The following note is work-in-progress.
It is part of an attempt to bring to an end the great mass of confusions that have cluttered discussion of ‘quantitative easing’ in the UK’s monetary policy debate and indeed the monetary policy debates of many nations, in the last few years.
Work on the note will resume next week…and perhaps in the following weeks.
How can the state create new money balances? – a simple account
In early 2009 the UK faced macroeconomic catastrophe. A standard macroeconomic principle – that the equilibrium levels of national income and wealth are functions of the quantity of money broadly- defined (i.e., of the quantity of bank deposits, in practice) – held then, as it always holds. But, as in other leading economies, a significant number of banks were under pressure to shrink assets because they could no longer fund their asset totals in the inter-bank market. If not offset by some other process, that asset shrinkage would cause a reduction in the quantity of bank deposits. More fundamentally, every bank was under new official pressure to operate with higher ratios of capital to assets than before. Again, if not offset by some other process, increases in banks’ capital (relative to their deposit liabilities) and reductions in banks’ assets (which implied a lower total of liabilities, including deposit liabilities) would cause falls in bank deposits (i.e., in the quantity of money).
So the private sector, largely in response to the demands from financial regulators, was about to destroy money balances on a large scale. This money destruction was likely to proceed at a rate of at least ½% a month or perhaps even of 1% a month, as experienced (disastrously) in the USA’s Great Depression. The members of the MPC – accustomed to the interest rate tweaking of the Great Moderation – were bewildered. They quickly slashed interest rates to zero (or as near as dammit to zero), but it was evident in the opening weeks of 2009 that even zero interest rates would not soon restore a positive rate of growth of bank credit to the private sector. The central priority of macroeconomic policy was therefore for the state to create new money balances. (The inauguration of a money squeeze by one economy in a strongly expansionary world economy ought to lead to a balance-of-payment surplus, pulling in money balances from the rest of the world. But this was not the situation in early 2009, as money contraction was threatened in nearly all the leading economies. It was threatened by much the same forces as in the UK, i.e., the regulatory push for more bank capital relative to assets and the funding difficulties of banks with large net debtor positions to other banks).
The creation of money by the state is, in essence, a simple matter. It was explained in a previous note (‘Is UK monetary policy on the right lines?’ 1 ½, sent out on 14th March) that nowadays banks expand their balance sheets by adding identical sums to both sides of their balance sheets. Because liabilities are dominated by deposits (which are money), any growth of bank assets creates new money balances, at any rate in the usual course of events. Money is therefore created, literally, ‘out of thin air’ and banks possess ‘the widow’s cruse’. The government is of course the most creditworthy entity in a nation because it has fiscal powers, that is, the powers by law to raise taxes and to print legal tender. All that is required for the state to create money is for it (or its agencies) to borrow from the banking system.
£100m. of new money is created when a bank lends to a company. The bank adds £100m. to the assets side of its balance sheet and this loan may stay unchanged as part of the bank’s assets for several years; it simultaneously adds £100m. to the liabilities side of its balance sheet, crediting that sum to the company’s deposit. The company then spends the deposit, which goes into another agent’s deposit, this second agent then spends the deposit, which goes into a third agent’s deposit, and so on. The value of the payments expedited by the new money may in a year be 50 times as much as the value of the newly-created money itself and, hence, over a decade 500 times its value. (Money circulates, remember. So over a decade the increase in the value of payments arising from the act of money creation is a multiple of that money quantity, with the multiple dependent on the so-called ‘velocity of circulation’. MV = PT, if you had forgotten.)
In the above paragraph merely substitute ‘the state’ for ‘a company’. Then the argument proceeds in exactly the same way. The creation of money by the state to neutralize a deep recession is, in its general outline, as simple as that. (I set out the detailed institutional mechanics in my February 2009 pamphlet for the Council for the Study of Financial Innovation on How to stop the recession.) Obviously, the state is increasing the quantity of money and thereby raising equilibrium national income relative to what would otherwise have eventuated. Also obviously, monetary policy is being conducted in order to influence the quantity of money, which explains why such operations became known as ‘quantitative easing’. In contrast to the Great Moderation, when monetary policy consisted in the tweaking of the short-term interest rate to influence the rate of private sector credit expansion (and hence money creation), monetary policy now had to include discussion of the size of ‘the state’s’ borrowing from the banking system. (I have apostrophized the words ‘the state’ for reasons which will become apparent in the discussion of the technical and institutional content of QE. There is a serious ambiguity here. Another minor caveat is that the deposit credited to ‘the state’s’ account is not usually included in the quantity of money, and with good reason because the government’s expenditure – unlike that of a private sector agent – is not affected by the size of its bank balance. But, when the state spends the balance in its account and the sum is credited to a private sector deposit, that private sector deposit is money, as normally understood in official definitions.)
