Unconventional monetary policy: introduction.

Armstrong, Angus ; Ebell, Monique

The world's four major central banks have turned to new forms of monetary policy to support demand during the Global Financial Crisis. (1) The conventional policy instrument of overnight interest rates was reduced to close to zero per cent within eighteen months of the crisis. (2) This presents a problem of providing further stimulus by lowering interest rates; if rates turn negative then depositors always have the option of holding wealth in cash at a zero interest rate. The option of holding cash is thought to create an effective floor on interest rates known as the zero lower bound (ZLB). Central banks have had to find new ways of deploying monetary policy, popularly known as 'unconventional' monetary policy, to provide further stimulus subject to the ZLB.

Central banks have also carried out many other interventions, from buying private or risky sovereign securities (accepting credit risk), to direct liquidity and solvency support to institutions and markets and implementing government guarantees to support lending. These interventions directly involve credit risk and so can be considered as distinct from monetary policy. The Bank of England (the Bank) and the European Central Bank (ECB) have been particularly keen to maintain this distinction. (3) Yet the validity of the distinction is unclear; monetary policy influences the risk premia on assets with consequences for financial stability; financial stability policies influence the transmission mechanism for monetary policy; and income from unconventional monetary policy is counted as a fiscal revenue.

In this issue of the Review, we accept this distinction of unconventional monetary policies from other measures taken by central banks. In particular, three types of unconventional policies have been widely used by the major central banks. First, forward guidance was introduced by the US Federal Reserve (the Fed) in December 2008 and has since been used by all four major central banks. (4) Second, all four central banks have engaged in large asset purchase programmes through the creation of bank reserves, known as quantitative easing (QE). Finally, there is a healthy debate about whether nominal interest rates can, in fact, safely be set substantively below zero per cent. Four European central banks have now introduced negative policy rates without any obvious sign of instability.

When assessing the suitability of unconventional monetary policies we face the quandary that mainstream economic models (e.g. New Keynesian) offer little rationale for intervening. According to these models the current interest rate and expected path of future interest rates fully encapsulates the impact of monetary policy on the economy. Unless unconventional monetary can change expectations about the future path of policy then there is no channel to influence the economy. Bernanke (2014) remarked that that QE "works in practice, but does not work in theory". (5) However, these strong results are generated in models that generally have no meaningful financial sector and assume that agents such as central banks can fully commit to policy targets. The financial sector continues to be underplayed in economic models and in practice. Its absence may lead to misleading assessments of the effectiveness of unconventional monetary policies.

An 'unconventional' history

While theory may not be much of a guide for unconventional policies, there is a rich history of intellectual debate on the appropriateness of such measures. During the Great Depression all three unconventional measures were widely debated. Indeed, Keynes (1936) set out a general economic system where the interest rate could not fall to the level required to support investment and output at full employment. An important part of Keynes's system became known as the 'liquidity trap'. (6) Keynes noted that "there is a possibility ... that, after the rate of interest has fallen to a certain level, liquidity-preference may become virtually absolute in the sense that almost everyone prefers cash to holding a debt which yields so low a rate of interest. In this event the monetary authority would have lost effective control over the rate of interest." (7)

Keynes considered and discussed two of the three forms of our modern day 'unconventional' monetary policies: forward guidance and quantitative easing. He recognised that monetary policy operates through short-term interest rates and could not necessarily influence the long-term interest rate which contains future expected interest rates and therefore is also important for investment. Indeed, he appreciated the time inconsistency problem facing monetary policymakers: "a monetary policy which strikes the public as being experimental in character or easy to change may fail in its objective of greatly reducing the long-term rate of interest". (8) Of course, "the same policy might prove easily successful ... if it is ... rooted in strong conviction, and promoted by an authority unlikely to be superseded". (9) Thus, Keynes effectively provides an early description of forward guidance.

The historical roots of QE have similarly close ties to Keynes. There is considerable dispute about how exactly QE functions, with the Bank of England offering many possible answers. Perhaps the clearest explanation is given in Governor King's speech in early 2009 soon after the policy was announced: "the success is not to be judged by the increase in bank lending. Its aim is to increase the money supply." (10) In an open letter to President Roosevelt (1933), Keynes described the idea that output and income can be raised by increasing the quantity of money as "like trying to get fat by buying a larger belt". Keynes described the quantity of money as only a limiting factor and the volume of expenditure is the operative factor.

