Why Accounting Matters: A Central Bank Perspective

Thank you very much. Thank you also for the invitation and organising this workshop. I am delighted to be here with you and present my humble contribution to the accounting research. To introduce my background, I am also not an accountant by training. I studied financial markets and asset pricing and econometrics. Yet again, my research concentrates on why accounting matters for central banks. And I am sure you are pleased to hear that after we analysed this topic with a broader perspective we concluded that it actually does matter for us as central banks. And accounting is more than just a mere reflection of financial information: It moves market participants and thereby influences our work. I hope I can give you some insights on that today.

Accounting frameworks do influence various activities of central banks and in this paper we concentrate on three different activities, which are the most important ones for a good number of modern central banks. First, monetary policy, and thus ultimately price stability, crucially hinges upon the financial strength and credibility of the central bank. And accounting obviously influences the financial strength of the central bank. So, through this channel accounting frameworks may have an impact on monetary policy or the transmission of monetary policy. To illustrate this, we have done a simulation exercise to collect some evidence on how accounting frameworks may influence the financial strength of a central bank. A second topic that has sparked renewed interest since the financial crisis is financial stability with many central banks being responsible for safeguarding financial stability. Accounting frameworks, as they are applied by commercial banks clearly impact the decisions of bank managers, but also investors who look at the accounts of banks to assess their health or strength. So, through that channel accounting frameworks influence also bank behaviour and thus financial stability. And another topic, that is very important for us now in the European Central Bank (ECB) since we have taken up office on the Single Supervisor Mechanism (SSM), is the micro-prudential banking supervision. Accounting data is basically the basis for the most important capital ratios and for many supervisory reports. Therefore supervisory assessment and analyses are largely influenced by accounting.

Let us get started with the most important area for a central bank – that is monetary policy. As I said before, monetary policy could be influenced by the credibility and the financial strength of a central bank. So, the term that we use as central bankers is ‘financial independence’ of a central bank: Financial independence is crucial to any central bank to be credible and it means that a central bank has sufficient capital resources to carry out its tasks. This is important because central banks, due to the nature of their operations, could face large-scale losses. If a central bank is financially weak, there are basically two consequences this could have. First of all, sub-optimal monetary policy could be conducted. Why is that the case? If the central bank incurs large losses and its capital is negative, governments might not be willing to refinance their central banks. This might put pressure on central banks to pursue sub-optimal policies to go back to the positive territory. But even if the government, who has its own agenda, recapitalises the central bank, there might be pressure on the central bank to conduct a monetary policy that is not optimal. The risk of government intervention could, on its own, lead to sub-optimal policy just because of the fear of losing independence. A second possible consequence of a weak capital position is that even if a central bank conducts optimal monetary policy, it might not be appropriately reflected in market expectations. So if the credibility of a central bank is affected, markets might not expect that the central banks are willing to pursue their goal of price stability and might reflect that in market prices. In other words, the transmission of the monetary policy would, in this case, be impaired. We say that credibility keeps inflation expectation anchored. Empirical evidence shows that there is indeed a negative relation between central bank’s capital position, so financial strength, and inflation. But we are not trying to say that if a central bank is financially weak, it automatically leads to a sub-optimal monetary policy. There have been examples where central banks have done just fine although they had a weak financial position, but also vice versa, it does not mean that a central bank that is financially strong will always conduct optimal monetary policy. Nevertheless it is clear that it is not the optimal position for a central bank, because it can constrain it in its actions or make its actions less effective.

Now, let us try to analyse more the channel of how accounting can actually influence the financial strength of a central bank. What makes a central bank financially strong is, apart from its ability to generate income, the level of its financial buffers available to cover losses. So how does accounting influence these buffers? There are three distinct sets of rules, which influence these buffers.

First of all, accounting frameworks: As you all know as accountants, they stipulate, inter alia, the measurement of assets and liabilities, the income recognition and the ability to set up general risk provisions.

Second, profit distribution rules: For us, the ECB, we distribute profits to the euro area National Central Banks and a lot of the National Central Banks distribute their profits then to their governments. The profit distribution rules in general determine the amount of profit that needs to be transferred to the shareholders and governments and the amount that can be kept and set aside for bad times as below the line buffers.