But why, then, are there so many debates about ‘quantitative easing’?
The section above suggests that the use of the state’s money-creation powers to stimulate an economy is straightforward. So indeed it is. Monetary policy-makers’ agenda consists in adjusting the size of the state’s borrowing from the banking system in order to maintain low and stable growth of the quantity of money. They need always to take a view on these adjustments, just as they need always to take a view on the banking system’s likely pace of credit extension (and hence money creation) to the private sector. In its decisions on the size of QE that is what the MPC has been doing, more or less, since early 2009. Over the last four years Mervyn King’s pronouncements on money creation, and the desirability of steady money growth, are readily interpreted in this framework.
(I am nevertheless baffled by his incessant calls for ever-more bank capital and his apparent indifference to the effect of such demands on money growth. I was even more baffled in late 2008 and early 2009 by his diktats that banks simultaneously expand their lending and operate with higher capital/asset ratios, and his suggestions that if they did not comply with these irreconcilable, indeed impossible demands they would be nationalized, presumably without compensation. The history books will handle King roughly for his central role in the disasters that led to the Great Recession.)
But – if the matter is so simple – why are there such a large number of debates about QE? I suggest that the debates arise for two main reasons,
- i. confusion (or indeed a multiplicity of confusions) about the transmission mechanism of monetary policy and, more specifically, about the macroeconomic variable (or variables) relevant to the determination of national income, and
- ii. confusion(s) about the technical and institutional content of the monetary policy operations that constitute ‘the state’s borrowing from the banking system’.
Even worse people who are confused (or downright mistaken) about the transmission mechanism of monetary policy are often confused/mistaken also about the technical and institutional details of QE operations. A good example of the muddles that have arisen in the QE era is provided by Adair Turner’s lecture at Cass Business School on 6th February on ‘Debt, money and Mephistopheles: how do we get out of this mess?’. As Turner is not (and never had been) a specialist in monetary economics, he deserves high marks for trying to teach himself about the subject while also serving as chairman of the Financial Services Authority. However, the lecture is full of misunderstandings and it is no excuse that these misunderstandings are at present widely shared.
The following discussion will start with a review of the confusions about the transmission mechanism of monetary policy and then review those concerned with the technical and institutional content of QE operations
Confusions about the transmission mechanism of monetary policy
The bare bones of the monetary theory of national income determination were outlined above, in the first two paragraphs of the section ‘A framework for monetary stability’. The theory is basic to understanding how an economy works and, in its essentials, it is very simple. (I have set out this theory on numerous occasions, most recently at some length – perhaps not at sufficient length – in the essays in part five of my 2011 book Money in a Free Society.) However, some aspects of the attainment of monetary equilibrium (i.e., the equivalence of the demand to hold money with the quantity of money created by the banking system) are far from intuitive, while expositors of the quantity theory of money have tended to be conservative in their politics and hence were unfashionable in academe in the second half of the 20th century. As a result of the academic pooh- poohing of the monetary theory of national income determination, it was no longer taught in most English-speaking universities by the 1970s and still is not much taught today. (The denigration of the quantity theory was particularly extreme in, for example, Cambridge and Oxford, England, and in Cambridge, Massachusetts, and related locales [i.e., Harvard, MIT and other Ivy League universities in the USA], with profound effects on the beliefs of economists at the Bank of England and the Federal Reserve.) A range of other theories has circulated, with three having particular prominence. They are summarized ‘in a nutshell’ as follows,
- Changes in expenditure are related somehow to ‘interest rates’ (defined in a number of ways), and so ‘interest rates’ are relevant to the determination of the level of national income and expenditure
- The level of (or the change in ) national income is a function of the level of (or the change in) the monetary base, meaning the cash liabilities of the central bank (which, again, can be defined in more than one way).
- Some concept of the level of or change in bank lending to the private sector is relevant to the determination of national income.
Of course, the nutshell versions of the theories are not entirely fair, but they do convey the emphases in well-known strands of policy thinking. In none of these three theories does the quantity of money as conventionally defined (i.e., the money balances held by private sector non-banks and dominated by the deposit liabilities of the commercial banking system) play any role whatsoever. So to many of the economists who believe in the three theories, and indeed the many variants of the three theories, the effect of QE on the quantity of money is neither here nor there! (I am not kidding.) Let me review the three theories in more detail.
