… Peter Stella’s contribution to understanding issues related to central banks’ negative capital
« I am aware that some people do not fully understand, from an economic point of view, why a central bank should be concerned about the soundness of its capital, » Toshihiko Fukui, Governor of the Bank of Japan, June1, 2003 (link)
« Part of the inability of some to understand the relevance of the financial situation of central banks is clearly linked to the situation in the countries where most economics textbooks were written » (countries where central banks have not experienced such problems), Stella and Lönnberg (2008)
« We, the Central Bank of Costa Rica, have a problem of negative economic value, and that is our biggest problem, » Francisco de Paula Gutierrez, Governor of the Central Bank of Costa Rica, 2004
Usefulness of the article: The article summarizes important points that help to better understand issues related to negative capital in central banks.
Summary:
-
We review the reasons for the emergence of the importance attached to the capital (and more broadly the financial situation) of central banks in the 1980s and 1990s, linked to the trend towards central bank independence and transparency.
-
We discuss the concept of a central bank’s « economic value, » which is important in this debate, and provide links to comprehensive articles that provide a full understanding of the related concepts.
-
We focus on the link between the financial situation of the central bank and inflation (thus ignoring the aspects of credibility, reputation, and political economy associated with the issue of negative capital), and discuss the empirical evidence on this aspect.
The issue of central bank capital has been the focus of attention in recent months. It is linked to what some refer to as a « debate »: the question of « canceling the public debt held by the European Central Bank. » While the « debate » on « the cancellation of public debt by the Central Bank » is more media-driven than academic these days, the issue of central bank capital has been the subject of much academic discussion, with more than a hundred academic articles on the subject to date.
The article often considered to be the « founding » work in this field is that of Peter Stella, former head of the IMF’s Central Banking Department, « Do central banks need capital? », published in 1997. It was followed by a dozen articles by the same author on this topic, including two frequently cited articles published a few years later, « Central Bank Financial Strength, Transparency, and Policy Credibility » and « Issues in Central Bank Finance and Independence » (co-authored with Aka Lönnberg).
In these articles, Peter Stella and his co-author lay the groundwork for this debate, drawing on a number of experiences they encountered during their work at the International Monetary Fund (IMF): what is the purpose of a central bank’s capital? To what extent do central banks need capital? What do the facts tell us? Why have some central banks suffered from negative capital, while others have not? Does the fact that some central banks with negative capital have not appeared to have any problems imply that capital is unnecessary? At what point should we start to worry about a central bank with negative capital?
1. Reasons for the emergence of this debate in the 1990s
Stella and Lönnberg (2008) begin their study by explaining why the debate on central bank finances is a relatively « recent » one. One of the main reasons cited is the trend toward central bank independence, which often includes a financial component, at least in theory.
Previously, the boundaries between central banks and the state were very blurred. There was therefore less reason to analyze the central bank (which is, fundamentally, only a « branch » of the state) as a separate entity. To put it simply: the financial problems of the state were also those of the central bank, and the financial problems of the central bank were also those of the state[1]. In the absence of this boundary, the currency sometimes issued by central banks to cover the state’s financing needs was well understood (by both Keynes and Friedman)[2] as a « government obligation » (to quote Friedman), in the same way as bond loans, and therefore had to be « backed, » like bond loans, by future primary surpluses in order to maintain a credible value.
The trend towards central bank independence meant that a central bank with financial problems could more easily find itself in a position to resolve them « alone. » The financial problems relevant to Stella’s analysis are mainly (but not exclusively) those of recurring financial losses (negative cash flows), which are therefore indirectly linked to the level of capital.[3] Continuously negative profits, for example, as observed in some Latin American central banks in the 1980s and 1990s (see Klüh and Stella (2008)), would require a continuous need for money creation in order to honor commitments. The central bank’s balance sheet (excluding capital) would theoretically continue to grow indefinitely and beyond, in the absence of any adjustment.
In the absence of a link with the state and therefore of recapitalization, there are several ways for a central bank to resolve[4] financial problems « on its own. » The most significant are discussed in Stella and Lönnberg (2007). The main way is to leave interest rates lower for a significant period of time, and thus allow inflation to run higher than initially desired[5]. Stella and Lönnberg (2007) refer to « policy insolvency « to describe the resulting situation, which could be defined more broadly as follows: a central bank will be in a situation of « policy insolvency » when its financial situation prevents it from achieving its economic policy objectives, in particular the operational target to which it has committed (e.g., its 2% inflation target or its exchange rate target in the context of a peg).
2. Negative economic value versus negative capital
Stella explains that the threshold for « political insolvency » is not, in theory, the same as the central bank’s book capital. A central bank may well have positive capital but be constrained in its economic policy objectives due to its financial situation. Conversely, a central bank may have negative capital without being in a situation of « political insolvency. » How, then, can this insolvency threshold be quantified, and how can this concept be defined and better understood?
