HEROES AND VILLAINS: THE ROLE OF CENTRAL BANKS - The End of Alchemy: Money, Banking, and the Future of the Global Economy - Mervyn King

The End of Alchemy: Money, Banking, and the Future of the Global Economy - Mervyn King (2017)

Chapter 5. HEROES AND VILLAINS: THE ROLE OF CENTRAL BANKS

‘There have been three great inventions since the beginning of time: fire, the wheel, and central banking.’

Will Rogers, American actor and social commentator, 1920

‘I promise on demand to pay,’ affords,

A sort of fascinating sounding words;

And if I’m not the most deceiv’d on earth,

The sound they make is nearly all they’re worth.

Anonymous, The Siege of Paternoster Row, 1826

Before the crisis, I was wandering around the stacks of the London Library one evening (not something I was able to do after the crisis started) when my eye was caught by a title: The Old Lady Unveiled. How could such a risqué title have found its way into the section on money and central banks? I soon discovered that the book, not held in any other library of which I was aware, was a devastating critique of the Bank of England. Written during the Great Depression, it began:

The object of this book is to awaken the public to the truth that the Bank of England, commonly believed to be the most disinterested and patriotic of the nation’s institutions, has been since its foundation during the reign of William of Orange a private and long-sustained effort in lucrative mumbo jumbo.1

Many would say that little has changed in the world of central banks. Certainly, they are more lucrative than ever, making large profits from their enormously expanded balance sheets. Although mumbo-jumbo surfaces from time to time, plain speaking is now very much the order of the day, with central bank governors giving press conferences, testifying regularly before Congress or Parliament, and appearing on television.

The cult of celebrity has reached even the gloomy halls of central banking. President Clinton was once asked by a journalist what it was like to be the most powerful man in the world. Pointing to Andrea Mitchell, White House correspondent for NBC, he replied, ‘Ask her. She’s married to him.’ Her husband was Alan Greenspan, then Chairman of the Federal Reserve. Put on its cover by Time magazine as the key member of the ‘Committee to Save the World’, lionised by former presidential candidate John McCain (who said in one of the debates, ‘I would not only reappoint Mr Greenspan - if Mr Greenspan should happen to die, God forbid, I would do like they did in the movie, Weekend at Bernie’s. I’d prop him up and put a pair of dark glasses on him and keep him as long as we could’), and subsequently vilified on stage and screen (not to mention in print) as the architect of the financial crisis of 2008, Alan Greenspan is unrecognisable, in either guise, as the thoughtful and careful central banker I knew. So it is vital to strip away the magic and mystique of central bankers and see them for who they really are: people. Only then can we ensure that the system in which individuals operate provides the right incentives to behave in a way that leads to the best outcomes for the rest of us.

Attempts by central bankers, such as Ben Bernanke, who followed Greenspan, to take personality out of central banking met with limited success or outright failure. In his bestseller The Lords of Finance, Liaquat Ahmed describes the four shadowy central bank governors who led the world into, and eventually out of, the Great Depression. Eighty years later, the equivalent group of governors, of which I was one, confronted an equally difficult challenge. But the difference was that in the build-up to the crisis that started in 2007, central banks had come out of the shadows into the sunlight. They were embarking on a journey from mystery and mystique to transparency and openness. In the 1990s, central banks, if not their governors, became financial idols. The governors themselves saw it differently. We took to heart Keynes’s advice, ‘If economists could manage to get themselves thought of as humble, competent people, on a level with dentists, that would be splendid!’2 Our goal was to make monetary policy as boring as possible.

It is fair to say that we failed in our ambition. You would have to be a hermit on a desert island to describe the past decade as boring. Around the world, central banks were thrust into the spotlight of controversy as the world plunged into its worst ever banking and financial crisis. Money and banking may seem boring technical topics. But they generate more than enough excitement when things go wrong. Rapid inflation can quickly turn into catastrophic hyperinflation, as in Weimar Germany in the 1920s, and deflation can lead to economic stagnation, as in Japan in the lost decade of the 1990s - and, as many fear, in other parts of the industrialised world today. In 2008 the banking crisis dominated the news.

If money and banks are almost as old as Homo sapiens, central banks are the new kids on the block. As institutions go, most central banks are youthful. Indeed, the reputation of central banks as wise and disciplined institutions, in contrast with the wild excesses of finance ministries, belies their respective ages. The first central bank was the Riksbank in Sweden, set up in 1668. To celebrate its tercentenary it endowed the Nobel Prize in Economic Science. But the Riksbank did not acquire its name until 1867 and was really only a commercial bank until 1897.3 The oldest central bank in continuous existence, the Bank of England, opened for business in 1694 to help the government finance military expenditure. Its tercentenary was a more low-key event: it held a concert and published a book of conference proceedings. Next came the Bank of Spain in 1782 and the Banque de France (founded by Napoleon) in 1800, followed by the Bank of Finland in December 1811 (well before the central bank of Russia, of which Finland was then a part, in 1860).

The United States had two false starts in central banking - with the First Bank of the United States between 1791 and 1811 and the Second Bank of the United States between 1816 and 1836 - before Congress legislated in 1913 to set up the Federal Reserve, which opened for business in 1914.4 A year after the 1907 financial crisis, when a panic in New York led to a fall in the Stock Exchange of 50 per cent and numerous bank failures, Congress set up the National Monetary Commission to report on ‘what changes are necessary or desirable in the monetary system of the United States’. Before recommending the establishment of the Federal Reserve System - a plan which it described as ‘essentially an American system, scientific in its methods, and democratic in its control’ - the Commission produced twenty-two volumes on monetary and banking systems elsewhere, especially in Europe. A complete set sits proudly in the Governor’s anteroom in the Bank of England. The authors noted that ‘the important place which the Bank of England holds in the financial world is due to the wisdom of the men who have controlled its operations and not to any legislative enactments’. They did not therefore see the Bank of England as a model, instead recommending the creation of an institution framed by legislation and responsible to Congress. After the Great Depression, further changes were made in the way the Federal Reserve operated, with greater powers for the board in Washington and the introduction of a clear national interest-rate policy.

Older central banks often had their origins as commercial banks and made the transition from private to public institutions as a result of their dominant position in their home banking market. For that reason they were deeply unpopular because they were seen as exploiting that position. From the outset, each time the Bank of England charter came due for renewal, a torrent of pamphlets condemned its privileged position. The Bank Charter Act of 1844, which gave the Bank the exclusive right to issue new banknotes, unleashed a new surge of anti-Bank sentiment. The Bank’s directors were variously described as ‘torpid as toads’ and ‘priests of Moloch’s blood-stained altar’.5Whatever criticisms were directed at central bankers during the crisis, we could console ourselves that they were more moderate in tone.

When, in 1832, President Andrew Jackson vetoed the renewal of the charter of the Second Bank of the United States, he argued that Congress had no constitutional right to delegate the issuance of paper money to any other body:

It is maintained by some that the Bank is a means of executing the constitutional power ‘to coin money and regulate the value thereof.’ Congress have established a mint to coin money and passed laws to regulate the value thereof. The money so coined, with its value so regulated, and such foreign coins as Congress may adopt are the only currency known to the Constitution. But if they have other power to regulate the currency, it was conferred to be exercised by themselves, and not to be transferred to a corporation. If the bank be established for that purpose, with a charter unalterable without its consent, Congress have parted with their power for a term of years, during which the Constitution is a dead letter. It is neither necessary nor proper to transfer its legislative power to such a bank, and therefore unconstitutional.6

After the experience of banking crises in the late nineteenth and early twentieth centuries, Congress was persuaded that a central bank was both constitutional and a good idea. What led to the change in view? During the era of free banking, described in Chapter 2, the US had no central bank. Banknotes issued by commercial banks often traded at a discount to their face value. That made them less useful as money that could be used to buy stuff or as a store of value. There were concerns that banks might issue too many notes in order to exploit the lack of information among depositors about the solvency of the bank. And when there was a crisis in the banking system there was no central authority to restore confidence - in 1907 the task of putting together a consortium of banks to support their weaker brethren fell to John Pierpont Morgan, founder of the eponymous bank. In a similar fashion, the German Reichsbank was set up in 1876, not to coincide with unification of the country in 1871 but in response to a financial crisis in 1873. Central banks acquired their modern role as the result of experiences of earlier monetary and banking crises. The position of central banks that started life as commercial banks developed into that of first among equals, organising occasional rescues and acting in effect as the secretary of the club of banks, above the competitive fray in which other banks were engaged.7

By the twentieth century, central banks were gradually evolving into today’s powerful institutions, responsible for managing the money supply and overseeing the banking system. Concern about the power of central banks remains a popular political position on both left and right, as expressed by the slogan ‘end the Fed’.8 Central banks were seen as heroes for delivering the decade of the Great Stability and for preventing a relapse into a second Great Depression after 2008. They were seen as villains for having failed to rein in the excesses of the banking system in the first place and then for creating money on a massive scale. Compared with the late 1990s and early 2000s when their reputation peaked, central banks are now on the back foot, defending their hard-won independence from the ambitions of politicians of all colours. When the Federal Reserve reached its centenary it felt obliged to describe the event as ‘marking’ rather than celebrating the milestone, and relied on charitable donations rather than its own funds to finance the accompanying exhibition in the American Museum of Finance on Wall Street. Even the courts are getting in on the action - Judge Thomas C. Wheeler of the United States Court of Federal Claims ruled in June 2015 that the Federal Reserve had acted beyond its legal authority in taking a large equity stake in the insurance company AIG in return for a bailout of the company during October 2008.9 In Germany, the Federal Constitutional Court has expressed reservations about proposals by the European Central Bank to purchase the sovereign debt of some periphery members of the euro area.10 Yet despite these challenges to their authority, governments have relied more and more on central banks - especially the Federal Reserve, the European Central Bank, the Bank of Japan and the Bank of England - to deliver a recovery from the Great Recession.