Views of the transmission mechanism which focus on ‘the interest rate’ (or ‘interest-rate-based macroeconomics’)
The first proposition expands into at least two bodies of thought, the first being what might be called ‘Taylor-rule macroeconomics’ and the second into one version of ‘creditism’. In Taylor-rule macroeconomics central banks are assumed able to set one particular interest rate (i.e., ‘the money market rate’), which gives them sufficient leverage over the entire macroeconomic situation. (The Taylor rule made the appropriate money market rate a function of inflation and the output gap, following a celebrated 1993 paper by John Taylor, a professor at Stanford University who was Under Secretary at the Treasury during the G. W. Bush Presidency.) The second idea is that ‘expenditure’ (in some sense) depends on ‘credit spreads’ (to be understood as the margin charged by banks and/or bondholders over a ‘safe’ rate of interest and/or a money market rate set by policy-makers). This idea derives from the ‘creditism’ of Ben Bernanke and associated authors (Gertler, Blinder), and some jointly-authored papers of about 10 to 15 years ago.
I am not going to spend much time here on Taylor-rule interest-rate-based macroeconomics. The dilemma of the current macroeconomic juncture arises – essentially – from the fact that the money market rate is as near to zero as it can be and, in that sense, ‘the rate of interest’ has ceased to be a weapon in macroeconomic management. (Taylor seems not to have considered the possibility that the interest rate recommended by his rule could go negative.)
But it should be noted that a version of QE is advocated by those who believe in the importance of credit spreads. Ostensibly QE could be implemented by the banking system lending to the state, and the state using the loan proceeds to purchase low-credit-quality assets, raising their price and hence reducing credit spreads. In other words, QE is used to affect a particular ‘interest rate’ (i.e., that at which private sector agents borrow), in the conviction – at least partly – that that this rate has a bearing on the quantity of bank credit (or even ‘credit’ altogether) and, at a further remove, the quantity of bank credit influences national expenditure and income. Many observers are attracted to this prescription, but – in my view – it is implausible.
The volume of transactions between private sector agents in securities of different credit-worthiness is a very high multiple of any official operations of the QE type. The relative yields on such securities depend above all on the preferences and knowledge of private sector investors. If QE does not alter the quantity of money, and if the relative default probabilities of different securities are given by non- policy forces, these preferences and knowledge must continue to determine relative yields in the long run. (I don’t deny that – if the state buys a significant proportion of mortgage-backed securities in a short period of time – the yield on MBS will fall. But how long with the effect last? There is a risk that the state will end up as the buyer-of-last-resort of low-quality securities and then have high losses on those securities. The state should operate in high-quality securities, ideally government securities, and focus on the quantity of money, not credit spreads.)
Views on the transmission mechanism which focus on the monetary base (or ‘base-ism’)
Another line of thought is that expansionary open market operations (or QE in some versions) should focus on the monetary base, because the level of equilibrium national income is a function of the monetary base by itself. The supposed importance of the monetary base (or ‘currency’) by itself was one theme of the Currency School in early 19th century debates on banking policy, but it was ignored over the many decades of the quantity theory’s heyday in the early 20th century. In the quantity theory’s heyday discussion concentrated on the quantity of money as such (i.e., a quantity dominated by bank deposits). The revival of base-ism in modern times is the work of the post-Milton Friedman Chicago School of Economics and particularly of Eugene Fama.
A problem of definition immediately arises. The monetary base can be defined in several ways, more specifically as
- The ‘broad’ monetary base, i.e., the notes and coin held by the general public, and the cash reserves of the banking system, including both the cash balance at the central bank, and the notes and coin held in the banks’ vaults/tills (i.e., ‘vault cash’), and
- The ‘narrow’ monetary base, which consists only of banks’ cash reserves.
Unfortunately, it is often unclear from their statements exactly which definition the base-ists believe to matter above all.
The broad monetary base includes the non-bank private sector’s cash holdings, i.e., the notes and coin held by the household sector, and the notes and coin held by retailers (which are the only significant holders in the corporate sector). There appears to be some logic in seeing a connection between such cash holdings and the level of retail sales (where cash is often used), which tends to be seen as emblematic of ‘aggregate consumption’. With consumption as the largest component of aggregate demand, the private sector’s cash holdings become of great macroeconomic significance. It seems that the adoption of M0 as its favourite monetary indicator by the UK Treasury in the 1980s originated in this sort of thinking.
The narrow monetary base is deemed a powerful macroeconomic variable for very different reasons. Evidence over many decades shows (or rather showed until 2007) that the ratio of banks’ cash reserves to their deposit liabilities was fairly stable. If one then believed, for example, that bank lending was important to demand, and that lending increased by a multiple of any cash added to banks’ reserves, operations to expand the narrow monetary base would have wider stimulatory effects because of this boost to bank credit.
