^[29]

When evaluating the profit distribution under the different accounting frameworks, external recapitalisations (through the pooling of monetary income in the case of the ECB) need to be taken into account, as they effectively represent a negative profit distribution. The “existing rules” scenario is the only scenario where external recapitalisation is allowed. The simulation results are displayed in Table 3. Since the inception of the ECB the monetary income of the Eurosystem NCBs was used twice to offset ECB losses, resulting in an average pooling of €0.10 billion leading to a profit distribution net of external capitalisations of €0.64 billion. Under IFRS, monetary income would have been used five times, leading to an average pooling of €0.64 billion per annum since 1999.

Table 3

Simulation results for the external capitalisation of the ECB.

Scenario	Accounting framework	Mean (in EUR billion)	Count
(i)	Eurosystem	0.10	2
	IFRS	0.64	5

The higher and more frequent recourse to external recapitalisation under IFRS also implies that the impact on profit distribution when moving from the “existing rules” scenario (i) to the “loss coverage from profit retention” scenario (ii) is greater under that accounting framework. However, when moving from the “existing rules” scenario (i) to the “loss coverage from limited profit retention” scenario (iii), the impact is not significant as the possibility of covering accumulated losses with the profits of the year is restricted.

The profit distribution in each year for the “loss coverage from limited profit retention” scenario (iii) is displayed in Chart 3. It confirms that the distributed profit under the Eurosystem accounting rules, while lower on average, is smoother and thus future shareholder profits are more predictable.

Chart 3

Simulation results for the profit distribution of the ECB under scenario (iii) – “loss coverage from limited profit retention” (in EUR billion).

3.2.5 Simulation results for the financial buffers of the ECB

The paper now turns to financial buffers, which are one of the main determinants of the financial strength of a central bank. Under the Eurosystem accounting rules, the financial buffers of the ECB comprise gold, foreign currency and security price revaluation accounts, the general risk provision and the general reserve fund, reduced by accumulated losses from previous years for the scenarios without loss coverage using the NCBs’ monetary income (i.e. scenarios (ii) and (iii)). In contrast, general risk provisions and security price revaluation accounts are absent from the simulations under IFRS rules.^[28] Currency revaluation accounts only exist for gold under IFRS. Table 4 shows the simulation results for the ECB’s mean financial buffers in the period 1999–2013. The financial buffers excluding gold revaluation accounts are displayed separately.

Table 4

Simulation results for the financial buffers of the ECB.

Scenario	Accounting framework	Financial buffers excl. capital		Financial buffers excl. capital and gold
		Mean (in EUR billion)	Volatility	Mean (in EUR billion)	Volatility
(i)	Eurosystem	13.75	9.21	7.41	4.63
	IFRS	7.54	6.00	1.20	1.20
(ii)	Eurosystem	13.31	9.47	6.97	4.97
	IFRS	4.45	7.40	−1.88	4.03
(iii)	Eurosystem	13.12	9.17	6.78	4.69
	IFRS	0.47	4.40	−5.87	4.98

^[31]

It can be observed that the financial buffers are considerably higher under the Eurosystem accounting framework, while mean balances are very similar irrespective of the profit distribution scenario.^[29] In contrast, under IFRS, the profit distribution scenarios do in fact make a substantial difference to the figures. In the “loss coverage from limited profit retention” scenario (iii), the average financial buffers available are even as low as €0.47 billion. The buffers in the corresponding scenario using Eurosystem accounting rules are about 28 times higher. It can therefore be deduced that automatic loss coverage is generally more valuable for central banks applying IFRS.

Due to the large amount of gold holdings and volatility in the price of gold, balances on gold revaluation accounts exhibit considerable variation. Given that, in practice, gold cannot be used, e.g. sold to offset losses,^[30] the same analysis is conducted excluding gold revaluation accounts – see the last two columns of Table 4. The variation in the financial buffers under the Eurosystem framework now decreases. Under IFRS, in the scenarios without loss coverage using the NCBs’ monetary income (i.e. scenarios (ii) and (iii)), the accumulated losses are so high that the buffers become negative on average when gold revaluation accounts are excluded. In scenario (iii) the paid-up capital of the ECB would not suffice to cover the negative buffers thus leading to negative equity on average.