The idea of engineering substantially negative interest rates also has a long history. Silvio Gesell (a German merchant living in Argentina) suggested that currency notes could be required to be stamped to maintain their legal tender status. The cost of the stamp would be a cost of holding the currency and set to be equivalent to a negative interest rate. Keynes (1936) appears to have been enthusiastic, dedicating much of chapter 23 to this idea, but remained unconvinced as the analysis was partial equilibrium. Another possibility for generating a negative interest rate was suggested by Robert Eisler (1932): when an exchange rate is set between bank reserves and cash (rather than the presumed parity rate). Since the interest rate on reserves can be set at any rate, the zero interest rate on cash could be made negative by depreciating the exchange rate of cash against reserves. This idea is developed in the first paper in this special issue.

Lender risk

There are of course similarities between the Great Depression and the global financial crisis. Both were preceded by the accumulation of private sector debts, rising asset prices and increased leverage in the financial system. (11) The relevance is that the fall in demand and resultant debt overhang raised doubts about the effectiveness of monetary policy. However, the limiting factor may not be the ZLB on short-term policy rates or sovereign bonds but the lending rate for enterprises. Keynes (1936) noted that two types of risk affect the volume of investment--borrower risk (associated with the riskiness of the project) and lender risk which is a "pure addition to the cost of investment and would not exist if the borrower and lender were the same person". (12) The exclusion of a meaningful financial sector in many economic models may downplay the cost of intermediation above the zero rate as the effective lower bound. While large firms can access capital markets, the vast majority of firms and around half of output is generated by small and medium enterprises which rely on the banking sector. The lower bound for these firms' borrowing costs is not zero, but the cost of intermediation. Two points follow. As long as many institutions continue to trade at market values below book values it is difficult for the cost of intermediation to fall further. Also, how the banking sector responds to unconventional monetary policies may determine their effectiveness.

Electronic money

The first paper in this Review, by Miles Kimball, takes up the challenge of how to deliver significant negative interest rates when necessary for economic stabilisation. The objective is to remove the ZLB with the minimum of social and political complications so that central banks could revert to negative interest rates within weeks if warranted. Kimball takes up Eisler's idea of introducing an exchange rate between bank money and paper (currency) money. Most of our transactions today are carried out using electronic money (bank transfers, cheques, credit and debit cards and touch cards) with currency only used for small transactions. Kimball suggests it would be straightforward to impose negative interest rates on electronic money. In such periods cash would no longer have a par exchange rate against electronic money but would depreciate very gradually over time to yield the same negative interest rate.

Items in shops would be priced in terms of electronic money, at least for larger items usually paid for by cards or transfers, which are also more interest rate sensitive. The cash exchange rate would depend on the depreciation rate set by the central bank and the time period since negative rates began. For example, a negative 4 per cent interest rate would be equivalent to a 0.011 per cent of a cent decline in the value of a dollar each day over one year. This would minimise disruption to cash balances for transactions purposes and discourage flight from electronic money stocks into currency. If negative interest rates are necessary (ruling out other options such as fiscal policy) then this idea of an exchange rate between electronic and paper currency is surely the most viable.

While Kimball notes that economic theory makes little distinction between positive and negative interest rates, we would like to know more about how profit maximising financial institutions might respond. An environment where significantly negative interest rates are required is likely to be one with great uncertainty and hoarding of liquid assets. Banks may be less willing to pass negative interest rates on to their depositors for fear of flight, perhaps into another currency or financial asset. As a result banks would receive a negative return on their reserves at the central bank. This would increase the financial pressure on banks and may even lead them to increase lending rates to maintain their profit margin. The effect of the negative interest rate policy on the economy may be muted or even counterproductive depending on the behaviour of financial institutions.

Quantitative easing and inflation

The second paper examines quantitative easing within the context of the quantity theory of money, in particular whether the accumulation of bank reserves will lead to an eventual rise in inflation. Bill Cline provides an expert account of why one might expect vast increases in bank reserves to lead to growth in the broad money supply and inflation. The reason that this has not occurred is the decline in the money multiplier (the transformation of reserves into broad money) rather than a particular fall in the velocity of circulation. The money multiplier is a function of cash balances, bank profitability and the amount of leverage banks are seeking. If financial institutions and households and firms are seeking to de-leverage this is consistent with a falling money multiplier.

It would be incautious to consider that because inflation has not arisen thus far central banks are in the clear. As the economy and financial institutions normalise, banks may be prepared to increase lending faster given their large reserve holdings. Cline summarises the Fed's exit strategy of increasing the interest rate on reserves. Higher reserve requirements are limited by law. Therefore the real risk is that there could be a dramatic switch from 'risk-off' to 'risk-on' behaviour whereby banks are not discouraged by the higher returns on reserves. Again this depends very much on the response of financial intermediaries.