Experience has demonstrated the importance of a public body - normally the central bank - responsible for two key aspects of the management of money in a capitalist economy. The first is to ensure that in good times the amount of money grows at a rate sufficient to maintain broad stability of the value of money, and the second is to ensure that in bad times the amount of money grows at a rate sufficient to provide the liquidity - a reserve of future purchasing power - required to meet unpredictable swings in the demand for it by the private sector (see Chapters 2 and 3 respectively). Those two functions are rather simple to state, if hard to carry out. They correspond to the twin objectives of price stability and the provision of liquidity by a ‘lender of last resort’.

Price stability - inflation targeting as a coping strategy

Eighteenth-century thinkers, such as David Hume and Adam Smith, understood the relationship between the amount of money in circulation and the prices at which goods and services were bought and sold: ‘if we consider any kingdom by itself, it is evident, that the greater or less plenty of money is of no consequence; since the prices of commodities are always proportioned to the plenty of money’.11 In the long run, more money means higher prices. The quantity theory of money, later refined and popularised by the American economists Irving Fisher and Milton Friedman, had been born.

Over the years, governments have been unable to resist the temptation to debase the currency, and, with the advent of paper money, to print as much of it as possible to finance their expenditures. Lenin is alleged to have remarked that the best way to destroy capitalism is to debauch the currency. To judge by the subsequent experience in Europe after the First World War, he was right. Even where market economies survived, inflation was a problem. In the twenty-five years before the Bank of England adopted an inflation target in 1992, prices rose by over 750 per cent, more than over the previous two hundred and fifty years.12 Inflation was simply taken for granted. Price stability seemed an unlikely state of affairs.

Alan Greenspan, former Chairman of the Fed, defined price stability as when ‘inflation is so low and stable over time that it does not materially enter into the decisions of households and firms’.13 Alan Blinder, the Princeton economist who was Greenspan’s deputy at the Federal Reserve Board, put it even more clearly. Price stability, he said, was ‘when ordinary people stop talking and worrying about inflation’.14 In recent years, we have started to take price stability for granted; so much so that some people have become exercised about the possibility of deflation - when prices fall. Deflation is just as damaging as inflation. In AD 274 the Roman emperor Aurelian tried to restore the integrity of the coinage, which had been adulterated by workmen in the mint. Aurelian exchanged bad money for good, and ordered the destruction of all accounts drawn up in the devalued currency. Prices fell overnight. Gibbon, in his History of the Decline and Fall of the Roman Empire, observed that ‘a temporary grievance of such a nature can scarcely excite and support a serious civil war’.15 And in the long run the operation did restore the value of money. But in the short run it caused hardship. Taking the Keynesian view that in the long run we are all dead, the population at the time rose in insurrection. Many of them found that they were dead in the short run as well; seven thousand soldiers and countless civilians perished during the suppression of the uprising.

In more modern times, governments, even if they profess a belief in price stability, have found themselves tempted to depart from the path of righteousness in order to obtain a short-term benefit by stimulating the economy prior to an election in the hope that the inflationary cost will become apparent to the electorate only after the vote. Once having given in to temptation, they are faced with an unpalatable choice between a recession to bring inflation back down again, or high and possibly accelerating inflation. Taken together, the verdict of economics, history and common sense is that both inflation and deflation are costly. Giving a central bank the exclusive right to issue paper money raises the question of how we can prevent the abuse of the power to issue money. We cannot commit future generations - or even ourselves - to a particular policy. So how can we design an institution to create the reasonable expectation that money will retain its value?

By tying the currency to the mast of gold it seemed that price stability over a long period was attainable, as indeed it was for much of the nineteenth century. But even the gold standard could not override national sovereignty, and, when the costs (in terms of lost output and employment) of adhering to the standard appeared too high, governments suspended the convertibility of their currency into gold, as happened on several occasions in Britain and other European countries in the nineteenth century during financial crises. There was an underlying need to find a way to retain domestic control over the supply of money and liquidity while at the same time retaining a long-term commitment to price stability.

Unfortunately, the switch from a fixed rule, such as the gold standard, to the use of unfettered discretion led to the failure to control inflation, culminating in the Great Inflation of the 1970s. Attention turned to the idea of delegating monetary policy to independent central banks with a clear mandate to achieve price stability. Central banks were not born with independence, they had it thrust upon them - literally, in the case of Germany when, after the Second World War, the Allies imposed the model of an independent central bank. The movement towards independence gathered pace in the 1990s as a reaction to the Great Inflation. The Bank of England and the Bank of Japan were made independent in 1997, the Swedish Riksbank in 1999, and in the same year the independent European Central Bank was set up, influenced by the track record of the Bundesbank in Germany which had, since its creation in 1957, achieved lower inflation than in other industrialised countries.16

Of course central banks may themselves be tempted to court popularity. One riposte to that concern is that central bankers are different from politicians. Central bankers who have the determination and strength of purpose to ‘take the punch bowl away just when the party is getting going’, in Federal Reserve Chairman McChesney Martin’s memorable phrase, clearly have the right stuff - so why don’t they ‘just do it’?17 It would be nice to think that all central bankers were made of the right stuff, and maybe they are. But there is no need to rely on our ability to identify superhuman individuals. Instead, the answer is to devise an incentive structure for the individuals appointed to run central banks.

If the elected government, or its advisers, understood exactly how the economy worked then it could write a contract specifying precisely what the central bank should do in each possible state of the world that might arise in the future. Monetary policy could then be delegated to an independent central bank tasked with implementing the contract. There are two problems with this idea. First, governments might be tempted to tear up the contract in precisely the same circumstances that they themselves would give in to the temptation to allow inflation to rise. Second, in a world of radical uncertainty we cannot write a detailed contract covering all possible future events. The future is literally undescribable. Economists have tended to devote more attention to the first of these problems than the second. I am inclined to think that the reverse should be the case. Coping with temptation is easier than coping with the entirely unknown.

Our inability to identify in advance the challenges that will arise in managing money means that it is sensible, indeed unavoidable, to grant the central bank a degree of discretion in responding to unfolding events. This is the basic idea behind inflation targeting, which originated in New Zealand in 1990. The idea soon spread to Canada in 1991 and the United Kingdom in 1992. The aim was to hold the central bank to account for achieving a numerical target for inflation over a specified period. Central banks were given discretion over the extent to which they responded to short-run movements in inflation.

Such movements are unpredictable. Prices and wages do not adjust instantaneously to clear markets whenever demand and supply are out of balance. Firms change prices only irregularly in response to changes in demand; wages adjust only slowly as labour market conditions alter; and expectations are updated only slowly as new information is received. Such ‘frictions’ or ‘rigidities’ introduce time lags into the process by which changes in money lead to changes in prices. These lags in the adjustment of prices and wages to changes in demand - so-called ‘nominal rigidities’ - and lags in the adjustment of expectations to changes in inflation - ‘expectational rigidities’ - generate short-run relationships between money, activity and inflation.18 Monetary policy affects output and employment in the short run and prices in the long run. Central banks care about both. This is captured by the so-called dual mandate of the Federal Reserve, which states its objectives as maximum employment and stable prices.19 The overriding concern of central banks is not to eliminate fluctuations in consumer price inflation from year to year, but to reduce the degree of uncertainty over the price level in the long run. People will then stop worrying about inflation.

An inflation-targeting monetary policy is a combination of two elements: (a) a target for inflation in the medium term and (b) a response to economic shocks in the short term. From time to time shocks - to oil prices or the exchange rate, for example - will move inflation away from its desired long-term level, and the policy question is how quickly it should be brought back. The answer depends on the relative costs of deviations of inflation from the target and of unemployment from its long-term equilibrium level, and central banks have discretion in making that judgement. From this perspective there is no essential difference between the actions of a central bank with a Fed-style dual mandate and a central bank with a single mandate to meet an inflation target. What is crucial is that households and businesses believe that prices will be stable in the long run.

Inflation targeting has been highly successful, both in its primary aim and as a way of ensuring the democratic accountability of powerful public institutions. Some economists have argued that central banks should be compelled to set policy according to a ‘policy rule’ set by legislators, or at a minimum to explain why their chosen policy deviates from that implied by the rule. Monetary policy rules have become a major area of research.20 Perhaps the most famous is the so-called Taylor rule, named after John Taylor of Stanford University. The Taylor rule implies that interest rates should rise if inflation is above its target and output is above its trend level, and fall when the converse is true. In 2014, Representatives Scott Garrett and Bill Huizenga introduced a bill that would require the Federal Reserve to provide Congress with ‘a clear rule to describe the course of monetary policy’.21 Such a rule would be a mathematical formula showing how the Fed would adjust interest rates in response to changes in the economy.

Although it is clearly desirable for the Federal Reserve to be held directly accountable to Congress for its actions, the fundamental flaw in this proposal is that there is no timeless rule that is likely to remain optimal for long. Since our understanding of the economy is incomplete and constantly evolving, sometimes in small steps, sometimes in big leaps, any monetary policy rule judged to be optimal today is likely to be displaced by a new and improved version tomorrow. Whatever rule might be mandated in legislation would be superseded by new research within a year.

A good example was the experience of the Federal Reserve and the Bank of England during 2013 and 2014, when they announced the rate of unemployment at which they would start to consider raising interest rates. What looked in 2013 a plausible unemployment rate that would trigger a rise in interest rates turned out to be much less plausible by 2014, when unemployment had fallen faster than expected without signs of a pick-up in inflation. Monetary policy in practice is characterised by a continuous process of learning. Learning from experience means that it is sensible to be prepared to deviate from a rule constructed even a year or two ago. Rather, the onus should be on the central bank to justify its behaviour in terms of presenting convincing economic arguments and evidence for them. Accountability and transparency are superior to the use of a fixed rule.