Chart 4 gives a graphical presentation of the ECB’s financial buffers (excluding capital) over the sample period. It reveals that the buffers under the Eurosystem accounting framework are not only higher on average, but also higher in each year of the sample period. Under IFRS, in the scenarios without loss coverage using the NCBs’ monetary income, the buffers excluding capital become negative in several years, despite the inclusion of gold revaluation accounts. Even if one were to allow for the amount of capital, the ECB would have had negative equity in several years.

Chart 4

Simulation results for the financial buffers of the ECB (excluding capital) (in EUR billion).

4.1 Fair value accounting

By reflecting current market conditions and providing timely information, fair value^[32] accounting was intended to increase transparency and thus reinforce disciplining effects imposed by the markets. In perfect and fully efficient capital markets with no information asymmetries, herd behaviour, build-up of bubbles and fire sales or liquidity problems, full fair value accounting can be seen as the most appropriate valuation method. In such circumstances, it should accurately reflect changes in the overall wealth of a reporting entity between two reporting dates. However, given market inefficiencies, some undesirable effects may occur, particularly in the case of financial institutions. For example, whilst recognising unrealised gains eliminates hidden reserves and thus helps improve transparency, making such gains distributable via the profit and loss account without any demonstrable proof that they are readily realisable may conflict with the accounting principle of prudence.^[33] Furthermore, “marking to market” in this way may increase the volatility of the income statement. Pro-cyclicality in bank lending may arise from the immediate and excessive recognition of losses in economic downturns, leading to a reduction in the capital base of banks. Since banks have to hold capital against the risks inherent in their operations, they may then be forced to reduce lending or sell assets. In economic booms these effects reverse such that banks may further expand lending or buy assets.^[34]

Another drawback is that pricing of non-marketable instruments or of assets with illiquid markets is based on models and assumptions, leaving considerable discretion to banks vis-à-vis the measurement of their assets. This increases opaqueness and reduces the comparability of financial statements.^[35]

When the first proposal for IAS 39 was published in 1999, the concept of fair value accounting was heavily criticised by banks and regulators. In particular, the application of fair value measurement to loans or other non-negotiable instruments and to some liabilities was seen as inappropriate.^[36] In response, the IASB introduced the mixed attribute model in 2002, which allows banks to measure certain financial instruments of the banking book at amortised cost, but also includes a fair value option, whereby a financial institution can choose to measure any asset or liability at fair value. Again, the supervisory community and the ECB expressed some concerns.^[37] Taking account of these concerns, in November 2004 the European Commission endorsed IAS 39, with the exception of two “carve-outs” relating to the fair value option and hedge accounting. The latter is not dealt with in this paper. After further changes, the fair value option was finally endorsed in full by the European Commission.

In response to the financial crisis that started in the summer of 2007, amendments to the IAS 39 rules on the reclassification of financial assets have been introduced by the IASB, and adopted by the European Commission, to mitigate the consequences of the crisis and give European companies the same flexibility as companies applying the United States Generally Accepted Accounting Principles (US GAAP).^[38] More specifically, from the third quarter of 2008, banks have been allowed to reclassify certain non-derivative financial assets that were initially classified as “held-for-trading” or “available-for-sale” and measured at fair value, to “held-to-maturity” assets that are measured at amortised cost under certain circumstances. This was intended to prevent the immediate recognition of large losses in asset values and to thereby alleviate a downward spiral.

The severity and spread of the crisis also made policy-makers reconsider the rules on fair value accounting. G20 leaders agreed that the accounting standard-setters should improve standards for the valuation of financial instruments and reduce their complexity, while reaffirming the concept of fair value measurement. Indeed, as mentioned above, the IASB is currently developing IFRS 9 to replace the requirements under IAS 39. But some valid questions remain: did fair value accounting exacerbate the recent financial crisis? And should central banks, as the guardians of financial stability, raise their voices against the excessive use of fair value accounting?

Hoogervorst (2012) argues that fair value accounting played, at most, a minor role in the crisis as volatility was inherent to banks’ business models. Furthermore, financial reporting should contribute to stability by virtue of providing transparency. Stability, however, should not be considered the primary goal.

