Open Market Operations and Financial Markets

David Mayes,Jan Toporowski

• Publisher: Routledge
• Year: 2007
• ISBN:
  9781134114344
• Edition: 1
• Topic: Business
• Subtopic: Business General
• Index: Business

A mixture of academic and practitioner research, this is the most detailed book available that provides an account of open market operations. With broad international appeal it includes discussions of central bank operations in Europe, North America, Australia and Japan. Exploring the effectiveness of short-term interest rates and other modern cent

1 Introduction

David Mayes and Jan Toporowski

The arrival of the ‘New Consensus’ as the guiding doctrine for monetary policy has coincided with a renewal of interest in the ways in which that monetary policy is implemented. Such a coincidence is not really surprising. It is obvious that the replacement of one guiding doctrine, laying out the effects of monetary policy on an economy, by another doctrine is not just decided by policy considerations, but also usually involves some re-examination of the way in which monetary policy is implemented. The practical operation of a guiding doctrine of the past is usually re-examined to show that not just administrative failures are responsible for the flaws in previous monetary policy. At the same time central bankers, operating in financial markets, need clear procedures for the implementation of the new policy. The last change of monetary regime, the switch to controls of monetary aggregates during the 1970s, was also anticipated by the critique of monetary operations from Milton Friedman and guidelines for the operation of new policy from William Poole (Friedman 1960; Poole 1970). The monetary procedures for the previous regime of active, Keynesian monetary policies after the collapse of the gold standard, and procedural errors in gold standard operations, had been clearly laid out by Hawtrey and Keynes himself (Hawtrey 1932; Keynes 1930/1971, 1945).

Similarly, the embrace by policy-makers of a ‘New Consensus in Monetary Policy’, the view that a central bank should set the short-term (overnight) rate of interest by regard to some target for future inflation, has also been associated with critiques of monetary policy procedures under the previous regime targeting monetary aggregates (e.g., Bindseil 2004b). Indeed, such discussion of their operating procedures has been invited by central bankers as a way of clarifying their obligations. For example, in a recent speech to Lombard Street Research, the Bank of England’s Executive Director for Markets, and member of the Bank’s Monetary Policy Committee, Paul Tucker urged further research in this direction: ‘The overall historical picture is not especially coherent. I suggest that the question of whether desirably or even optimally, there might be some mapping from monetary regimes to operating frameworks warrants research by the academic community’ (Tucker 2004, p. 372). Tucker refers to the Bank’s procedures as its ‘operating system’, an intriguing example of the influence of technology on the language of economics.

The operating target of New Consensus policy-making is the overnight rate of interest, as opposed to the money supply in the previous doctrine. The new system is a major and welcomed simplification in economic modelling, since the relationship between the interest rates that are the independent variables in models of the monetary transmission mechanism and the money supply, while elegant in theory, always proved troublesome in practice. Charles Goodhart has remarked in the past on the tendency of the money supply to elude control, and the Volcker experiment (1979–1982) in stabilising the monetary base also succeeded in destabilising the interest rates through which monetary policy was supposed to be transmitted to the rest of the economy. Since changes in the money supply were supposed, in any case, to operate through the rate of interest (the IS component of macroeconomic models, from which the Phillips Curve was derived), it makes sense where possible to control that rate of interest directly. This inevitably raises the question of how market interest rates can be influenced, and the role of open market operations in that system of control.

Central banks have relatively little direct control of interest rates. Operations in the money market, where overnight interest rates are set, require the co-operation of counter-party banks. In the case of the longer-term rates that are crucial for the monetary policy transmission mechanism, the influence of central banks is even more tenuous. Even the Bank of England’s Bank Rate under the gold standard, which is sometimes referred to by partisans of the ‘New Consensus’ as the golden age of interest rate targeting (e.g., Bindseil 2004b: 10–16; Tucker 2004, Appendix 3; Woodford 2003: 93–4), regularly lagged behind money market rates. Indeed, once it became clear that money market interest rates, rather than the amount of base money, were the targets of central bank monetary operations, the practical need to concentrate money market rates around the central bank’s preferred rate became a key factor in changing central bank operating procedures, both in the Euro-zone and in the U.K. The setting of an official discount or lending rate may of course have a significant ‘signalling’ effect in the money markets. But, without operations in the money markets, such signalling may have only a marginal impact on interest rates in those markets (Friedman 1999).