Delegating policy to an independent central bank operating under a well-specified regime of ‘constrained discretion’ was seen as the answer to the unappealing choice between adopting a fixed rule and giving unfettered discretion to an independent body. Such a framework required a clear definition of the constraints to be imposed on central banks. One was a numerical target for inflation, and a second was the establishment of a regime under which central banks could be held accountable for their decisions. From the outset, inflation targeting was conceived as a means by which central banks could improve the credibility and predictability of monetary policy.

Since its adoption in New Zealand, Canada and the United Kingdom in the early 1990s, inflation targeting has spread to more than thirty countries around the world.22 The big central banks now all have an inflation target of 2 per cent, with the Federal Reserve adopting it in 2012 and the Bank of Japan in 2013. In the language of Chapter 4, delegating monetary policy to an independent central bank with an inflation target is a coping strategy. Its clarity and simplicity mean that the target provides a natural heuristic for central banks and the private sector alike. The heuristic for the former is to set policy such that expected inflation is equal to the target rate, and for the latter it is to expect inflation equal to the target rate. Since expectations of inflation have a major influence on the setting of wages and prices, and hence on inflation itself, anchoring expectations on the target is a key element of any credible monetary policy. And the heuristic frees the central bank from having to commit to any one particular model of the economy when making its judgement about the likely future path of inflation. The great attraction of an inflation target is that it is a framework that does not have to be changed each time we learn something new about how the economy behaves.

Inflation targeting is about making and communicating decisions. It is not a new theory of how money and interest rates affect the economy. But, by anchoring inflation expectations on the target, it can in theory reduce the variability and persistence of inflationary shocks - and has done so in practice. And it has done so without pretending to commit to a rule that is incredible because it is not expected to last.

Old problems and new instruments

There are, however, deeper reasons to ask why central banks should worry only about consumer price inflation rather than the state of the real economy. Inflation targeting is designed to mimic the behaviour of a competitive market economy, one that exhibits none of the nominal or expectational rigidities that prevent prices from adjusting immediately. This makes perfect sense within the confines of the conventional economic models used by central banks. But those models take no account of radical uncertainty. And the problem that central banks need to confront is whether there are other significant imperfections in the economy that justify departing from an inflation target. Confronted with radical uncertainty, it is natural that households and businesses make occasional ‘mistakes’, for example about their future incomes, and realise their errors only after a considerable time lag. Such mistakes can accumulate into substantial deviations of spending and output from a sustainable path, even though they may have little impact on inflation in the short run. This is not the outcome of short-run rigidities but of misjudgements about the nature of the future.

The practical significance of this question has been highlighted by the current disequilibrium in the world economy. Should central banks take responsibility for trying to correct such mistakes before households and businesses come to a true appreciation of the situation, or should they stay focused solely on targeting inflation a year or two ahead? Did central banks contribute significantly to the crisis by not trying to correct the big mistakes made by the private sector? To suggest that monetary policy has the purpose of preventing the economy from getting into an unsustainable position is tantamount to arguing that central banks should, on occasions, target the real equilibrium of the economy and not just price stability - a much deeper and more difficult question than that of whether a central bank should have a dual or a single mandate. The fundamental question is whether central banks should take responsibility for preventing substantial deviations of real variables, such as spending and output, from their normal levels, because the cost of permitting the continuation of a large and growing disequilibrium is a crash at some point in the future, followed by economic stagnation and persistently low inflation.

The proper role of a central bank in guiding the economy is, therefore, a thorny and controversial issue. I shall return to it in a concrete context in Chapters 8 and 9, where I ask whether central banks could have prevented or reduced the severity of the crisis by following a different monetary policy, and consider what they should do today. Related concerns about the desirability of inflation targeting have been raised by those who believe that central banks should focus at least as much on ‘financial stability’ as on ‘price stability’, meaning that monetary policy can and should try to affect much more than just short-run movements in inflation. The difficulty with this proposition is that failure to achieve financial stability covers a multitude of sins. It may reflect the consequence for asset prices of ‘mistakes’ by economic agents - the rapid increases often loosely described as ‘bubbles’. Or it may reflect excessive fragility in the banking sector resulting from excessively high leverage and interconnectedness. The appropriate policy response depends on the cause of the instability. The provision of an appropriate amount of ‘emergency money’ (see below) can ameliorate the immediate consequences, but to prevent instability arising may require either action on monetary policy or a change in the structure of the banking sector. Rather than thinking deeply about the framework for monetary policy or radical change in banking, policy-makers have sought new instruments to deal with the potential causes of financial instability. The most important such instruments go under the name of ‘macro-prudential policies’.

Macro-prudential instruments include direct controls on financial markets - for example, setting limits on the size of mortgage loans relative to incomes - and indirect controls - such as requiring banks to use more equity finance if they increase lending to areas that are judged particularly risky. These quantitative controls are equivalent to setting different interest rates for different types of transaction. At the Bank of England, the Monetary Policy Committee decides on the level of the official short-term interest rate (Bank Rate) and a new committee, the Financial Policy Committee, set up in 2011, decides on the macro-prudential measures that act as addons. The distinction between monetary and macro-prudential policies is not clear-cut. A crude way of thinking about the difference is that the former is about the amount of money in the economy and the latter is about the allocation of credit across sectors.

Before the crisis, central banks believed that their role was not to enter into the allocation of resources, but rather to guide the economy by sending price signals (in the form of interest rates) about the appropriate relative prices of spending today and in the future. Today, the use of measures to intervene in particular asset markets is all the rage. But the scope for tensions between the two sets of decisions is evident. For example, in the Swedish Riksbank from 2010 to 2013 there was a sharp, and often bitter, division between two groups. The first wanted to raise interest rates because of concerns about the pace at which house prices and indebtedness were rising. The second thought that responsibility for dealing with the housing market should be left with the supervisory authority, which, as it happened, was more relaxed about housing developments than were the majority in the central bank.

Quantitative restrictions on credit are by no means a new policy instrument. They were deployed in many advanced economies in the 1950s and 1960s, and still play a role in many emerging and developing economies. As banking and financial markets were liberalised in the advanced economies in the 1970s and 1980s, and opened up to foreign competition, most of these controls were scrapped. Although central banks can determine interest rates in their own currency, they cannot easily restrict the lending activities of foreign banks. Cooperation between regulators has been improved since those days, but how far macro-prudential measures will be successful in today’s world of borderless capital markets remains an open question. Nimbleness and the ability to respond quickly to events are important features of interest-rate policy. It will be more difficult to act, and to defend and explain rapid changes in restrictions on lending (for example, the maximum ratio of a mortgage loan to the value of the house) than in interest rates.

Out of a political consensus on the importance of ensuring monetary stability emerged an agreement that democratic societies would delegate to an unelected central bank the power to set interest rates, even though this would have effects on the distribution of income and wealth. But the entry of central banks into the field of direct controls on mortgages and lending more generally is bound to raise the question of whether this is taking delegation too far. It is possible to debate the merits of intervening in the market allocation of credit to help, for example, first-time homeowners and small businesses. But decisions on that type of interference with the market should properly be left to elected politicians with a mandate to take such action. It is hard to see why central banks should want the power to intervene in the microeconomic allocation of credit. If, as has happened over the past twenty years, saving and spending get out of kilter, then central banks will come under increasing pressure to intervene in particular financial markets to correct so-called distortions, which are in fact the result of a macroeconomic disequilibrium. And ‘distortions’ and ‘excesses’ will then pop up in other markets. Down that road lies a degree of intrusion into individual decisions on saving and credit that is incompatible with an innovative market economy.

Amid the post-crisis confusion about whether central banks should focus solely on price stability, or whether they should take responsibility for guiding the economy to a new equilibrium, or deal with potential ‘bubbles’ in asset prices, one might be forgiven for thinking that central bankers should follow the example of the Church of England in making a general confession: we have not targeted those things which we ought to have targeted and we have targeted those things which we ought not to have targeted, and there is no health in the economy.23

Expectations and communications

When I joined the Bank of England in 1991, I asked the legendary American central banker Paul Volcker for one word of advice. He looked down at me from his great height (a foot taller than I) and said, ‘Mystique.’24 He was talking about the importance of businesses, households and financiers having confidence in the central bank. Today that confidence has to be earned in a much more transparent way. During my time at the Bank of England, it became apparent that politicians and central bank governors were on a divergent path. As they try to make an impression on the electorate, politicians have become taller and taller, whereas central bank governors have become shorter and shorter. Paul Volcker was followed by Alan Greenspan, Ben Bernanke and Janet Yellen, a steady decline in height. At the Bank of England, Gordon Richardson, the counterpart of Paul Volcker during the 1980s debt crisis, was followed by Robin Leigh Pemberton, Eddie George, myself and Mark Carney. It is evident that central banks have come to rely less on height and hauteur and more on transparency and the ability to look the average person straight in the eye.25

Businesses and households base their decisions on expectations of the future, and so the way we expect monetary policy to be conducted in the future affects economic outcomes today. Consider a simple and stark example. Suppose that there were no frictions or time lags in the way the economy responded to changes in monetary policy. Imagine a central bank which, in those conditions, was successful in controlling inflation perfectly by responding to all shocks instantaneously. The outcome would be a constant inflation rate. Interest rates would move around but with no apparent link to or effect on inflation. To an observer - whether journalist or econometrician - interest-rate changes would appear to have little to do with inflation. The central bank would appear to be behaving almost randomly. By assumption, that inference would be false. Indeed, if people did expect the central bank to behave randomly, then the behaviour of households and firms would change and inflation would no longer be stable.