Some studies suggest that it was not the accounting rules on fair value accounting per se,^[39] but other rules (such as regulatory requirements or collateral requirements) that accelerated the crisis by provoking a downward spiral. For example, banks relied heavily on repurchase agreements as refinancing instruments and also had large amounts of derivative contracts outstanding. Both types of business take into consideration the market values of the collateral provided, leading to margin calls in a market downturn that subsequently result in asset fire sales,^[40] as was the case. In addition, Barth and Landsman (2010) and Laux and Leuz (2010) find that banks had a large portion of their assets in portfolios that were not subject to fair value accounting.^[41] Furthermore, banks were mostly regulated based on prudent values for equity capital, using prudential filters to calculate the regulatory capital. In this way, the recognition of unrealised gains was neutralised.

Other studies find that the accounting rules did indeed have a significant effect on banks’ income statements and may thus have exacerbated the impact of the crisis. Jarolim and Öppinger (2012) empirically analyse the reaction of the European banking sector to the aforementioned amendments to the rules on reclassification of financial assets from October 2008 by focusing on 52 of the 80 banks included in the STOXX® Europe TMI Banks index.^[42] They show that this reclassification option was used quite extensively by some banks and, on average, it avoided recognition of accounting losses of almost €900 million per bank.^[43] The study reveals that banks could have run into substantial problems if the rules had not been amended at the peak of the crisis.

In line with this, Georgescu and Laux (2013) find that three prominent German banking failures involved banks that were regulated using data based on historical costs. In addition, policy-makers reject the notion that historical cost accounting would be more suitable for promoting financial stability in the future. Among others, the IMF (2008) has concluded that despite the problems encountered with fair value accounting, it is still the most appropriate way of valuing financial instruments.

Although some current research fails to find clear empirical evidence that fair value accounting caused or significantly worsened the crisis, many academics find the contrary. Yuan and Liu (2011) find that fair value accounting is generating asset price bubbles and played a major role in the financial crisis. In line with these findings, Biondi and Giannoccolo (2014), among others, argue that fair value accounting relative to historical cost accounting leads to increased market volatility and exuberance and thus to more instable financial systems. Also regulatory and financial stability policy-makers believe that the excessive use of fair value measurement, including the recognition of unrealised gains as income, artificially increases the volatility of profit and loss accounts and may exacerbate pro-cyclicality, which would not be beneficial for the resilience of the financial system. The ECB has itself stated that the “[financial] crisis confirmed the scepticism that fair value is not the most relevant measurement for all asset portfolios at every moment”.^[44] Summing up, rules to limit the excessive use of fair value measurement are promoted by both the academic literature and regulatory and financial sector policy-makers.

4.2 Loan loss provisioning

Under IAS 39, the rules for impairment of loans currently follow the so-called “incurred loss model”. This model assumes that loans will be repaid in their entirety and that actual indications to the contrary must be identified before any impairment losses are recognised. Critics of the accounting standard argue that the recognition after identification of evidence, such as a counterparty failing to meet its contractual obligations, is much too late because the expenses in the income statement for impairments then accumulate in economic downturns when losses materialise. This provisioning regime therefore increases pro-cyclicality. In good times, when lending is already at a high level, banks are not required to set aside buffers for expected losses, and thus overstate the economic value of the loan portfolio and understate losses in the income statement. As a result, lending can be expanded beyond the amount that would be possible under a different accounting regime. In economic downturns high credit losses occur, but the lack of available provisions increases the losses reported in banks’ income statements, which reduces capital and may force banks to recapitalise or reduce lending and sell assets. Hence, provisions set aside in good times could serve as a buffer against risk; one that alleviates the impact of these effects and reduces the likelihood of banks becoming insolvent.

In contrast to the incurred loss model, the so-called “expected loss model” implements a forward-looking methodology. Under this model, impairments can be made in a timelier manner, potentially dampening pro-cyclicality. Credit losses that are expected to occur are reflected over the life span of credits, providing useful information for investors.