Central bank operations in the money markets may be conducted through open market operations, or through the use of standing facilities, sometimes also called the discount window. The previous monetarist, monetary policy regime undoubtedly favoured the use of open market operations. In part this was a legacy of the 1930s, when open market operations seemed to offer a direct way of counteracting a catastrophic credit contraction (Hawtrey 1932; Simons 1946). This preference for conducting monetary policy through open market operations was encouraged in recent central bank practice through the influence of Simons’s most prominent student, Milton Friedman. Even prior to the monetarist regime, open market operations were a favoured way of implementing policy. For example, in the early 1980s the Bank of England described its monetary operations as:

…setting, and periodic variation, of an official discount or lending rate, which, when necessary, is “made effective” by open market operations in the money market. “Making Bank rate effective” means restraining a decline in market rates from an unchanged Bank rate, or bringing them up to a newly established and higher Bank rate; it is accomplished by limiting the availability of cash to the banking system so as to “force the market into the Bank” to borrow at the somewhat penal rate of Bank rate.

(Coleby 1983, p. 213)

Under the monetarist regime, the conduct of monetary policy operations was supposed even to exclude standing facilities, or discount window operations. As an authoritative paper by Goodfriend and King on U.S. Federal Reserve policy argued ‘the discount window is unnecessary for monetary policy… Open market operations are sufficient for the execution of monetary policy. It follows that unsterilized discount window lending is redundant as a monetary policy tool’. This was followed by a cautionary note: ‘Nevertheless, over the years the Federal Reserve has employed unsterilized discount window lending extensively, together with discount rate adjustments, in the execution of monetary policy. Though it remains puzzling, use of the discount window this way seems to be connected with the use of secrecy or ambiguity in monetary policy’ (Goodfriend and King 1988; see also Schwartz 1992). In fact, the diversity of banks in the different regions of the Federal Reserve system has traditionally been a factor in the use of the discount window in the USA.

In a somewhat confessional (for a central banker) aside the Bank of England’s Executive Director for Markets admitted: ‘With no deposit facility… the OMO rate was a natural way to express policy and we slipped into thinking of it as how we actually implemented policy too. That was a fallacy’ (Tucker 2004).

The ‘New Consensus’ view of monetary policy has reversed the accepted view on the relative importance of open market operations and standing facilities. If standing facilities are available to participants in the money market, then the standing deposit and borrowing rates form a ‘corridor’ between which the market rate will fluctuate. How it will fluctuate depends on the amount of reserves that banks need on any one day; the amount and frequency of open market operations; and the credit activities of banks. For convenience the latter is sometimes modelled as a stochastic variable, e.g. in Davies (1998). If minimum reserves are required to be held at the end of every day, and that minimum is sufficiently large in relation to the daily fluctuation in credit activities, then, without accommodating open market operations, the overnight rate in the money market will tend to the upper and lower bounds of the corridor. One way of moderating this drift to the margins is to allow banks to average their reserve requirements over a maintenance period. In that case, the overnight rate will fluctuate between the deposit and lending rate, but will tend to end up on one of the corridor margins at the end of the maintenance period. The new arrangements for implementing monetary policy by the Bank of England envisage averaging with a wide corridor (100 basis points on either side of the official rate), to discourage use of standing facilities on a daily basis, but a narrower corridor (25 basis points on either side of the official rate) on the final day of the reserve maintenance period (Clews 2005, p. 211).

Thus, in the operational framework for the ‘New Consensus’ monetary policy, open market operations become redundant for the purpose of keeping the overnight interest rate close to the official interest rate. For example, the leading theoretician of the ‘new consensus’ Michael Woodford has argued that even with the zero reserve requirement that is implied by his assumption of a ‘pure credit’ economy, all that is required to keep the overnight money market rate at the official rate is for the central bank to offer a deposit facility at the official rate (Woodford 2003: 32–33). However, this is because the deposit facility he envisages would only provide a risk-free asset to the banking system, giving the money market a benchmark rate of interest on such assets. In the ‘pure credit’ economy that he envisages, all autonomous movements in banks’ currency would be accommodated in ‘complete markets’. Hence not only the absence of reserve requirements, but also the reduction of the banking system’s autonomous reserve requirements for payments purposes to zero, would eliminate the need for open market operations.

However, Ulrich Bindseil has recently raised another issue that has not been discussed in the academic literature, although it appears among the practical considerations that have been advanced in the establishment or reform of central bank operating procedures (e.g., Bank of England 2004a). This is the degree to which open market operations that deprive the banking system of reserves in order to induce the borrowing of reserves from the central bank thereby cause the central bank effectively to replace the activities of the money market (‘bringing the market into the bank’). His argument is that ‘open market operations should ensure that the recourse to standing facilities is not structural, but covers only non-anticipated probabilistic needs… Today, the essential argument advanced for open market operations is that they do not, in contrast to standing facilities offered at market rates, dry up the short-term inter-bank money market’ (Bindseil 2004b: 144 and 177). His concern is to minimise the tendency of commercial banks to draw routinely on standing facilities. Unchecked, this may turn the central bank into a giro-clearing system for the banks, as the German Reichsbank was before the First World War. In such giro-clearing all autonomous movements in currency and reserves end up as book-keeping transfers in the central bank’s balance sheet. The current view is that such routine drawing on standing facilities would require central banks to price the riskiness of lending to individual banks on a routine day-to-day basis, something that they would prefer the money market to do (Clews 2005). This is an aspect of central banks’ operations in money markets that has not been adequately discussed in the academic literature.