This observation leads to what we might call the Maradona theory of interest rates. The great Argentine footballer, Diego Maradona, is not usually associated with the theory of monetary policy. But his performance against England in the World Cup in Mexico City in June 1986 when he scored twice is a perfect illustration of my point. Maradona’s first ‘hand of God’ goal, when he deliberately punched the ball into the England net unseen by the referee, was obviously against the rules. He was lucky to get away with it. His second and quite brilliant goal, however, was an example of the power of expectations. Maradona ran sixty yards from inside his own half, beating five players before shooting into the English goal. The truly remarkable thing, however, is that, as cameras positioned above the stadium showed, Maradona ran virtually in a straight line. How can you beat five players by running in a straight line? The answer is that the English defenders reacted to what they expected Maradona to do. Because they expected Maradona to move either left or right, he was able to go straight on.

Monetary policy works in a similar way. Market interest rates react to what the central bank is expected to do. In recent years there have been periods in which central banks have been able to influence the path of the economy without making large moves in official interest rates. They headed in a straight line for their goals. How was that possible? Because financial markets did not expect interest rates to remain constant. They expected that rates would move either up or down. Those expectations were sufficient - at times - to stabilise private spending while official interest rates in fact moved very little. An example of the Maradona theory of interest rates in action was seen in the UK in 2002. During that year the Bank of England was able to achieve its inflation target by moving in a straight line with unchanged official interest rates. But, although interest rates scarcely moved, expectations of future interest rates - as revealed in market interest rates - did move around as the economic outlook changed from the expectation of a swift recovery to worries about a protracted slowdown. And in turn those changes in expected future rates affected activity and inflation. In other words, monetary policy was able to respond by less than would otherwise have been necessary because it affected expectations.

Of course, if developments in the economy continue to evolve in the same direction then interest rates will eventually have to move and follow expectations. It should be clear that, just as Maradona could not hope to score in every game by running towards goal in a straight line, so monetary policy cannot hope to meet the inflation target by leaving official interest rates unchanged indefinitely. Rates must always be set in a way that is consistent with the overall strategy of keeping inflation on track to meet the target; sometimes that will imply changes in rates, at other times not. But the key point is that the power of expectations about future rates can often be more important than the current level of the official interest rate itself.

Because the expectation of what central banks might do has become as important as their immediate actions, if not more so, an entire industry of private sector central bank watchers has grown up. They now comment ceaselessly on when the Federal Reserve or other central banks will change interest rates. But they are like characters in a John le Carré novel, working in the shadows and inhabiting a world of double-talk, coded language and private vocabulary. One of the aims of central banks has been to put this industry out of business and to move to a world of simple, clear language. The advent of inflation targeting saw a move from the old central banking tradition of mystery and mystique to openness and transparency. The old world was illustrated by Lord Cunliffe, the Governor of the Bank of England during the First World War, who, when giving evidence before a Royal Commission on the size of the Bank of England’s gold and foreign exchange reserves, replied that they were ‘very, very considerable’. When pressed by the commission to give an approximate figure, he replied that he would be ‘very, very reluctant to add to what he had said’. Today the figures are published monthly. Until February 1994, believe it or not, the Federal Reserve did not even reveal what the official interest rate was, or whether it had changed. Analysts and researchers had to infer from market interest rates whether or not the Federal Reserve had changed policy. Today, the Federal Reserve publishes the decision of each meeting along with minutes of the discussion and reasons for its actions.

Transparency is, however, not an end in itself. Any requirement for transparency in a central bank’s deliberations should have the aim of improving the quality of its decisions. The publication of the minutes of meetings of the policy boards of central banks, as well as of regular monetary policy reports or inflation reports, has provided information both to guide expectations as to how the central bank will respond to future events and to explain past decisions. They are the basis for accountability. But the publication of transcripts of meetings can inhibit free and open discussion, and the style of meetings of the Federal Open Market Committee has undoubtedly changed since such transcripts were first disclosed in 1994; prepared formal statements are read out, while the important private discussions take place at earlier, often bilateral, meetings.26 In any policy setting, there has to be room for private conversations. There are limits to the desirable degree of transparency.

It is also important for central banks to be honest about what they do not know. A case in point was the recent, and rather short-lived, experiment in ‘forward guidance’ adopted by the Federal Reserve and the Bank of England in 2013. Both central banks wanted to provide more information about the likely future path of official interest rates. In the first instance, this was a laudable attempt to reduce uncertainty about how they might respond to developments in the economy. But it soon became an attempt to predict the future path of interest rates.

They were not the first to be tempted down this path. For some years, the Reserve Bank of New Zealand and the Swedish Riksbank have published forecasts of their own policy rates. This has not been an entirely happy experience, especially in the latter case when the markets did not believe, correctly as it turned out, the Riksbank’s forecasts about its own policy actions. The danger is that markets and commentators read too much into central bank forecasts of their own future actions. When, as is almost inevitable, the future turns out to contain surprises, interest rates will deviate from the forecast path. Although the latter is not intended to represent a commitment by the central bank to pursue that path, it is only too easy to paint it as such.

In turn, central banks were reluctant to concede that the path should be adjusted. The confidence that central banks wanted the private sector to have in their forecasts was not consistent with the inherent degree of uncertainty surrounding those forecasts. To retain credibility, it is important that central banks do not claim to know more than they in fact do. And it is clear that central banks are not able to provide accurate forecasts of their own actions. Policy must confront the fact that ‘stuff happens’. Making forecasts is inherently difficult. They always turn out to be wrong. The most egregious example of wrong forecasts by central banks was the prediction before the crisis that the Great Stability would continue. Central banks were using forecasting models that ignored the lessons explored in Chapter 4.

New problems and old instruments

Until recently, central banks thought of monetary policy in terms of setting interest rates rather than fixing the supply of money. The two are, of course, closely related. Reducing interest rates stimulates the demand for borrowing and if banks increase their lending, the supply of bank deposits rises. That pushes up the money supply. But frequent and volatile shifts in the demand for money have led central banks to choose interest rates as their principal policy instrument.

Instabilities in the demand for money are not new. In the early years of the Bank of England, there were unexpected shifts in the demand for money and credit resulting from the uncertain arrival times in the Port of London of ships laden with commodities from all over the world. The uncertainty derived from changes in the direction and speed of the wind carrying ships up the Thames. To cope with this, the Court Room of the Bank of England contained a wind dial linked to a weather vane on the roof, which provided an accurate guide to these shifts in money demand - the weather vane is there to this day, and it still works. If only monetary policy could be as scientific today!

To prevent a repetition of the Great Depression, central banks during 2008 and 2009 cut interest rates virtually to zero, at which point influencing the supply of money directly was the only remaining monetary instrument. The new problem they faced was what to do when interest rates are zero and cannot be lowered any further. When official interest rates have reached zero, modern Keynesians draw the conclusion that monetary policy is impotent and only fiscal policy can return the economy to full employment. Central banks did not accept this proposition, and took steps to expand the money supply.

My own explanation was simple. For most of the post-war period, governors of the Bank of England had been trying to prevent the amount of money in the economy from growing too quickly. If it were to expand at a rate much faster than the ability of the economy to grow, then the result would be inflation. But the problem facing the Bank in 2009 was that the amount of money in the economy available to finance spending was actually falling. The reason was that banks had begun to contract their balance sheets by refusing to roll over loans and no longer making new ones, thus reducing their total assets. The automatic counterpart on the liabilities side was a corresponding reduction in deposits as loans were repaid. Since most money comprises bank deposits, the fall in deposits meant that the amount of money available to finance spending actually fell. If left unchecked, that threatened a depression. So the task of the Bank was to ensure that the amount of money in the economy grew neither too quickly nor too slowly.

In the particular circumstances of 2009, that meant creating more money. It did not create inflation for two reasons. First, the increase in the supply of money was matched by a sharp increase in the demand for highly liquid reserves on the part of the banking system and the economy more generally. Second, the total supply of broad money, including bank deposits, rose only moderately. The ‘emergency money’ created by the Bank was necessary to prevent a fall in the total money supply. It was precisely because the demand for money and liquidity changed so sharply that monetary developments mattered. It is ironic, therefore, that economists who believe that money matters (for example, Milton Friedman) argue that ‘the demand for money is highly stable’, whereas Keynesian economists argue that money does not matter because its demand is unstable.27 Both groups are wrong - money really matters when there are large and unpredictable jumps in the demand for it.

The method used to create money was to buy government bonds from the private sector in return for money.28 Those bond purchases were described by many commentators as ‘unconventional’ monetary policies and became known as ‘quantitative easing’, or QE. They were regarded as newfangled and untried. If history is what happened before you were born, then many of the commentators must be extremely young. For open market operations to exchange money for government securities have long been a traditional tool of central banks, and were used regularly in the UK during the 1980s, when they were given the descriptions ‘overfunding’ and ‘underfunding’.29 What was new in the crisis was the sheer scale of the bond purchases - £375 billion by the Bank of England, almost 20 per cent of GDP, and $2.7 trillion by the Federal Reserve, around 15 per cent of GDP. The need for purchases on such a scale reflected the fact that since government and central banks control directly only a small proportion of the money supply - less than 10 per cent, as we saw in Chapter 2 - a large percentage increase in the printing of money is required to create even a moderate increase in the total money supply.

Economists produced convoluted explanations of how and why this extra money might affect the economy through changes in risk premiums and other arcane aspects of the financial system.30 Ben Bernanke, then Chairman of the Federal Reserve, said in January 2014 that ‘the problem with QE is it works in practice, but it doesn’t work in theory’.31 Perhaps there was a problem with the theory.

How does QE work? Such asset purchases inject money into the portfolios of the private sector. Those investors, such as pension funds and insurance companies, who have sold bonds to the central bank will reallocate their higher money holdings among all possible other assets, such as common stocks, corporate bonds and foreign investments. Those purchases change the prices of private sector financial assets, which in turn affects wealth and spending. For example, if investors use their new-found money to buy corporate bonds, the higher price of those bonds will correspond to a reduction in their yield and hence the cost to companies of obtaining finance for new investment.