With a few nuances, we can simplify by considering that Stella largely equates this political insolvency threshold with the economic value[6] of a central bank, conceptually « the value that a fully informed and risk-neutral shareholder would be willing to pay if he were to buy a central bank » ( Stella, 1997). This concept, sometimes misunderstood by some but well understood by major investors (such as Goldman Sachs, 2015), was analyzed in a more mathematical form a few years later by Ricardo Reis and Christopher Sims (Reis, 2013; Del Negro and Sims, 2015) and used in other applied studies (Buiter, 2008; Pinter and Pourroy, 2019).
As Stella (1997) explains, economic value differs from book value: the book value of a central bank’s capital does not reflect future profits, for example (whereas economic value does). A central bank may therefore well have negative book value but positive economic value. In this case, we could consider that the « political insolvency » mentioned above should not be relevant: the central bank will normally be able to achieve its operational target without being constrained by its financial situation. Low capital could still be problematic for the conduct of monetary policy, but not via the channel typically at work in the case of negative economic value (see Klüh and Stella (2008) and Pinter (2018) as well as related articles, and for a concrete example, see the recent Goldman Sachs note of July 2, 2020, expressing concern about the situation related to « unbacked » monetization in Indonesia[7]). When, on the other hand, capital is so negative that economic value becomes negative, the situation can clearly pose a problem if it is certain that the state will not intervene.
3. The link between financial situation and inflation in practice
Stella (1997), Stella (2003) and Stella and Lönnberg (2007) cite numerous cases where a central bank’s financial situation has caused it real problems in achieving its objective. This is particularly true in emerging or developing countries, which can probably be explained in part by a lack of strong institutions to prevent such problems.
The best-known case is probably that of Costa Rica, whose governor stated in 2004 that « we, the Central Bank of Costa Rica, have a problem of negative economic value, and that is our biggest problem. » The central bank had indeed suffered numerous losses, resulting in negative capital of nearly 8% of GDP at the end of 2003, despite two episodes of recapitalization by the state in 1999 and 2001, amounting to nearly 10% of GDP (see Ize (2005)). As an IMF report explains (IMF, 2019), this financial situation contributed to major problems for the central bank, affecting inflation and the exchange rate. The « adjustable peg » exchange rate regime followed by the central bank saw the Costa Rican currency depreciate by nearly 500% between 1990 and 2004. Inflation averaged 20% between 1982 and 2006, confidence in the currency collapsed, and dollarization increased.
In other cases, the seemingly « fragile » financial situation of a central bank was not recognized as a direct source of problems for controlling inflation. This is particularly true in Chile, as discussed at length in Stella (2003) and Ize (2005). The relatively negative capital from 1998 onwards does not seem to have constrained the central bank in achieving its inflation targets, although the governor at the time confided at a public seminar at Harvard in 2019 that the negative capital level had caused him real problems in other areas. A recapitalization plan by the Chilean government was initiated in 2006 (Stella and Lönnberg, 2007), and studies by the Central Bank of Chile in 2009 pointed to the need for further recapitalization (Central Bank of Chile, 2009).
Beyond the qualitative studies mentioned above, some researchers have attempted to empirically analyze a possible causal link between the financial situation of central banks and economically important variables such as inflation (which is a particularly difficult task). Klüh and Stella (2007) conclude that there is a causal link in cases where the financial situation is severely deteriorated, and Pinter (2018), analyzing 82 constitutional laws on central banks to select only those countries where recapitalization is clearly not guaranteed, concludes that there is a causal link only in cases where the state « clearly does not have the means to recapitalize . » These empirical studies remain imperfect and do not provide any indicators of « economic value, » for example, only proxies. At this stage, the most informative discussions on the cause-and-effect relationship (apart from cases of hyperinflation) are certainly to be found in Stella’s papers discussed here, as well as in IMF and central bank reports based on historical archives or direct discussions with operational staff.
4. Conclusion
Issues relating to central bank finances have been the subject of much debate in recent years. Peter Stella is undoubtedly one of the pioneers in the analysis of these issues. His articles provide a rich basis for reflection, inspired and illustrated by numerous concrete examples to which the author has been exposed through his experience as a « field » economist at the IMF.
A very clear case where the financial situation of a financially independent central bank can make it difficult for it to control inflation, for example, is when the economic value of the central bank is negative. Economic value is a different concept from book capital and relates to the structural profitability of the central bank.
In exploring these debates, Stella and his co-authors go well beyond the « tempting » but simplistic idea that the power to create money makes the issue of central bank finances irrelevant, which Stella (2005) considers to be a fallacy.
Peter Stella’s articles, by clarifying concepts and bringing together numerous anecdotes from an economist directly exposed to these issues, are of undeniable value to anyone interested in these debates.
NB: this article is simply intended to shed light on the issue of negative capital in central banks. The fact that a central bank can have negative capital and positive economic value cannot be seen as an argument for monetary financing without collateral being a preferable method of financing to bond financing, see Stella and Pinter (2020), among others.
