Fashion, Trends and Overshooting: Accounting standards for central banks


It is a well established part of the human psyche to be more comfortable when one is in the company of others, part of the consensus, following the common trend. Fashion houses rely on it, markets give ample evidence of it (witness the well known saying “the trend is your friend”).  But it is also well known that trends can go too far:  what starts as a logical and good idea is often taken to excess. The creations of the catwalk become ever more extreme; markets periodically overshoot.

This short paper will argue that central banking is no less subject to fashions, trends and indeed overshooting, and one such fashion, the call for central banks to be accountable and produce standard financial accounts, has now gone too far. The conclusion is that what was undeniably a good idea initially has now overshot.

Trends and fashions in central banking

That central banking exhibits trends and fashions cannot be denied. 25 years ago the orthodoxy was for monetarism, and central banks based their monetary policy decisions largely – in a few cases almost exclusively – on changes in the money supply.  This paper does not intend to debate the rights and wrongs of this approach, but one notes that fashion has moved on from an exclusive reliance on the money numbers, and few central banks would claim to be even partly monetarist now.

15 years ago, the orthodoxy changed to inflation targeting, independent central banks and (as a corollary) increased security of tenure for top central bankers. Again, there is much merit in the main argument, that central banks should have the control of inflation as their main objective and should be able to pursue this free of the political need to bend to the latest public opinion poll.  So strong was the trend that anyone who raised even legitimate concerns about the democratic deficit and the suitability of a largely unaccountable public body in modern society was shouted down.  Yet even this trend could overshoot, and few will now argue that absolute security of tenure of office is ideal, in the light of events at the Banca d’Italia earlier this year.

About 8 years ago the fashion changed to a drive for greater transparency. Central banks were urged to report more openly, in more detail and more quickly, what they were doing.  Once more, the idea was grounded in good sense:  the Asian crisis of 1997-98 showed that too many central banks were opaque and secretive, and too much of the data they issued was incomplete, misleading or simply too late to be useful.  But once more, once the idea gained common currency it became extremely difficult to argue against it:  witness the ECB’s long (and in this author’s opinion entirely justified) struggle to resist the pressure to publish minutes of Council meetings.

At the heart of the ECB’s defence is the fact that ultimately, transparency, or the act of seeing what the central bank is doing as it is doing it, is not a neutral act. It is a well-known phenomenon of physics[1] that the act of observing sub-atomic particles affects those particles and changes their speed or position. So also with those determining monetary policy:  the act of being observed or reported on risks changing both the debating stance of the participants and the final result of their deliberations.

The ECB’s long and principled stand has brought realisation that, while transparency in central banking is in most cases a good thing, it is not universally desirable and should not be taken to extremes. Like the preparation of much of our food, or a major operation in a hospital, some aspects of central banking are best judged by their results not observed as a process, and although many countries show their parliamentary processes on live TV (and some even show their courts of justice) no-one has taken the trend of central bank transparency to its logical conclusion of live TV at the FOMC or MPC!

The latest fashion – full IFRS-compliant accounts

The latest fashion to engulf the central banking world is for central banks to produce full mark-to-market financial accounts. On the surface, there is nothing unusual in the requirement that central banks should produce accounts.  They are, after all, financial corporations, and – unlike departments of central government, for example – can draw up both a balance sheet and a profit and loss account in a fairly straightforward manner.  And if they can, the argument goes, then they should;  and furthermore, these accounts firstly should be as comprehensive as possible, including a full analysis and valuation of the FX reserves for those central banks that own their country’s reserves themselves, and secondly, should use accepted international accounting standards and methods.  Transparency and accountability – the touchstones of 21st century central banking – are thereby enhanced all round.

This logic is extremely seductive. Who, indeed, could argue with it?  But this paper will contend that it results in a major anomaly, is at heart based on a major assumption that is at best debateable and may even be counter-productive and damaging, and has a major unintended consequence.

The anomaly of a central bank’s accounts becomes more apparent when one moves from considering the central bank as a financial institution and thinks of it more as a policy organ of the authorities.  Although it is an anathema to most central bankers to consider themselves on a par with the rest of central government, a central bank is in fact at heart a policy-making body, whose true worth is in most cases measured by the effectiveness of their policy operations rather than their profit and loss accounts.  In this light, the central bank is alone as part of the authorities in attempting to produce full accounts:  government departments typically do not do so, except for their incidental expenses, and it is only because a central bank conducts its policy operations in the monetary sphere that it is able to produce financial accounts of those policy operations[2].

The assumption behind the production of central bank accounts is that the accountability of a central bank is improved by its issuance of accounts.  The neatness of the word association (in the English language at least) leads one to make this assumption almost automatically.  Yet accounting standards are by and large designed for companies whose objective is making profits.  If one’s objective is to make profits, it is fair to be measured by a system of rules and customs that records one’s success in doing just that.  But for those who have other, non-financial objectives, it is not obvious that financial accounts are a suitable measure of their success or even able correctly to record what they do[3].