The reduced scope of open market operations is reflected in the reduction of the Bank of England’s operations from two or three each day, to one each week, plus another operation on the last day of each maintenance period, although additional open market operations will be undertaken to prevent a build-up of reserves that would render the banking system independent of the central bank’s official rate (Clews 2005). In the ‘New Consensus’, in which monetary aggregates are no longer supposed to matter, but monetary policy is conducted by movements in the official rate of interest, the new function of open market operations is not a monetary one, in the sense that the scale of these operations is unrelated to the rate of interest that the central bank seeks to enforce in the money markets, or to the monetary policy stance that the central bank is adopting, i.e. the trend in interest rates that the central bank seeks to indicate to the financial markets. The function of open market operations in the new consensus is to prevent settlement banks from ‘forcing the money markets into the bank’ by using remunerated standing facilities as a form of cash management service. Monetary ‘shocks’ are now supposed to be modelled as changes in interest rates, possibly in exchange rates, rather than as unexpected increases or decreases in the money supply, that may be offset by open market operations. Similarly, the monetary transmission mechanism is activated by changes in interest rates, rather than injections of money through open market operations.

The changed scope and significance of open market operations in the New Consensus monetary policy therefore raise important questions of theory, policy and modelling. In 2004, the Bank of Finland, together with SUERF, took the initiative of calling a conference to discuss these questions. The conference took place in Helsinki in September 2005. We were fortunate in being able to secure the participation of a wide range of experts from central banks, the academic milieu, and commercial banking and finance. The papers in this volume therefore represent a selection of those papers, enlarged and improved by the discussions at the conference.

The structure and argument of the rest of the book

The twelve chapters that follow fall into three groups. The first, containing chapters by David Laidler, Bill Allen, Ulrich Bindseil and Flemming Würtz, Jens Forssbæck and Lars Oxelheim, Laurent Clerc and Maud Thuaudet, and Noemi Levy Orlik and Jan Toporowski, lays out the ingredients of the ‘New Consensus’ and what it implies. The second group, with papers by Dan Thornton, Beata Bierut and Michal Kempa, looks much more closely at what the new consensus implies for the behaviour of monetary policy implementation and in particular explores its limits. The remaining group considers wider questions, with Sheila Dow, Matthias Klaes and Alberto Montagnoli exploring the nature of signalling in implementing monetary policy, Xinsheng Lu and Francis In the impact of OMOs on financial markets and, finally, Takero Doi, Toshihio Ihori and Kiyoshi Mitsui the interaction with fiscal policy and public debt in face of the threat of deflation.

While there is a larger share of papers on the Eurosystem, the chapters explicitly cover the situation in the United States, Japan, the UK, smaller EU countries, Australia and emerging markets, with a particular focus on Mexico, so as to cover a wide spectrum of experience and regimes.

The flavour of the early chapters is to look beyond the ‘new consensus’ into what open market operations could be in different circumstances and into where thinking on the subject appears to be going. They form a group, each contributing to the overall understanding. The initial purpose is to set out what the new consensus constitutes and build on it from there. Put crudely, what we see is a shift from a focus on quantities in the money market directly influenced by the central bank to a focus on a rate of interest. The essence of the system is in effect a simple three-equation model. Aggregate demand in the economy is affected, inter alia, by the rate of interest. Inflation is a function of expectations, some specific factors and the gap between aggregate demand and some measure of sustainable supply – the gap being labelled the output gap. Central banks seek to control inflation by setting interest rates in such a way that future inflation is likely to remain within acceptable levels.

David Laidler’s chapter, which follows, seeks to embed these views in the literature of monetary economics. In many ways his chapter offers a critique of Michael Woodford’s (2003) book, ‘Interest and Prices: Foundations of a Theory of Monetary Policy’, that has done so much in developing a consensus model that has a proper foundation in economic theory. However, he provides a fascinating account of how thinking in monetary economics has developed over the last 75 years or so and how that development has interacted with the monetary policies that central banks have sought to implement.

The key message in his analysis is that, attractive and elegant though the Woodford framework is, it is lacking in some of the core elements necessary to provide a helpful basis for policy making. His ‘cashless’ economy removes the role of money as a means of exchange and assumes away the problems of what will happen if markets do not clear. Traditional theory at least offers a buffer stock role for money in enabling people to correct for all the various errors they make in pricing and in interpreting information.