Naturally, this raises the question whether the expected loss model will actually achieve these desired effects in practice. In 2000, Spain had already adopted a dynamic provisioning approach that allowed for early recognition of future credit losses. Previously, Spanish banks had had the lowest ratio of provisions to non-performing loans amongst the Organisation for Economic Co-operation and Development (OECD) countries,^[45] whereas, by 2006, this ratio had become the highest among Western European countries. Nevertheless, there have been a vast number of write-offs of Spanish banks since the start of the crisis due to the real estate crisis in Spain. Hence, did dynamic provisioning fail to prevent the crisis for Spanish banks? It is obvious that the situation for Spanish banks would have been considerably worse without the large buffers set aside in good times. Moreover, Fernández de Lis and García Herrero (2009) find that the Spanish provisioning model reduced pro-cyclicality but did not eliminate it.

Meanwhile, Fillat and Montorial-Garriga (2010) analyse the effects that the Spanish provisioning model would have had on US banks during the crisis. In this hypothetical situation, about half of the US banks that received government support would not have required it. However, the provisions would not have been enough to cover all losses incurred, suggesting that the Spanish model is a good way of covering losses in “average” downturn periods, but would not suffice for a financial crisis that is as severe as the most recent one.

Chan-Lau (2012) notes that, in theory, dynamic provisioning should weaken pro-cyclicality. However, in a simulation exercise applied to the Chilean banking sector, he finds that while dynamic provisioning increases the resilience of banks it may not dampen the credit cycle. Therefore, dynamic provisioning might not be effective for dampening pro-cyclical effects in all jurisdictions and policy-makers should not rely solely on this approach to solve the problems encountered during a crisis. Instead, Chan-Lau recommends that additional counter-cyclical measures, such as the regulatory buffers proposed by the Basel III regime, should be considered. In line with these arguments, Hoogervorst (2012) states that: “The lesson is that economic cyclicality can be too powerful to be dented significantly by mere accounting”.

While the advantages of the expected loss model for the soundness of the financial system are acknowledged, there are certain critical aspects that should be considered when judging this model. Bushman and Williams (2012) examine banks across 27 countries and find that discretionary forward-looking provisioning can also be misused to smooth or disguise earnings. This could weaken market discipline, as transparency and comparability of financial statements may then be reduced.

As mentioned above, the IASB aims at introducing a forward-looking impairment method during the course of 2014. In general, academics, the banking sector, central banks and regulatory authorities largely welcome the introduction of an expected loss provisioning regime. Yet, the details of how this should be implemented remain a matter of discussion.^[46] Nonetheless, the ECB has expressed its support for moving to provisioning based on the expected loan loss.^[47] It is believed that adequate forward-looking provisioning rules will mitigate pro-cyclicality and dampen the negative effects of economic downturns, thereby contributing to the solvency of banks and the soundness of the banking system.

4.3 Implications for central banks

Given the role played by fair value measurement and the loan impairment methodology in the recent financial crisis, there are two main conclusions to be drawn by central banks acting as guardians of financial stability.

First, when assessing the overall resilience of the financial sector, the implications of accounting frameworks should be taken into account.^[48]

Second, central banks should assess accounting standards applicable to the entities in their jurisdiction from a financial stability perspective, and contribute to the public discussions on accounting standards as well as the standard-setting process.

Criteria to assess accounting standards from a financial stability perspective were presented by the European System of Central Banks (ESCB) Banking Supervision Committee in its report of December 2006 on the “Assessment of accounting standards from a financial stability perspective”.^[49] The report identifies the risks to financial stability stemming from the accounting treatment of the banking sector. The Maystadt report requested by the European Commission (see Maystadt, 2013), which focuses on strengthening the EU’s voice in the field of accounting and financial reporting, even proposes formalising financial stability considerations in the IFRS adoption process in the EU by amending the IAS Regulation.^[50] It suggests introducing two additional adoption criteria for IFRS standards in the EU, namely that the standards should not endanger financial stability or hinder economic growth. In addition, the report proposes to reform the European Financial Reporting Advisory Group (EFRAG).^[51] It suggests, inter alia, giving the ECB and other European supervisory authorities the right to be represented on EFRAG’s supervisory board. In this way, the ECB could formally contribute to the comment letters of EFRAG addressed to the IASB and also to the advice to the European Commission on the endorsement of standards.