Bibliography:
Adler et al (2012) « Does central bank capital matter for monetary policy? », Open Economies Review
Central Bank of Chile (2009) « Macroeconomy, Monetary Policy and Central Bank Capitalization, » Economia Chilena 12(1):5-38
Buiter, Willem H., 2008. « Can Central Banks Go Broke? », CEPR Discussion Papers 6827, C.E.P.R. Discussion Papers.
Del Negro, Marco & Sims, Christopher A., 2015. « When does a central bank’s balance sheet require fiscal support? », Journal of Monetary Economics, Elsevier, vol. 73(C), pages 1-19.
International Monetary Fund (2019) “Foreign exchange intervention in inflation targeters in Latin America, chapter 10 Costa Rica: learning to float”
Goldman Sachs, 2015 « European Economics Daily: Accounting capital vs. Economic capital – the case of the SNB, » Goldman Sachs Macro Research, August 10
Goldman Sachs, 2020 « Bank of Indonesia buying zero interest bonds: needed stimulus or slippery slope? », Goldman Sachs Macro Research, July 2
Ize A (2005) Capitalizing central banks: a net worth approach IMF Staff Papers, vol 58 International Monetary Fund, Washington, DC, pp. 289–310
Kluh U, Stella P (2008) Central Bank Financial Strength and Policy Performance: an Econometric Evaluation, International Monetary Fund, Working Paper 08/17, Washington.
Julien Pinter, 2018. « Does Central Bank Financial Strength Really Matter for Inflation? The Key Role of Fiscal Support » Open Economies Review, Springer, vol. 29(5), pages 911-952, November.
Pinter Julien and Pourroy Marc « How can financial constraints force a central bank to exit a currency peg? An application to the Swiss franc peg, » SSRN Working Paper.
Ricardo Reis, 2013. « The Mystique Surrounding the Central Bank’s Balance Sheet, Applied to the European Crisis » American Economic Review, American Economic Association, vol. 103(3), pages 135-140, May.
Stella, Peter (2003) « Central Bank Financial Strength, Transparency, and Policy Credibility, » International Monetary Fund, Working Paper, Washington.
[1]Transparency was not the same as it is today either. It should be remembered that it was only in the 1990s that the Fed, for example, began to explain its decisions to the public; see this article onthe subject. The trend toward transparency is another key reason cited by Stella and Lönnberg (2008) and Stella (2005).
[2]See Keynes’ analogy in « A Tract on Monetary Reforms » (p. 62). Friedman , speaking in India , said, « There is no difference between a promise to pay in the form of a two-rupee note and a promise to pay in the form of a bond except that the bond pays interest and the two-rupee note does not. The notes outstanding are non-interest bearing obligations of the government. « On the illusion of a crucial difference between central bank financing and bond financing, see in particular the work of Sims, Sargent, Wallace, Cochrane, and Leeper (often considered rigorous on this topic).
[3]The capital of a central bank is linked to its ability to generate positive structural profitability. This can be understood with the following example: imagine that the capital of the central bank increases by 10 billion, which means that the central bank’s assets have increased by 10 billion, while its liabilities (excluding capital) have not changed. The interest rate on the additional $10 billion in assets allows the central bank to generate more income in the long term.
[4]Note that « resolve » is used here in the sense of « ensuring that structural profits become non-negative again. » The fact that negative structural profits can be a problem for an independent central bank is the very subject of all the literature related to this article.
[5]This should lead to a significant increase in seigniorage profits in the long term. See Reis (2013), Ize (2005), Adler et al (2012), Pinter and Pourroy (2019) and the papers cited by Ize for more comprehensive discussions.
[6]Or « net value, » in English the term is « net worth. » Buiter (2008) refers to « comprehensive net worth » or « comprehensive equity » to describe a concept that is essentially similar. A discussion of this terminology can be found in Stella (2003). In Stella, the concept of « political insolvency » is more general than that of « central bank insolvency » proposed by Reis (2013) or Del Negro and Sims (2015).
[7]In this note, economists point to the risks of potentially uncompensated zero-rate monetization in Indonesia, implying « a broad decline in central bank profits for the future, » potentially constraining it in possible future sterilization operations.
[8]The paralogism cited by Stella (2005) is as follows: « 1. Commercial banks require financial strength (capital) to absorb losses while meeting their financial obligations in full and on time and hence remain in operation. 2. Central banks have an unlimited costless ability to create the means to pay their financial obligations in full and on time in domestic fiat money. 3. Central banks, therefore, do not require financial strength (capital). »See also Stella and Lönnberg (2007) : « This neglect of central bank corporate finance may be attributed to several factors which have led to the view that central bank finances can be ignored as they are either: macroeconomically insignificant; properly analyzed only within the consolidated public sector accounts; or irrelevant owing to the central bank’s unlimited ability to create money. Each of these factors will be considered in turn in this paper … ».