Channels of quantitative easing

Jagjit Chadha addresses some of the limits of the mainstream approach to monetary policy. In his view, QE is simply an extension of conventional monetary policy, i.e. an extended open market operation involving 'unsterilised' bank reserves, but applied to the ZLB. However, if such operations influence the risk premia on the assets bought, then this may be an important channel by which QE influences the economy. Chadha opens the discussion on financial intermediation and differences in borrowing and lending rates, showing how this can lead to distributional consequences as well as alter the impact of QE. In the policy space this draws him to discuss how QE might complement macro-prudential measures. This inevitably brings unconventional monetary policy into the realm of actions which have fiscal consequences. The paper concludes with a call for more consideration of how the array of tools now available to central banks can be used in a complementary fashion.

Jack Meaning and James Warren provide a quantitative assessment of the extent and channels by which QE influences longer-term interest rates. They consider how the Bank of England's acquisition of 30 per cent of the stock of gilts in private hands may have influenced bond prices in two ways: through the New Keynesian channel of changing expectations of future monetary policy; and through market segmentation theories where bonds are imperfect substitutes. They show that QE may have lowered bond yields by 25 basis points; at the lower end of economists' estimates. This may be consistent with a greater shift in short-term interest rates. The authors suggest that when QE is unwound the increase in supply may raise interest rates over and above any expected changes in monetary policy. This also highlights the inevitable interaction between fiscal and monetary policy.

An important issue is the impact of QE beyond the direct affect on sovereign bond yields. The impact on financial intermediaries may in fact be as important for the real economy. A recent paper by Lambert and Ueda (2014) provides an assessment of QE in the US on banks' profitability. They reject the common assumption that QE is beneficial for banks. Rather they provide evidence that the flattening yield curve may damage bank profitability and even encourage greater risk taking. This may accord with the increased concentration and size of banks since the crisis. It is the impact of QE on financial intermediation together with the sovereign yield curve which will determine the overall effectiveness of the policy on the real economy.

The papers in this issue of the Review provide leading assessments of the viability and impact of unconventional monetary policies. We are extremely grateful to the authors. Our judgement is that incorporating profit maximising financial intermediaries is required both in understanding economic cycles and for a full assessment of the desirability of unconventional monetary policies.

NOTES

(1) The four major central banks in this introduction refer to the Federal Reserve, European Central Bank, Bank of Japan and Bank of England.

(2) Central banks differed in their urgency to deploy conventional monetary policy. The Fed reduced its target interest rate by 100 basis points in the twelve months before the economy entered recession in the first quarter of 2008. The Bank of England raised interest rates in July 2007, two months before the demise of Northern Rock, and the Bank Rate was 5 per cent in October 2008--after two quarters of economic contraction. The ECB began raising interest rates in 201 I just ahead of the full consequences of its sovereign debt crisis.

(3) The ECB defines large asset purchases where credit risk remains within national boundaries as monetary policy (Public Sector Purchase Programme) while the purchase of distressed sovereign assets where credit risk is shared (either the Securities Markets Programme or Outright Market Transactions) is defined as improving the transmission mechanism rather than traditional monetary policy.

(4) The Federal Reserve minutes (16 December, 2008) stated that "weak economic conditions are likely to warrant exceptionally low levels of the federal funds rate for some time".

(5) Bernanke (2014), p. 14.

(6) The term 'liquidity trap' is generally credited to Robertson (1940).

(7) Keynes (1936), p. 207.

(8) Keynes (1936), p. 203.

(9) Keynes (1936), p. 203.

(10) King (2009), p. 3.

(11) There are also many important differences between the Great Depression and the global financial crisis; in particular, the policy response of supporting banks allowing fiscal deficits to absorb the fall in private demand. The UK equivalent of the great depression happened in the 1920s and no financial institutions were lost.

(12) Keynes (1936), p. 144.

REFERENCES

Bernanke, B. (2014), 'Central banking after the Great Recession: lessons learned and challenges ahead', discussion at The Brookings Institute, 16 January.

Keynes, J.M. (1933), 'An open letter to President Roosevelt', New York Times.

--(1936), The General Theory of Employment, Interest and Money, The Royal Economic Society, 1973.

King, M. (2009), Mansion House Speech, Bank of England, 17 June.

Lambert, F. and Ueda. K. (2014), 'The effects of unconventional monetary policies on bank soundness', IMF Working Paper, WP/I4/I52.

Robertson, D.H. (1940), Essays in Monetary Theory, P. S. King.

Angus Armstrong and Monique Ebell *

* National Institute of Economic and Social Research and Centre for Macroeconomics. E-mail: a.armstrong@niesr.ac.uk or m.ebell@niesr.ac.uk. The authors are supported by the Economic and Social Research Council through its Centre on Constitutional Change.