Third, the loss coverage arrangements: Not all central banks have loss coverage rules laid down in law. For the ECB there is a loss coverage arrangement via the monetary income of the Eurosystem. The loss coverage arrangements, if they exist, tell you whether and how the losses of central banks are covered.

As I mentioned before, to obtain empirical evidence, we have conducted a simulation of the ECB financial results. We compared the Eurosystem’s accounting framework applicable to the ECB and to all Eurosystem’s Central Banks, to IFRS. In addition, we analysed three different profit distribution and loss coverage scenarios. Let me briefly talk you through the scenarios.

The first scenario is the existing rule scenario, using the rules that apply to the ECB. This means, 80 % of the ECB profits are distributed as a minimum and 20 % of the net profit can be put aside for bad times via the general reserve fund. But there is a limit to this: Amounts set aside to the general reserve fund, and the general risk provision, are limited to 100 % of our capital. If we incur losses, meaning if all reserves are depleted we take recourse to the monetary income of the Eurosystem.

And the second scenario, I call ‘loss coverage from profit retention’. In this scenario we assume that the ECB cannot take recourse to the monetary income of the National Central Banks, the NCBs as we call them. Then however no profit would be distributed, if we have accumulated losses on our simulated balance sheet.

The last scenario, I call ‘loss coverage from limited profit retention’; meaning that here again we do not have the monetary income of the Eurosystem to cover losses; however we would always distribute at least 80 % of our profit to our shareholders regardless of whether we have accumulated losses.

For each of these scenarios we simulate results under the Eurosystem accounting framework and IFRS. I am sure as accountants you all know IFRS by heart or at least better than I do, so I am just going to brief you on the difference between the IFRS and the Eurosystem accounting framework.

Unrealised results are treated in an asymmetric manner in the Eurosystem accounting framework, meaning that unrealised gains are taken to what we call revaluation accounts and unrealised losses go to the Profit and Loss account (P&L) if the revaluation accounts built for the same accounts are depleted.

Second, we have the possibility to create general risk provisions, which are forbidden under IFRS. Gold is also treated differently: It is treated as a currency, not as a commodity. However, for the purpose of our simulation we did not take into account the IFRS treatment even in the IFRS scenarios: The problem is that IFRS is inadequate, when it comes to the treatment of gold of central banks. This is because of the massive amount of gold that central banks hold and also the purpose for which central banks hold gold. The central banks are also limited in the amount of gold they may sell at a point in time. That is why we treated gold as a currency to avoid any distortion in the simulation due to the massive upward movement of the gold price over the last years; we treat it in the same way under the IFRS scenarios.

Let us jump to the results of our simulation. On the graph on the right-hand side, I present you the mean profit of the simulation that is conducted for the years 1999–2013. So this is the time series mean that I am showing to you here. The dark blue bars are the results under the scenarios using the Eurosystem accounting rules and the lighter blue bars under IFRS. And what is striking to see, is that under IFRS, the profit is much higher, almost double as high as the Eurosystem profit on average over the years. However, what you cannot see in this picture, but you can see from the figures in the paper is, IFRS figures are also more volatile. This profit volatility is a characteristic that a central bank, which stands for stability, may not want to have. Our conclusion is that the Eurosystem rules effectively smoothen the profit, and from our analysis we can see that this is mainly due to the asymmetric treatment of unrealised results but also due to the ability to build general risk provisions. We can also see that the loss coverage scenarios have very little influence on the mean profit. And, what I do not show here in this presentation, but what you can also see in the paper, is that the results for the profit distribution are very, very similar to the results for the profit.

Now comes the more interesting part, and that is the influence on the financial buffers, which are, as I said before, key for the financial strength of a central bank and thus for its credibility. So, here on this slide, on the right-hand side, the graph is showing the average results of the financial buffers, so it is again time series means that you can see. The financial buffers comprise the revaluation accounts, the provisions and the general reserve fund under the Eurosystem accounting rules. Under IFRS we only account for the gold revaluation accounts and the reserves. We exclude capital from this picture because capital is normally not used to offset losses just like gold that is also not used to offset losses. That is why I also give you the picture when gold re-evaluation accounts are excluded from the buffers. And what you can see here is that the buffers under IFRS are significantly lower. What is also striking to see is that the loss coverage scenarios do not have much impact when we look at the Eurosystem accounting rules. However, when we look at the IFRS scenarios, we see it makes a huge difference in which profit distribution scenario we are in. We conclude from that, that automatic loss coverage as we have it for the ECB is much more valuable for central banks that would apply IFRS. Summing up, we would say that the Eurosystem accounting rules are appropriate for the purposes of the ECB accounts, and, they have been beneficial for building up financial buffers and the ECB’s financial strength.