So, extremely low official interest rates do not exhaust the ammunition of central banks. But when interest rates at all maturities, from one month right out to thirty years or more, have fallen to zero, then money and long-term government bonds become perfect substitutes (they are both government promises to pay which offer zero interest), and the creation of one by buying the other makes no difference. To be sure, a number of advanced countries are close to that position, though none has yet quite reached it. Central bank official interest rates are virtually at zero in many countries. Long-term bond rates, by contrast, as measured by yields on ten-year bonds, are still above zero. In late 2015, bond yields were around 2 per cent in the United States and most other advanced economies, apart from Germany and Japan, where rates were around 1 per cent and 0.5 per cent respectively. Only in Switzerland, of the major economies, were ten-year bond yields slightly negative.

When the yield curve is completely flat, central banks may still create money by purchasing assets other than government bonds - either private sector assets, such as corporate bonds, or overseas currencies (the latter was the main strategy pursued by the Swiss National Bank in a vain attempt to prevent a sharp appreciation of the Swiss franc against the euro). But this means taking on credit risk of a very different kind from that involved in monetary policy, which is limited to buying and selling government bonds of different maturities, and has long been accepted as a legitimate role for central banks. Taking on credit risk, which ultimately falls on taxpayers, means that monetary policy is entering the world of fiscal policy. At this point, it is for governments to take the responsibility of deciding which sectors of the economy should be favoured over others. To be sure, there are circumstances in which the central bank and government, working together, can improve matters. For example, in the midst of the crisis some financial markets (for short-term commercial paper, for example) seized up, and both the Bank of England and the Federal Reserve intervened for a short period as a market-maker of last resort until those markets returned to some semblance of normality. But those crisis conditions in financial markets ended a long time ago. The challenge today is to deal with a period of prolonged weakness in demand.

It is not that monetary policy is completely impotent when central bank interest rates are close to zero. It is that monetary policy runs into diminishing returns; although continually falling real interest rates encourage households to bring forward spending from the future to the present, there comes a point when they are reluctant to sacrifice more and more future spending to increase current spending. I shall return to this quandary in Chapter 9.

Now that official interest rates are virtually at zero, an even more extreme version of forward guidance has been proposed by some economists as a way of stimulating the economy.32 The idea is that central banks should promise to allow inflation to go above their normal target at some point in the medium term so that real interest rates - nominal rates less expected inflation - can fall to more negative levels, so stimulating spending. This is a counsel of despair and is literally incredible. Suppose businesses and markets believed that inflation would indeed be higher in the future and that the resulting lower real interest rate did indeed stimulate recovery. The central bank would then be faced with the following dilemma. Should it proceed to allow inflation to rise above the target despite the recovery, or should it be grateful for the recovery and then set policy to keep inflation on track to meet the target? It is not hard to see that for any central bank governor the latter would be more attractive than the former. Markets will anticipate that reaction and so not believe that inflation will be allowed to rise above target. But then real interest rates will not fall and the recovery will not take place. The strategy of promising to generate an inflationary boom is ‘time inconsistent’; in other words, what you say you will do in the future is not what you will want to do when you get there.

If it really were thought desirable to generate an inflationary boom in order to bring down real interest rates, then there would be a much simpler answer - abolish independent central banks and return interest-rate decisions to government. Markets would then certainly expect higher inflation in future. But we would be back to where we started, when central banks were handed independent responsibility for controlling inflation.

Inflation is not a beast that can be killed once and for all. Success is a matter of the patient application of policies designed to maintain price stability. Central bankers are like doctors - they need to be on top of the latest technical developments, have several years of experience and a good bedside manner. Even then, it may be impossible to do much more than avoid big mistakes and promote a healthy way of living. Stability is like dieting - it is no good alternating between binge and starvation, boom and bust. It is necessary to follow a few principles consistently and in a sustained manner. Inflation targeting represented a healthy way of living for central banks charged with the task of ensuring monetary stability.33 Accountability and transparency provide the incentives for central banks to meet the inflation target. Such a framework of ‘constrained discretion’ is far removed from the world of 1930, when the Deputy Governor of the Bank of England explained to the Macmillan Committee that ‘it is a dangerous thing to start to give reasons’.34

An event in 2007 illustrates the change in the way in which monetary policy was conducted after inflation targeting was introduced and independence was granted to the Bank of England. At 12 noon on Thursday 10 May, Tony Blair announced his resignation as Prime Minister after ten years at Number Ten. At exactly the same moment the Bank of England announced an increase in interest rates of 0.25 percentage points. Nothing could symbolise more vividly the change in the monetary regime in Britain than that conjunction. Before the Bank of England became independent it would have been inconceivable that interest rates would have risen on a day when there was an important government announcement.

Monetary policy in bad times - emergency money

No central banker should be spared the experience of a financial crisis, and few are. Since 2008, central banks have reoriented their focus towards financial stability and, with inflation low and often below their target, have placed less emphasis on monetary stability. To some extent this change of focus is a return to the historical origins of central banks. It has gone hand in hand with the view that the instruments available to central banks need to be extended well beyond those used during the Great Stability. Inevitably, attention turned to refining the instruments designed and deployed in the dark days of 2008 and after, and to the macro-prudential powers discussed above. What has been missing, however, from this broader vision of central banks is a coherent unifying framework within which to analyse policies in both good and bad times, and especially in the best and worst of times. I shall try to show that an integrated framework can be constructed and, indeed, is necessary if we are to challenge the alchemy of our current system.

As I described in Chapter 3, banks borrow short and lend long. This leaves them open to runs by the people who make short-term unsecured loans to banks. In theory, this could be the result of a temporary shortage of cash in the bank. But often it is related to concerns about losses on the loans the bank has made. Because depositors cannot easily coordinate their actions, if a run begins it is rational to join it. As many depositors discovered in nineteenth-century America, being last at the counter is a recipe for leaving empty-handed. Banks are inherently unstable.

Over the past 150 years the conventional wisdom has been that central banks should stand ready to be the ‘lender of last resort’ (LOLR) and supply liquidity to the banking system when the public loses confidence in one or more banks. It is a little-known, though not uninteresting, fact that the Wimbledon Championships are viewed in more countries than belong to the International Monetary Fund, and they are certainly more entertaining than the Annual Meetings of the IMF. Just a stone’s throw from Centre Court is the house in which Walter Bagehot wrote his classic study of central banking, Lombard Street. Setting out the doctrine of the LOLR - lend freely against good collateral at a penalty rate to banks facing a run - it became the bible for central bankers wondering how to respond to financial crises. Ben Bernanke at the Federal Reserve, and other central bank governors, often referred to Bagehot when explaining the measures they had taken to support banks during the recent crisis.35

Although the policy is widely attributed to Bagehot, it can be traced back to Henry Thornton in An Enquiry into the Nature and Effects of the Paper Credit of Great Britain published in 1802, in which he writes: ‘if any one bank fails, a general run upon the neighbouring ones is apt to take place, which, if not checked in the beginning by pouring into the circulation a large quantity of gold, leads to very extensive mischief’.36 And even before that, in response to the first financial crisis in the United States - the panic of 1792 - Alexander Hamilton, then US Treasury Secretary, intervened to stem the crisis and in so doing was arguably the first person to discover the benefits of a LOLR.37He certainly was the first in a long line of US Treasury Secretaries who believed in bailouts.

When Bagehot wrote his seminal work, there were no Wimbledon Championships and banking was very different from the global industry we see today. His vision of a central bank playing the role of LOLR was inspired by the banking crises in London in 1825, 1836, 1847 and 1866. After the end of the Napoleonic Wars, Britain experienced a period of prosperity and the banks were sufficiently carried away that they invested in widely speculative ventures in Latin America, including the country of Poyais that turned out not to exist. In 1825, although many banks failed, the Bank of England, albeit belatedly, was able to stem the resulting panic; as described by one of my predecessors as Governor, Jeremiah Harman, it lent ‘by every possible means, and in modes that we had never adopted before … and we were not upon some occasions over nice; seeing the dreadful state in which the public were, we rendered every assistance in our power’.38 That description was later used by Bagehot to illustrate the proposition that liquidity support is not designed to save an individual bank but is carried out in the collective interest of the system as a whole. In doing that, the central bank may need to take strong and unpopular action. The Bank also issued around one million £1 notes, which helped to ease the crisis when gold ran short - an early example of the creation of ‘emergency money’.

Of particular significance to Bagehot was the failure of Overend, Gurney & Co., an erstwhile competitor of the Bank of England. On Thursday 10 May 1866 the bank announced that it would immediately suspend its activities following a severe run. As The Bankers’ Magazine put it at the time, the announcement generated ‘the greatest possible excitement in the City’.39 The Bank of England lent unprecedented sums to other banking houses, but it did not step in to prevent the failure of Overend, Gurney itself. For several years, there had been concerns about the health of Overend, Gurney and the reaction when it became a public company in 1865 had been decidedly mixed. But it was more than three years after its collapse when it emerged that the directors of the company, who stood trial on charges of fraud, had published a false prospectus concealing the fact that the firm had been, in essence, bankrupt before it went public. The bank had expanded out of its traditional business of short-term lending in the money market into activities closer to present-day investment banking, becoming an investor in railways and ships, among other things. Despite the scale of lending by the Bank of England, other banks failed and a recession followed.