The interest rate route also appears to offer a problem when the lower bound of a zero nominal rate is reached in the face of deflation. Here, traditional theory suggests using open market operations to flood the market till the point that the economy does turn round – a point that Nakaso returns to in discussing the very last chapter in the book on the case of the quantitative easing in Japan. Laidler remarks with some irony that perhaps ‘the seemingly serious limits imposed on the powers of orthodox monetary policy by the nominal interest rate’s zero lower bound is not so much a property of the real world as of monetary policy models that focus too exclusively on interest rates’. A monetary policy that simply focuses on the new consensus and does not bear regard for the function of monetary quantities would tend to miss out on the times of difficulty when ‘open market operations should be promoted from the technical fringes of monetary policy to its very centre’.

Allen turns the focus on its head by arguing that the technical facets of modern open market operations can have clear macroeconomic and microeconomic consequences that need to be explicitly addressed. Hence it is important to step back from technical objectives, such as achieving a particular short rate of interest and consider whether the point of such actions might, for example, be better achieved by having a little flexibility. Without price signals, banks may pay only limited attention to cash-flow management and spreads may be narrowed, with consequences for macroeconomic management. It seems odd, for example, that some central banks provide free intraday credit, yet credit has a price in the overnight market for a similar fraction of a day. Similarly, allowing commercial banks to average their reserve holdings at the central bank over a predetermined period to achieve a given minimum requirement (subject to never running into overdraft at the end of any day) in order to keep short run interest rates smooth is likely, in Allen’s view, to limit market activity and accord unequal competitive advantage to the banks. While the market needs to be orderly, great smoothness in short interest rates has no particular macroeconomic implications and may allow inefficiencies, such as a bank with problems being able to get through to end of the (maintenance) period before the problem becomes apparent.

The two principal avenues of open market operations through the central bank buying or selling short-term securities and through overnight collateralised lending and borrowing facilities form the heart of much of the debate in the rest of the book. While central banks may be able to manipulate short rates it depends very much how much the market agrees with their assessment as to how much those rates will be transmitted along the yield curve and affect real activity and inflation. Allen’s work thus provides a basis on which many of the subsequent arguments are built.

Bindseil and Würtz, whose chapter immediately follows, is a case in point. It considers the relative roles of the two forms of operation – open market operations and standing facilities in achieving the desired short-term interest rates and keeping them stable. It begins with a robust critique of the Reserve Position Doctrine in the Federal Reserve over period from the 1920s to the 1990s. Gavin, in his comment, focuses on this aspect of the chapter, which is not surprising as he works for the Federal Reserve System. The authors come out in favour of the standing facility end of the spectrum, preferring stability in interest rates over a more ‘vivid’ interbank market, as they put it. The flavour of their remarks is thus in contrast to Allen, who puts a stronger weight on the importance of a market to ensure the removal of distortions – but a market in which the commercial banks are not particularly priviledged. As Gavin also remarks in his comment, exposing banks to the interbank market means that there will be a greater element of discipline in their risk-taking than would be the case from automatic access to the central bank’s stand...

Table of contents

• Cover Page
• Title Page
• Copyright Page
• Contributors
• Preface
• 1. Introduction
• 2. Monetary Policy and Its Theoretical Foundations
• Comment On Laidler
• 3. The Scope and Significance of Open-Market Operations
• Monetary Policy and the Allocation of Resources: Comment On Allen
• 4. Open Market Operations – Their Role and Specification Today
• Comment On Bindseil and Würtz
• 5. Monetary Policy In a Changing Financial Environment: A Case for the Signalling Function of Central Banks’ Operating Framework
• Comments On Clerc and Thuaudet
• 6. The Interplay Between Money Market Development and Changes In Monetary Policy Operations In Small European Countries, 1980–2000
• Discussion of Forssbæck and Oxelheim
• 7. Open Market Operations In Emerging Markets: The Mexican Experience
• 8. Open Market Operations and the Federal Funds Rate
• Comment On Thornton
• 9. On the Optimal Frequency of the Central Bank’s Operations In the Reserve Market
• Comment
• 10. Money Market Volatility – A Simulation Study
• Discussion of Kempa – Complexity and Challenges of Modelling the Central Bank Operational Policies
• 11. Monetary Policy By Signal
• 12. The Impact of the Reserve Bank’s Open Market Operations On Australian Financial Futures Markets
• 13. Sustainability, Inflation and Public Debt Policy In Japan
• Discussion of Doi, Ihori and Mitsui – Bank of Japan’s Open Market Operations Under the Quantitative Easing Policy
• References