5.1 The European System of Financial Supervision and the new role of the ECB

In response to the financial crisis a new institutional architecture was developed for the financial sector of the EU: the European System of Financial Supervision (ESFS) which started operations in 2011 and focuses on macro-prudential and micro-prudential supervision. Responsibility for the former lies with the European Systemic Risk Board (ESRB), while micro-prudential supervision is conducted by the three European Supervisory Authorities (ESAs), i.e. the European Banking Authority (EBA), the European Securities and Markets Authority (ESMA) and the European Insurance and Occupational Pensions Authority (EIOPA).^[52]

To promote further integration of the European financial system and restore market confidence, the European Commission proposed the adoption of a banking union in September 2012. The SSM is an integral part of the banking union^[53] and assigns specific banking supervision tasks to the ECB, and its work formally commenced in November 2014. The ECB is responsible for the direct supervision of significant credit institutions^[54] and also for the effective and consistent functioning of the SSM. As such, it grants and withdraws authorisations for credit institutions, carries out supervisory reviews, ensures compliance with the relevant legal acts, and cooperates closely with all other institutions of the ESFS. It will also work closely with national competent authorities with regard to the supervision of all other banks. Consequently, national authorities are continuing to play an important role in the euro area. The ESFS is depicted in Figure 1.

Figure 1

European System of Financial Supervision.

5.2 Regulatory requirements and their reliance on accounting values

Given that accounting values are the basis for calculating regulatory capital ratios and preparing a number of supervisory reports, accounting rules can substantially influence the work of central banks in charge of banking supervision. The paper now examines the reliance of supervisory reports on accounting data and the differences that accounting frameworks can make to supervisory ratios, using the leverage ratio as an example.

In the wake of the financial crisis, various initiatives to strengthen regulatory requirements and develop additional reporting requirements have evolved. Probably the best-known one is the Basel III framework developed by the Basel Committee on Banking Supervision, which aims at strengthening the quality and level of the capital base of financial institutions by increasing own funds requirements. In addition, Basel III introduces multi-dimensional regulation and supervision, with capital, liquidity^[55] and leverage ratios covering the whole balance sheet of banks.

In the EU, Basel III is implemented by the so-called “CRD IV package”, consisting of the Capital Requirements Directive^[56] and the Capital Requirements Regulation^[57] (CRR), which took effect on 1 January 2014.^[58] Accounting rules are of great importance because the mandatory supervisory ratios and reports specified in the new directive and regulation are largely based on accounting data, as discussed in more detail below. Even in cases where accounting data are replaced by prudential data in the context of regulatory requirements, accounting rules may still have an impact on regulatory capital ratios.^[59]

5.3 Implications for supervisory authorities

Banking supervisors have good reason to call for the application of harmonised accounting frameworks. The supervisory assessment of banks largely depends on the regulatory ratios and reports that are submitted to them. And, of the regulatory ratios, the pivotal capital ratios and the leverage ratio are determined using accounting figures. COREP reports contain, to a large extent, adjusted accounting figures, and FINREP is based solely on accounting data. Furthermore, not only do these regulatory ratios and reports mostly depend on accounting figures, but also, as shown by the FDIC, the type of accounting framework implemented can make a substantial difference to the numbers. Consequently, an unequal treatment of commercial banks by a supervisory authority might be the result of the application of different accounting frameworks.

Therefore, it is important for supervisors to understand the accounting rules behind the figures and to promote sound accounting practices. G20 leaders had proposed the development of “a single set of high-quality global accounting standards”, which would facilitate comparability of accounting data and supervisory ratios. Greater harmonisation of accounting frameworks is also supported by the ECB.^[70] Given that IFRS in the EU apply only to the consolidated accounts of listed companies, not all of the banks here have set up their accounts according to IFRS. There is a risk of not fully understanding the differences in accounting rules among Member States. Consequently, from the SSM perspective, it would be preferable to require that the banks to be supervised apply harmonised standards such as IFRS, and thereby ensure equivalence in any regulatory assessment of data dependent on accounting figures.