From his vantage point as editor of The Economist, Bagehot observed the 1866 crisis and drew the conclusion that when faced with a sudden and large increase in the demand for liquidity by the public, in other words a bank run - a ‘panic’ - the responsibility of the central bank was to meet it: ‘a panic, in a word, is a species of neuralgia, and according to the rules of science you must not starve it’.40 One of the unique roles of central banks is the ability to create ‘liquidity’.41 Banks create money, but if people lose faith in banks, the ultimate form of money is that created by the central bank - provided it is backed by the tax-raising power of a solvent government. In Germany in October 1923, as the hyperinflation was nearing its peak, the government was close to insolvent with only 1 per cent of its expenditure financed by taxation. Commercial banks have accounts at the central bank, and in a crisis, the central bank can lend to them against the collateral of their assets. Bagehot’s doctrine was that in a crisis central banks should lend freely against the security of good collateral, at an interest rate above normal levels to ensure that the central bank was the lender of last resort not the lender of first resort. What Overend, Gurney revealed to Bagehot was that, although in a crisis it might be difficult to know whether a bank was solvent, it was nevertheless safe for the central bank to lend against good collateral.

His view has since become the conventional wisdom. So much so, that the phrase ‘lender of last resort’ is widely misused to refer to any action that deals with a financial crisis by dousing the fire with a massive injection of liquidity. It is used to urge the European Central Bank to lend to sovereign governments within the euro area, and to imply that the IMF should lend to any country in difficulty; I have even heard it used by sports teams in financial trouble who believe that the league in which they play should bail them out. The expressions ‘lender of last resort’ and ‘bailout’ have become synonymous. It is only a matter of time before there is a demand for a LOLR for the Bank of Mum and Dad. Bagehot’s argument was very different. In essence, the problem was that the banking system was an intermediary financing illiquid assets by promising instant liquidity to depositors. For the economy as a whole, the promise cannot be met. When enough depositors want their money back, the banking system cannot provide it. If this additional demand for liquidity is temporary, then the provision of emergency money by the central bank can tide the system over until the panic subsides. But if the assets have genuinely lost value, then the central bank must be careful not to subsidise insolvent undertakings. The problem is that in a world of radical uncertainty it is never clear whether a bank is solvent or insolvent, and in a crisis there is rarely time to find out. Actions by a LOLR can prevent a liquidity problem turning into a solvency problem, although not all solvency problems can be converted into liquidity problems by LOLR lending, as governments have painfully discovered in recent times.

Even in Bagehot’s time, however, his views attracted sceptics. Thomas Hankey, a former Governor of the Bank of England, published a book just one year after the failure of Overend, Gurney in which he recognised the alchemy of the banking system: ‘the mercantile and Banking community must be undeceived in the idea that promises to pay at a future date can be converted into an immediate payment without a supply of ready money adequate for that purpose’.42 If banks came to rely on the Bank of England to bail them out when in difficulty, then they would take excessive risks and abandon ‘sound principles of banking’.43 They would run down their liquid assets, relying instead on cheap central bank insurance - and that is exactly what happened before the recent crisis. The provision of insurance without a proper charge is an incentive to take excessive risks - in modern jargon, it creates ‘moral hazard’. Both Bagehot and Hankey were right, in their own way. Once a panic has started, the provision of liquidity to others can prevent widespread contagion. But the design of the LOLR mechanism must be thought through carefully beforehand in order to avoid incentives to excessive risk-taking. As time passed, it became easier to flood the system with liquidity when problems arose than to design a framework that would counter moral hazard.

If Bagehot’s ideas grew out of his study of earlier financial crises, can we too learn from financial episodes after 1866? Milton Friedman and Anna Schwartz, in their monumental study of the monetary history of the United States, laid the blame for the depth of the Great Depression on the Federal Reserve for failing to create sufficient money and act as a lender of last resort to the many banks that subsequently failed.44 During that period the money supply fell by around 30 per cent. The Fed was culpable for failing to prevent that contraction of the supply of money, rather than for failing to meet a sudden increase in the demand for liquidity. For lessons on the LOLR role, a more relevant historical episode is the financial crisis at the beginning of the First World War and the creation in unusual circumstances of emergency money.45

The centenary of the First World War witnessed a veritable blizzard of new books on that dreadful conflict - almost as many as on the recent financial crisis. Few people have drawn comparisons between the two episodes. But the outbreak of the First World War saw the biggest financial crisis in Europe, at least until the events of 2008, and an equally severe crisis in New York, albeit that the Great Depression was a bigger economic crisis in terms of its impact on output and unemployment. An examination of the crisis of 1914, and in particular of the difference between the outcome in London and New York, throws light not only on our recent experience but on financial crises more generally. After September 2007, when the latest banking crisis began, I often publicly compared current events with those of 1914. Yet I found that few people knew much about the financial crisis of 1914. Even the war memoirs of the Chancellor of the Exchequer at the time, David Lloyd George, devote only fourteen out of 2108 pages to the financial crisis he faced.

So what happened in 1914? Historians have long documented the prevailing disbelief in the likelihood of a European war. Among other things, the economic cost would simply be too high. Complacency among financial policy-makers showed in the failure of both governments and markets to take seriously the likelihood and economic consequences of such a war. Just two days before Britain declared war in August 1914, the Governor of the Bank of England, Lord Cunliffe, was lunching on the yacht of the wealthy and well-connected Clark family, moored off the west coast of Scotland. As Kenneth Clark, the art historian, wrote in his memoirs, ‘On the second he [Cunliffe] lunched with us on the yacht. I had fallen in love with his daughter, who had red hair and wore a monocle, and so was glad to be present. “There’s talk of a war,” said Lord Cunliffe, “but it won’t happen. The Germans haven’t got the credits.” I was much impressed.’46 John Maynard Keynes too was not immune to the mood of the moment when, on 24 June 1914, he wrote to the Treasury: ‘In a modern panic it is improbable that the big banks will come to grief.’47 Gaspar Ferrer, the key adviser to Lord Revelstoke, the Chairman of Barings Bank, said later: ‘The war came like a bolt from the blue.’48 Even after the assassination of Archduke Franz Ferdinand (heir to the Austro-Hungarian throne) in Sarajevo on 28 June 1914, there was barely a ripple in London markets. It was almost a month before financial markets woke up to the significance of the unfolding political events, and it was the ultimatum from Austria to Serbia (demanding that Serbia take draconian steps to suppress the expression of nationalist opinions) on 23 July that finally changed sentiment.

The next two weeks saw panic in markets and among banks. European stock markets fell sharply, and several were closed. There was a flight to safety, especially to cash, and liquidity dried up in all major markets, including those for foreign exchange, stocks and shares. Three-month interest rates more than doubled. At 10.15 a.m. on Friday 31 July the London Stock Exchange was closed in order to postpone settlement of transactions and thereby prevent a wave of failures among its members as prices plummeted.

That same day, lines formed in Threadneedle Street outside the Bank of England as depositors queued, as was their right, to convert deposits or notes into gold sovereigns, which commercial banks would not provide to them. As Keynes later put it, ‘the banks revived for a few days the old state, of which hardly a living Englishman had a memory, in which the man who had £50 in a stocking was better off than the man who had £50 in a bank’.49

Meanwhile, on the other side of the Atlantic, the position was no less precarious. Europeans had begun to sell their investments on Wall Street and convert the dollars they received into gold to bring back to Europe. The dollar fell sharply. What happened over the next few weeks, however, was to result in New York displacing London as the money centre of the world. Although the Federal Reserve System had been created by Congress in December 1913, nominations to the Federal Reserve Board stalled in the Senate Banking Committee, and it met for the first time only in August 1914. So although William McAdoo, as Treasury Secretary (and, coincidentally, the son-in-law of President Woodrow Wilson), was also the first Chairman of the Board, the Fed did not enter the playing field until after the financial crisis had come and gone. As a result, in 1914 the major decisions in dealing with the crisis were taken not by central banks but by the respective finance ministers - Treasury Secretary McAdoo and Chancellor of the Exchequer Lloyd George.

Despite falls in stock prices of around 10 per cent earlier in the week, a meeting of bankers on the evening of Thursday 30 July had seen no reason for closing the New York Stock Exchange. But McAdoo intervened and, worried that if New York remained the only open exchange European investors would take the opportunity to sell and repatriate gold, ordered the exchange to close on Friday 31 July, a matter of hours after the closure of the London Stock Exchange.50 As it was, there were substantial outflows of gold from New York to Europe during that final week of July. The start of the First World War saw the end of the period during which most countries had fixed the price of their currency in terms of gold (and hence to each other) - the gold standard. In Europe the demands of wartime finance led governments to suspend the convertibility of paper into gold and conserve their holdings of bullion. If the United States could retain its fixed rate between the dollar and gold then it could aspire to be the world’s financial leader; in 1914 the British pound sterling and not the US dollar was the safe haven currency. From McAdoo’s perspective, it was no time to allow large withdrawals of gold that might force America off the gold standard.

Back in London, on the following day, Saturday 1 August, Basil Blackett, then a Treasury civil servant, wrote to John Maynard Keynes: ‘I tried to get hold of you yesterday and today, but found you were not in town. I wanted to pick your brains for your country’s benefit and thought you might enjoy the process. If by any chance you could spare time to see me on Monday I should be grateful, but I fear the decisions will all have been taken by then.’51Keynes lost no time. On Sunday he rode down from Cambridge in the sidecar of his brother-in-law’s motorcycle and went straight to the Treasury.

While Keynes and Blackett were conferring in Whitehall, McAdoo left Washington by train to meet with more than twenty senior bankers in the Vanderbilt Hotel in New York. All those present were desperate to avoid a repetition of the panic in 1907 when a run on the Knickerbocker Trust Company led to the suspension of cash withdrawals. During that earlier crisis, John Pierpoint Morgan had added to his fame by organising a private consortium of banks to lend to banks under suspicion, so averting a major collapse of the banking system, though other banks did suspend payments and there was a sharp contraction of the US economy. From that experience came the impetus to create the Federal Reserve System, which would be able to lend to banks that were temporarily short of funds - to act as a lender of last resort - obviating the need for a Morgan or similar to organise a private consortium to prevent a banking failure. But with the Federal Reserve not yet in operation, what could McAdoo offer the bankers in the Vanderbilt?