Let us get to the second topic of interest for a central bank – accounting and financial stability. As I said before, accounting is much more than a mere reflection of a financial position: It influences decisions of management; accounting information is used by investors to assess the health and the performance of private entities. Thereby it also influences their decisions. The most prominent topics that have been debated contentiously in the last years in this area have been fair value accounting and the loan-loss provisioning. And I will briefly elaborate on them in turn. Fair value measurement has been heavily criticised by investors, banks and also policy makers. The said problem is that loss cumulates in economic downturns and this could force banks to suddenly sell assets or reduce lending in an economic downturn and thereby it increases pro-cyclicality. However when you look at the empirical evidence, it is inconclusive with respect to whether fair value measurement exacerbated the crisis. I guess more research is needed on this front. I would assume Mr Yuri Biondi, you would agree on that; and also on how and how much, if it did indeed exacerbate the crises. So, obviously the IASB stance as a defender of IFRS, they say: “Well, volatility is inherent in their business models”, so basically accounting should provide transparency and should not mask volatility. However, they have amended their own rules on fair value measurement several times, so maybe it is also not too convincing. The ECB stance on this is very clear: Fair value accounting is appropriate for a lot of business models and instruments, but it is not always the most relevant measurement at all times. Briefly on loan impairment using the incurred loss model: It was also said to increase the pro-cyclicality, losses were recognised too-little-too-late. IASB stance is that they agree that there were some pitfalls and now they are developing the expected loss model for provisioning. In the literature, there are concerns that the expected loss model could be used for smoothing earnings. We have no empirical assessment of that yet. However the official ECB stance is that we think that the merits are larger than the costs of implementing this new model, and therefore we support this. We believe altogether, it will be conducive to mitigate pro-cyclicality. To conclude on the finance stability part, central banks which are the guardians of financial stability should analyse and take into account the implications of accounting standards when they assess the overall resilience of banks or banking sectors and the financial sector altogether.

Now, very briefly on the last topic, which is the banking supervision. As I said before supervisory reports depend on accounting data. Since the crisis we had many new regulatory requirements. The most renowned are the Basel requirements, which were implemented via the CRD IV package in the EU. And the most central aspect of the Basel requirements is the capital ratios, which are reported under the COREP forms. And when you analyse the components of the capital ratios, you will see that numerators and denominators both highly depend on accounting values. So, really the key aspect of the Basel requirements depends on accounting values. Even more so for the FINREP requirements, the financial reporting requirements to be submitted to the supervisors: These are pure accounting data, without adaptions such as prudential filters.

Another example that I want to give you is the leverage ratio. So, the leverage ratio was intended to complement the capital ratios, which are risk-based, to be a non-risk adjusted measure, which is tangible; simple, which is internationally comparable. However, the US FDIC conducted a study on the largest US banks, the US G-SIBs. And they analysed the leverage ratio that these banks would have had under an IFRS estimate. What you can see is that the results are striking: I mean the differences are from 6.2 % under US GAAP to 3.9 % under IFRS. So what about international comparability? G20 leaders have requested that we create a level playing field in accounting standards and to make them more comparable. However I am questioning that this has happened such that we are at a satisfactory stage. The implications for us, the central banks, first of all, we need to understand the different accounting rules. So even in our own jurisdiction in the euro area we have various different sets of accounting rules. We need to understand the differences; we need to understand how it influences the balance sheets of banks. We should support the harmonisation of accounting standards, which is most desirable for us, particularly for the SSM.

Wrapping-up, why accounting matters? I hope that I could give you some convincing arguments why accounting matters for us as central banks or for us as the ECB. Monetary policy is most important for almost all central banks and can be impacted by accounting; also financial stability is impacted via the accounts of private banks and banking supervision hinges on accounting numbers. Thank you very much.