I am not going to arbitrate on the relative merits of the broad and narrow monetary bases, as I regard both lines of argument as unimpressive. The focus on the role of the narrow concept of the base in retail sales is misguided, in my view, for at least two reasons. First, the volatile elements in retail sales are the big-ticket items that are typically bought by a payment from a bank deposit, indeed frequently a bank deposit created by the granting of credit. Secondly, apart from these volatile elements, retail sales are of little macroeconomic interest, being far from a dominant element in total demand and very much something that is determined by past and present income and wealth, rather than something which determines future expenditure and income. Meanwhile the belief that banks’ holdings of monetary base assets is critical to their balance-sheet strategies has been debunked in the current cycle. Banks have held higher ratios of cash to total assets than at any time since the 1940s, but that has not led to a great burst of new bank lending.
One reason for mentioning base-ism and its associated muddles is that in his Cass lecture Turner referred to one great monetary economist, Milton Friedman, and – I am afraid – seriously misunderstood him. The reference was to a 1948 article in which Friedman conceptualized a banking system with assets consisting entirely of (or at any rate in which changes in assets consisted entirely of) government securities. Friedman advocated that the state should at the economy’s normal level of working have a balanced budget. In period of beneath-normal working (i.e., of above-normal unemployment) the state would therefore run a budget deficit, which should be monetized in full; in a period of above-normal working (i.e., of beneath-normal unemployment) the state would have a budget surplus, which should enable the state to repay its borrowings from the banks. So the cyclical variation in the budgetary position would enable contra-cyclical variations of changes in the quantity of money.
Turner has misunderstood Friedman in three respects. First, he says that ‘Friedman’s proposal [assumes] that all money is base money, i.e., that there is no private money creation’. (See p. 8 of Turner’s lecture.) No, Friedman’s proposal related to the quantity of money as usually understood (i.e., an aggregate dominated by bank deposits). (The 1948 paper clearly envisaged the continuation of a privately-owned banking system with the tasks of ‘the provision of depository facilities’ and ‘the facilitation of check clearance’. The references to a money aggregate in the article are all to ‘the quantity of money’, ‘the stock of money’ or ‘the supply of money’, not to the monetary base.) Friedman in 1948 (aetat 36) did indeed take his cue from Henry Simons (a fellow Chicago academic) who wanted in a 1930s’ article to abolish the private creation of money and, in effect, to abolish the lending side of commercial banking. But Friedman dropped this kind of thinking later in his career. (It has often puzzled me that Friedman’s 1948 article has been so lightly criticized, as it was unrealistic ‘blue-sky’ theorizing of an extreme and unfortunate kind. But much work in a similar vein – implicitly hostile to a capitalist financial system – was produced in the very unsettled political environment of the 1930s and 1940s, even by such authors as Henry Simons and Irving Fisher. These authors have often been described subsequently as defenders of the capitalist order!)
Secondly, Turner nowhere states that Friedman repudiated the 1948 article. But he did do so, with a subsequent strongly-stated and often-reiterated preference for a constant growth rate of the quantity of money as the key to macroeconomic stability. (See, for example, p. 194 of my Money in a Free Society, which mentions a 1996 interview between Friedman and two British economists, Brian Snowdon and Howard Vane. In that interview Friedman rejected contra-cyclical variations in the quantity of money as ‘more complicated than necessary’.)
Finally, in the passages relating to Simons and Friedman Turner seems enthused by their dislike (in the 1930s and 1940s) of a fractionally-reserved banking system. Turner then talks about a fractionally-reserve banking system as one which operates with a ratio of capital to assets of well under one and, of course, often with a capital/assets ratio of under 10%. But banking systems are both fractionally-reserved, in the sense that cash constitutes only a fraction of assets, and highly-leveraged, in the sense that capital constitutes only a fraction of liabilities. The fractional-reserving and the high leverage that characterize bank balance sheets are different aspects of those balance sheets, and they need to be carefully distinguished. (Simons and Friedman in fact said very little in their articles in the 1930s and 1940s about the role of banks’ capital in monetary management. Friedman’s 1948 article contained several references to a fractionally-reserved banking system, but – unless I have missed something – it contained not a single reference to banks’ capital and the high leverage that is seen in bank balance sheets.)
[To be continued….]Tags: Bank of England, Central bank, European Union, Federal Reserve System, Interest rate, London, Mark Carney, Monetary policy, Monetary Policy Committee, United Kingdom withdrawal from the European Union