Out of the 1907 crisis came another solution to the problems of 1914. The Aldrich-Vreeland Act of 1908 permitted banks to deposit government bonds or short-term paper issued by American companies with the US Treasury and receive ‘emergency notes’ in return. Emergency banknotes, embossed with each bank’s own name and logo, worth $500 million were printed in advance and stored with the Treasury in a new underground vault. Here was a source of emergency money that could be distributed to banks in exchange for collateral without the need for a central bank. There was a limit on the value of the notes that could be obtained of 90 per cent of the value of the bonds and 75 per cent of the value of commercial paper deposited by the banks with the US Treasury. And there was a tax on the value of the emergency notes drawn.

At his meeting with bankers on Sunday 2 August 1914, McAdoo found men in urgent need of emergency currency, and plenty of it. The emergency money began arriving in New York on Monday 3 August, and the printing presses operated around the clock to print additional money. The fact that the Bank of England was not immune from the possibility of a run by depositors energised the American financial community to support the creation of this emergency money. In contrast with 1907, when the money supply fell by over 10 per cent, in 1914 the creation of emergency money allowed the money supply to rise at an annual rate of around 10 per cent. Demand for emergency money peaked at the end of October 1914 and fell gradually, disappearing altogether by the middle of 1915. Despite the absence of any help from the new Federal Reserve Board, not yet up and running, McAdoo had shown how a government could act as a lender of last resort.

In London, Monday 3 August was, fortunately perhaps, a bank holiday. Britain declared war on Tuesday 4 August. The bank holiday was extended by an additional three days. During that first week of August, a series of extraordinary measures was introduced by the Treasury and the Bank of England. The UK government decided to intervene and use taxpayers’ money on an unprecedented scale in a mission to ‘save the City’. The problem was that London had made, as was then usual, large short-term loans by ‘accepting’ or guaranteeing - for a price - loans to borrowers on the continent of Europe. Such guarantees could be traded and were known as bills of exchange. In normal times those bills could be bought and sold in the London market. As Lloyd George put it, ‘the crackle of a bill on London with the signature of one of the great accepting houses was as good as the ring of gold in any port throughout the civilised world’.52 With the onset of war in Europe, providing guarantees had become risky. London banks had underwritten loans that were not being repaid when due and might never be. As Lloyd George went on, ‘when the delicate financial cobweb was likely to be torn into shreds by the rude hand of war, London was inevitably thrown into panic’.53

The first measure - on the Tuesday - was emergency legislation to impose a moratorium on all London bills of exchange for one month. Three days later this became a general moratorium. Debts, except for wages, and taxes and debts owed by foreigners, could not be enforced. The legislation provided that, if necessary, the moratorium extended to bank deposits. This provided temporary relief but did not tackle the underlying solvency problem.

Unfortunately, unlike in 1825, the Bank did not have sufficient stores in its vaults to meet the sudden increased demand for low denomination notes as gold coin was conserved to rebuild the nation’s gold reserves. So the second measure to be taken - on the Thursday - was the passage in one day of the Currency and Bank Notes Act, allowing the Treasury to print special £1 notes, to a much lower standard - in the interests of speed - than the Bank of England would have accepted. This amounted to temporary removal of the limits on the fiduciary note issue of the Bank of England, and required the suspension of the Bank Charter Act of 1844, as had previously happened in 1847, 1857 and 1866. Britain had not learned from the US experience in 1907 and had printed no store of emergency money to distribute in a crisis.

Third, on Friday 7 August the government decided that the bank holiday should end and the banks reopened. The crisis had been contained, if not solved. As Keynes put it later:

In the dark and uncertain days, which seemed to divide by an interminable period the last Thursday of July from the first Thursday of August, the City was like a very sick man, dazed and feverish, called in to prescribe for his own case. Its great houses, suspecting the worst, could not then gauge exactly how ill they really were; and the leaders of the City were many of them too much overwhelmed by the dangers, to which they saw their own fortunes and good name exposed, to have much wits left for the public interest and safety.54

In the weeks that followed, measures were taken to deal with the underlying insolvency problem in London. In wartime conditions, debtors on the Continent would be unable to repay London banks, discount houses, and other institutions that had accepted bills, and hence many of the assets held by British financial houses would lose their value, leaving those institutions insolvent. As a contemporary author described the position: ‘the banking system, was, to put it quite bluntly - “bust”. They could not pay what they owed. They had not the money. The outbreak of the War at once revealed the hopeless make-believe of the whole pass-the-buck debt-generating process.’55 Over five months, Chancellor of the Exchequer Lloyd George and Treasury officials recapitalised the City. The Bank of England purchased a large volume of bills amounting to around 20 per cent of total assets held on bank and other financial balance sheets in London, all potential losses being indemnified by the government on behalf of the taxpayer. The Bank bought one-third of the entire stock of bills, amounting to some 5.3 per cent of GDP. As Lloyd George admitted in his memoirs, by offering the guarantee the government had ‘temporarily assumed immense liabilities’.56 He took the risk of losses on the assets he guaranteed without seeking any compensation. It was a gamble that could have been taken only in wartime.

It worked. The City was saved. When the recapitalisation was complete, the Stock Exchange was able to reopen on Monday 4 January 1915, and the assembled financiers sang all three verses of the National Anthem. The financial crisis ‘was over’.

Across the Atlantic, McAdoo’s problem was different. It was widely understood that the conflict in Europe would lead to higher prices and volumes of exports of American commodities. The Great War strengthened the US economy. The US banking system faced no solvency problem. But to reopen the Stock Exchange would lead to a resumption of sales of stocks by Europeans anxious to convert dollars into gold, which could then be shipped back to Europe. The solution was to increase European demand for dollars. And the way to do that was to meet the increased European demand for US exports arising from war needs.

To supply that demand more ships were required, and two factors ensured their availability. The first was the creation of the US Bureau of War Risk Insurance in August 1914 to insure American-registered vessels against war damage. The second was the unexpected announcement in October that Britain would not regard cargoes of cotton (essential for the production of explosives) destined for Germany as contraband and they would therefore not be at risk of seizure. Exports of cotton, and agricultural commodities more generally, grew rapidly. The New York Stock Exchange reopened on 15 December 1914. The link between the dollar and gold was maintained. The US government, unlike its British counterpart, had no need to assume ‘immense liabilities’ to solve the challenges to its financial system. Within a matter of months the dollar had begun its inexorable rise to become the dominant international currency and in due course the United States would replace Britain as the world’s leading financial power.

Two broad lessons emerge from the experience of 1914. The first is that the key function of the monetary authorities, whether as government or central bank, is to determine the supply of money in both good times and bad. In 1914 McAdoo had the advantage that the crisis of 1907 had alerted the US authorities to the need to prepare a stock of emergency money, whereas the British had forgotten the events of 1825. It was another example of a law that I saw on many occasions - countries learn only from their own mistakes and not from those of others (the failure of the UK to develop a resolution regime for failing banks despite US experience in the 1930s was another). The second is that a crisis will not be resolved by the provision of liquidity if there is also an underlying solvency problem; in other words, a shortage of capital available to absorb losses and prevent default. In 1914, London had a solvency issue and New York did not. In 2008, the turning point of the crisis was when governments were eventually persuaded that the banking system was suffering from a shortage not just of liquidity but of capital, and had to be recapitalised, forcibly if necessary.57

The concept of ‘emergency money’ is important. It captures the need for a sudden increase in money when there is a sharp rise in the demand for liquidity. A jump in the demand for liquidity can arise for many reasons, including a loss of confidence in the public sector itself. Not long after McAdoo was sending banknotes to New York, local and regional governments in Germany began printing new banknotes: Notgeld, the German word for emergency money. In Germany (and also in France and Belgium) the war had led to a sharp increase in the demand for, and hence the value of, metal. Coins made out of metal disappeared because their face value as minted coinage was less than their value to producers of armaments. Soon there was a big shortage of small change. In response, and in the absence of a centralised solution, local and regional communities began to print low-denomination banknotes. The temptation to use Notgeld to promote their local communities proved irresistible, and ‘regional memories and loyalties revived, and found an exuberant, colourful expression of local identity and civic pride’.58 The designs included serious scenes, featuring people such as Luther and Goethe, as well as amusing stories of life at the seaside. In the end, no central bank can act as a LOLR unless there is confidence in the government that underwrites it.

Banks today are very different from banks in Bagehot’s time or during most of the twentieth century. They are much bigger, their assets are more complex and difficult to value, they hold far fewer liquid assets, they finance themselves with far less equity capital, and they wield greater political power. As a result, the maxim ‘lend freely against good collateral at a penalty rate’ is outdated. This creates two problems for the LOLR role envisaged by Bagehot.

First, the definition of ‘good collateral’. Both Alexander Hamilton and Walter Bagehot knew that the authorities could lend against the security of government stock. They could not know whether the banks to which they might consider lending were solvent or not, but that did not matter if they could lend against good collateral. Until well into the post-war period, banks held around 30 per cent of their assets in the form of government securities, most of them short-term and all of them liquid. In that world, the central bank could lend $100 against the value of stock of $100. Today, however, bank assets comprise largely illiquid and often unmarketable assets such as loans or complex financial instruments. Such illiquid assets can be turned into good collateral if the central bank lends only a proportion of the value of those assets. If a commercial bank wants to borrow $100 from the central bank, it will therefore have to provide collateral worth more than $100, so that if for some reason it cannot repay the loan there is a suitable margin to provide confidence to the central bank that it will be able to sell those assets for at least the value of the loan. The difference between the amount lent by the central bank and the value of the collateral offered in return is described as the ‘haircut’ on that type of collateral. In practice, haircuts range from only a percentage point or two, when the collateral takes the form of highly liquid financial assets such as government bonds, to very high haircuts, of 50 per cent or more, in the case of individual loans on which little information is available.