And this leads on to the unintended consequence of producing accounts, which is that central banks risk being assessed on the basis of those accounts.  Once the central bank has issued financial results, it stands open to comment and criticism on the numbers.  Even when producing a profit is not the main objective of the bank, it will increasingly be at risk of comment on its financial outturn from uninformed or malicious commentators, and it is the fortunate few central banks who are so sure of their reputation that they can wholly ignore the risk of negative comment arising from a large loss.

This turns the accounts from being merely neutral records of decisions already taken, into an influence for shaping decisions as well. As I observe in an earlier paper (Nugée (2003)), this influence can take two forms:  the Weak Form, in which the central bank is discouraged from doing the right thing because of the accounting consequences, and the Strong Form, in which it is led to do the wrong thing to achieve an accounting result.  Neither are in the best interest of sound policy making.

This is a particular challenge for those central banks for whom the FX reserves are a dominant part of their balance sheet.  Reserves are usually held not for their financial value but for their operational or policy value, and while the monetary value of the reserves is not wholly irrelevant, their true value is in their policy effectiveness (for example, the amount of confidence they give to an exchange rate regime or the amount of intervention they represent).  It is debateable, at best, whether such qualities are best measured in domestic currency terms.  Furthermore, returns on the reserves can swamp the financial pluses and minuses from other operations[4], and large unrealised profits or losses on the reserves arising from currency moves can both hide inefficiencies elsewhere in the central bank’s finances and also – in the case of profits – result in pressure from central government to pay them over in the form of a dividend, whether they are realised or not[5].

Currency moves can also produce large unrealised losses, and even have the potential to leave a central bank with negative capital.  Both Banco Central de Chile and the Czech National Bank (CNB) have negative capital arising from exchange movements.  For a private sector company, such a position would be a cause of great concern.  Yet, as Jan Frait of CNB observes (Frait (2005)), their negative equity does not stop the central banks concerned operating effectively and – given their ability to generate future profits from seignorage – may not even give the banks net negative worth. This is a case where the application of standard accounting analysis to the two central banks’ accounts can be highly misleading, even to the extent of casting unjustified doubt over the effectiveness of the banks concerned.


It is the current “received wisdom” that central banks should aim to produce accounts that conform to the most rigorous international standards, and any voices raising caution and querying this are usually met with incomprehension.  But this paper has argued that it is far from clear that accounting systems and methods of analysis designed for private sector profit-making companies are optimal for central banks;  it is naïve to assume that the act of drawing up these accounts is neutral and does not affect either the central bank’s behaviour or its reputation;  and finally it is by no means self-evident that just because a central bank can draw up full financial accounts as if it was an ordinary private-sector bank, it is optimal for it to do so.


Frait, Jan:   “Exchange Rate Appreciation and Negative Central Bank Capital”, Remarks prepared for the Bank of England CCBS Expert Forum on Central Bank Finances, August 2005

Nugée, John:   “Issues in Accounting for Central Banks”, Accounting Standards for Central Banks, Central Banking Publications 2003

Sermon, Chris:   “Accountancy’s Golden Puzzle”, Central Banking Journal, August 2005


[1]              The Heisenberg Uncertainty Principle says that one cannot know the precise position and precise motion of a particle at the same time, as attempts to determine its position affect its motion and attempts to determine its motion affect its position.

[2]              The anomaly becomes clear if one considers how, for example, a nation’s health department might try to produce accounts of its policy operations.  How, for example, would they include in their accounts improvements in the physical and mental health of the country, which is after all one of the main objectives of the department.  Yet without valuing this, the health department – in crude financial terms – inevitably runs at a huge cost or “loss”.  Does this make it less worthwhile?  Likewise, a central bank’s accounts, which do not include a measure of the financial health of the country, are similarly an incomplete measure for the value of the central bank to the country.

[3]              As a example of this, see Sermon (2005), in which the author points out that standard accounting principles have no way of recording and accounting for central bank gold holdings.  This is because standard accounting principles are designed for standard companies, none of whom holds gold for the reason central banks do.  The point is that central banks are atypical, and forcing them to conform to accounting standards designed for other institutions can have unexpected and indeed counter-intuitive results.

[4]              Especially if the reserves are fully marked to market, as is increasingly seen as best practice.  This introduces the potential for very large volatitlity in the central bank’s balance sheet, which it could be argued is a high price to pay for little corresponding increase in clarity.

[5]              An example of the danger of producing seemingly precise accounts for FX reserves operations is the criticism levelled at the UK authorities for the – ultimately unsuccessful – defence of sterling’s parity in the ERM in 1992.  Rational analysis of whether the policy was sound and the defence justified, or at what point the authorities should have surrendered to the market, or even whether the UK’s membership of the ERM in the period 1990-92 was beneficial overall, are all coloured (and in some commentaries totally obscured) by the somewhat arbitrary figure of £3 billion or so that some analysts have put on the “cost” of this operation.

This essay was first published by State Street Global Advisors as part of their Insight programme