When the European Central Bank lent several hundred billion euros to banks all over Europe in December 2011 and February 2012, many thought that the crisis in the euro area was solved. They were soon to be disillusioned. Although the ECB was willing to lend for three years at a very low interest rate, it still expected banks to repay those loans. A bank that had too little capital to absorb likely future losses was still too risky to attract funding from the private market. So unless the ECB was to provide a permanent source of funding, European banks remained an accident waiting to happen. Moreover, like any central bank, the ECB would lend only against good collateral. What that meant in practice was that banks had to provide the ECB with collateral in the form of claims on their loans before the ECB would provide them with cash. Many of those loans were far from ‘good’. So the ECB would lend only a proportion of the estimated value of the loans surrendered by banks. In the operation in February 2012, the haircuts on some of those loans amounted to 60 per cent of their face value. In other words, banks surrendering such loans would receive just 40 per cent of their face value in cash. Only in that way could the ECB be confident that the collateral they received would be sufficient to compensate for the failure of a bank to repay its loan.

But that created a serious problem for those in the market, such as pension funds and insurance companies, who were thinking of lending to banks. For as the ECB increased its lending to banks, so a larger and larger proportion of the assets of those banks were encumbered by claims on them in the name of the ECB. The proportion of a bank’s assets available to act as collateral for debt provided by the market diminished. So the cost of unsecured funding - loans made without the security of collateral - to banks started to rise. Banks depend on financing a sufficiently high proportion of their balance sheet by unsecured funding, such as deposits or securities issued to pension funds and insurance companies, to enable them to function. Any attempt to fund the whole balance sheet in secured form - borrowing against collateral - could work only if the haircut required by those lending to the bank was literally zero.

Consider the simple hypothetical example of a bank with $100 million in total assets, financed by $90 million of short-term deposits and $10 million of equity capital. Suppose that for some reason $30 million of deposits are withdrawn, and the bank turns to the central bank for temporary liquidity support. In order to protect itself and, ultimately, taxpayers, the central bank decides that the appropriate haircut on pledged collateral is 25 per cent. The bank must then pledge $40 million of assets as collateral to the central bank so that after the haircut it will receive cash loans of $30 million. The bank has maintained its funding but the remaining depositors, who have a prior claim over shareholders, can see that there are only $60 million of assets left to support $60 million of deposits, whereas before there were $100 million of assets to support $90 million of deposits. And the effect of the haircut is to impose an upper limit of $75 million on the LOLR assistance that can be made available.

When the Bank of England lent to Northern Rock in 2007, it was possible to predict when the LOLR assistance would reach its maximum limit. The limit was duly reached on the date predicted and the government had to take over the financing of the bank and the associated credit risk. The more a central bank lends, the lower the proportion of assets to deposits available to support loans from the private sector. A LOLR supplies temporary funding but at the expense of increasing the incentive for a run by the private sector depositors or short-term creditors. In extreme cases, the LOLR is the Judas kiss for banks forced to turn to the central bank for support.

The second problem with the LOLR role is that banks may be reluctant to accept assistance because of the implied stigma from the revelation that they are in need of liquidity support. The information revealed to the market by the decision to accept central bank liquidity may damage the ability of the bank to obtain funding from elsewhere - partly because of the information itself and partly because of the Judas kiss effect in reducing available collateral - and so make it reluctant to turn to the central bank. This turned out to be a major problem in 2007 and 2008, when attempts to provide liquidity led to great caution on the part of banks reluctant to give a signal that they were in need of support. As a result, central banks created auction facilities in which banks could bid anonymously for liquidity.

But the LOLR must be ready to supply liquidity at any moment, and auctions cannot be organised on a continuous basis. At the height of the crisis, weekly auctions played a useful purpose, but they still needed to be supplemented by direct LOLR when banks experienced an urgent need for additional liquidity. In an effort to overcome the problem of stigma, central banks have traditionally been reluctant to publish details of institutions benefiting from access to special lending until the need for such support has come to an end. Delayed reporting of access by banks to central bank facilities helps to reduce the stigma problem. For example, in 2008, the details of the support for Royal Bank of Scotland and HBoS, although approved by the Chancellor of the Exchequer, were not revealed publicly until many months later when there was no longer any need for such secrecy. Legislators, however, are keen to impose greater disclosure requirements on such facilities. Details of those accessing the central bank ‘discount window’, at which banks can exchange collateral for money, are published, if at all, only after some delay. In the United States, Congress has mandated the publication of the names of all borrowers at the Federal Reserve discount window no later than two years after they borrow.

The stigma problem is not new. In 1914, there were concerns ‘about the potential reputational damage of borrowing from the Bank of England’ if it became known to others that advantage had been taken of such a facility.59Equally, the creation of emergency money in the United States under the Aldrich-Vreeland Act of 1908 was an opportunity not used by banks until the collective crisis of July 1914. No such emergency money had ever been demanded by a bank until the events of the summer of 1914. As the Comptroller of the Currency warned when the Act was passed, ‘the issue of so-called emergency notes … would at once be a confession of weakness and a danger signal that no bank would dare make until in desperate condition’.60 And, as the financial historian William Silber later wrote, ‘Emergency currency lost its stigma during the first week of August as the Great War threatened major banks throughout the country.’61 There is no simple or, for that matter, complex solution to this problem, which continues to trouble the designers of central bank liquidity facilities. The importance of stigma is that over the course of history, the reluctance of banks to be seen to accept support from the central bank has probably meant that the size of LOLR operations has probably been too little and too late, exacerbating the severity of crises. Central banks cannot afford to be ‘over nice’.

Aware of the importance of ‘lending freely’ in a crisis, and the problems caused by the need to lend against good collateral and at a penalty rate, central banks and governments have relaxed the conditions of a traditional Bagehotian lender of last resort and turned to bailouts. This creates risks to taxpayers and incentives for banks to expand the alchemy of the system. The solution is to convert the crisis function of the lender of last resort into a regime that determines the amount a bank can borrow in both good times and bad, and ensures that a bank will have posted sufficient collateral in good times that it will have access to liquidity in bad times adequate to meet the demands of short-term creditors. I will explain how this works in Chapter 7.

The future of central banks

Despite some ups and downs, central banks are starting this century well ahead of where they were a century ago. There are more of them, and they have greater power and influence. But has their reputation peaked? Will future historians look back on central banks as a phenomenon largely of the twentieth century? Although central banks have matured, they have not yet reached old age. But their extinction cannot be ruled out altogether. Societies have managed without central banks in the past.

Before the crisis, central banking seemed rather simple. There was a single objective - price stability - and a successful framework within which to make decisions on interest rates - inflation targeting. It seemed a successful coping strategy. Communication became more important, and central banks moved from mystery and mystique to transparency and openness. During the crisis, however, many of those assumptions were challenged, as we learned how inadequate our understanding of the economy and the financial system was. There is more to managing the economy than hitting a target for consumer price inflation. Most of the models used by central banks to forecast the economy proved deficient in explaining the disequilibrium in their own and the world economy, as described in Chapter 1. As a result, policy-makers failed to realise that the period of low inflation and steady growth during the Great Stability was unsustainable, and would probably come to an end with a crash of some sort.

Central banks have a role to play in changing the heuristic used by households and businesses when they see a serious disequilibrium building up. In such circumstances, what is required is a clear and convincing explanation of why it may be desirable to allow inflation to run above or below target for a period in order to restore a sustainable path for the economy. It would be a big mistake to jettison inflation targeting altogether. It is a valuable heuristic for central banks, provided there is room to deviate when circumstances demand.

The crisis also brought home the importance of the framework for liquidity provision to the banking system. Because both banking and central banks have changed out of all recognition since Bagehot wrote his book, it is time to reassess the old doctrines. The Bagehotian lender of last resort is a concept in need of reform. A number of new ideas and instruments were developed on the hoof during the crisis and undoubtedly some will persist. But what has been missing is an integrated single framework within which to analyse the provision of money by central banks in both good and bad times. Such a framework is the key to ending the alchemy of our monetary and financial system, and I shall return to this question in Chapter 7.

Despite those problems, the surprising outcome has been that central banks have been handed greater responsibilities than before the crisis. The Bank of England I left behind was twice as big as the one I inherited. There is a risk in expecting more from central banks than they can, in fact, deliver. Some people seem to believe that central banks are the answer to all of our economic problems - the ‘only game in town’. Any central bank that allows itself to be described as the ‘only game in town’ would be well advised to get out of town. In the end, expecting too much from central banks will produce disillusionment with the central bank independence that played such an important role in the conquering of inflation. With careful design of the mandate for central banks they can continue to be one of the three great inventions since the beginning of time.

Appreciation of central banks’ actions was well captured by a pamphleteer in the middle of the eighteenth century:

There certainly never was a body of men that has contributed more to the Publick Safety and Emolument than the Bank of England, and yet even this great, this useful Company, has not escaped the invectives of malicious tongues … This flourishing and opulent Company has upon every emergency always cheerfully and readily supplied the necessities of the Nation … and it may very truly be said that they have in many critical and important conjunctures relieved this Nation out of the greatest difficulties, if not absolutely saved it from ruin.62

The popularity of central banks has waxed and waned over the past couple of centuries. One point, however, stands out from that history. The freedom of a central bank to act in a crisis depends on its legitimacy. In turn that requires a clear mandate providing the assurance for the legislature to delegate powers to an independent central bank. Democratic legitimacy has been built up over the years, in part through greater transparency and accountability, and works at national level. It is far from clear, however, that any democratic mandate can function at a supranational level. The attempt to break the link between money and nations has always been fraught with difficulty.