More Information

Information about the Australian Economic Review is available from

More information about the 2009 David Finch lecture is available from


The following is the unedited speech given by Sir Howard Davies, Director of the London School of Economics, at the 2009 David Finch Lecture.  Now in its twelth year, the annual Faculty of Economics and Commerce lecture focuses on the subject of evolving practices and principles and related issues pertaining to international monetary, financial, fiscal and trade integration and cooperation.  A final revised copy of the text will be available in the December 2009 issue of the Australian Economic Review. 

Pricking Bubbles in the Wind: could central banks have done more to head off the financial crisis?

I am honoured to have been invited to deliver the 2009 David Finch lecture. Finch was a very distinguished Australian, and  he was of course a graduate of the LSE. We share many common interests in international finance and in higher education. I do not think we ever met, though we were certainly in the same building at the same time. He led the famous IMF mission to the UK in 1976, when I was working in the Treasury, but too junior to be involved in discussions with the receivers. If he were around today his experience then would be relevant. We may be receiving another IMF visitation before too long.

David Finch’s general perspective on the current crisis would have indeed be very valuable, with his academic detachment allied to a first-hand understanding of the complexity of global policy-making at the IMF. I will try my best to match that combination this evening, though in a different way, my own policy-making experience is largely domestic.

The question I will explore tonight may seem a little obscure to the man on the Carlton omnibus. But it is the one which in central banking circles today generates the most heat, partly because there are genuine disagreements on how best to calibrate monetary policy, but perhaps also  because the answer determines the degree of culpability which attaches to central banks for the crisis which continues to engulf the world, financial markets and the global economy.

Australia, living up once more to its ‘lucky country’ designation, is less badly affected than most. Your economy has not contracted anything like as severely as the UK’s, and your banks are in better shape – no doubt due to the careful supervision exercised by APRA’s LSE educated boss John Laker. Yet, in spite of a large fiscal stimulus,  even here unemployment is rising, and the prospects for growth in the coming years are subdued, at best.

Globally, the latest estimates suggest that the cost of the financial crisis will exceed $9 trillion, and unemployment will rise by tens of millions. Forecasters expect 3 million out of work in the UK by the end of next year. Estimates in Australia suggest a rise to over 8 per cent.  In Spain joblessness is close to 20 per cent. For the British government this is already the 4th largest fiscal shock in history, after the Napoleonic War and the two World Wars. In the US, the position is even more stark.

In these circumstances it is not surprising that a blame game is well under way. The famous ‘special relationship’ between the US and UK has come under strain, with Gordon Brown regularly claiming that the crisis was, like Bruce Springsteen, ‘Born in the USA!’. The French lump the two of us together and see the crisis as one of ‘Anglo-Saxon Capitalism!’. In the US itself a Boston Globe survey showed that 32 per cent of Democrats blamed the Jews. No doubt the Greeks blame the Turks, and vice-versa. Australians will no doubt also be true to type and blame the umpires.

In the financial world politicians blame the regulators for being asleep at the wheel. Regulators blame the banks, the commercial  banks blame the investment banks, and the accounting standard-setters for requiring them to reveal losses they would rather have kept quiet, or ‘smoothed’ as they prefer to put it. The investment banks blame the hedge funds for shorting their stock and the private equity houses for persuading them to lend as highly leveraged deals.  And everyone anathematizes the rating agencies. They are the guys we all love to hate.

In a crisis of this magnitude, there is plenty of blame to go round. Even academics have come in for their share, for continuing to teach efficient markets theory in finance classes, and focusing too much on growth theory in economics. One of our own economics professors has written a lengthy mea culpa on behalf of the profession, entitled ‘‘the unfortunate  uselessness of most monetary economics!” That went down well in the common room, as you can imagine. Paul Krugman made similar points in a lecture at the LSE last month, describing the recent period as a ‘Dark Age’ for the profession.

So is there any opprobrium left for central banks? Well, I think we may be able to find a little, so that they do not feel neglected. It’s far worse to be ignored than to be criticised, as we all know.

What charge may be laid at their door? Charlie Bean, Deputy Governor of the Bank of England, hinted at it recently when, having listed all the manifold sins of the banks and their regulators which ingnited the conflagration , he noted that ‘you need fuel to make a fire, too. And that was provided by the ex ante excess supply of global savings over investment, which pushed real interest rates on safe assets to historically low levels, reinforced by loose monetary policy.’

Others put forward a  strong-form version of that critique, and maintain firmly that the crisis was “made in the Fed”. Steve Roach, the former chief economist of Morgan Stanley and a longstanding critic of Alan Greenspan, does not mince his words. “Central banks”, he argues, “were asleep at the switch. Central banks have failed to provide a stable underpinning to world financial markets and to an increasingly asset-dependent global economy…it is high time for monetary authorities to adopt new procedures- namely taking asset markets into explicit consideration when framing policy options…As the increasing prevalence of bubbles indicates, a failure to recognize the interplay between the state of asset markets and the real economy is an egregious policy error”.

Roach’s call for a shift away from “one-dimensional fixation on CPI-based inflation” has been reinforced by arguments that the CPI has been giving false readings. Competition from China held down prices of traded goods, while leading participants in financial markets grew incontinently rich, recycling excess liquidity created by central banks, misled into a belief that inflation had been conquered and that a productivity revolution was under way. The critics consider that this reading of inflation was  fundamentally wrong. The CPI itself was really “Chinese Price Inflation”,  held artificially low by the Chinese export boom. In fact during the so-called Great Moderation a massive expansion of credit was under way, leading in turn to mispricing of risk and asset price bubbles, all under the noses of central bankers myopically monitoring their narrowly defined inflation objectives.

Indeed the period demonstrates an interesting asymmetry. Central banks used imported low, or even negative inflation to declare victory in hitting their inflation targets while running a relatively loose monetary policy, as measured by the low level of real interest rates, for example, or by the Taylor rule, for much of the 2000s. [The Taylor rule specifies by how much a central bank should change interest rates in response to the divergence of actual from potential GDP, and actual from target inflation. Specifically, it says that the interest rate should be one and a half times the inflation rate, plus half the GDP gap, plus one]. But when commodity prices rose they argued that it was necessary to exceed the target for a period.

 John Taylor himself points out that interest rates in the US were, from the end of 2001, held significantly below the level which his rule would have indicated. Indeed by 2004 rates were fully three percentage points lower. He argues that this deliberately loose monetary policy was the direct cause of the house price boom and subsequent bust. He further shows that European countries where the deviation from the rule was greatest experienced similar house price bubbles.

Less aggressive critics than Roach and Taylor acknowledge that the judgements central bankers made from 2001 to 2007 were defensible, and that it is unreasonable to expect them to manage asset prices as they control consumer price inflation,  but nonetheless argue that they could have done more than they did to moderate the massive escalation in asset prices and credit expansion which preceded the crash of 2007. Real interest rates were very low indeed. Sushil Wadwhani, who was an external  member of the Bank of England’s MPC for three years , believes his colleagues were captives of an outdated, almost religious belief in efficient markets.“They would rather carry out inflation forecast-targeting policy on the assumption that financial markets are efficient and there are no bubbles”.

Robert Shiller has advanced a related critique, pointing to the dramatic moves in asset prices in the early years of the century, which were quite unprecedented, and impossible to justify by reference to supply and demand, in the case of housing, or earnings in the case of equity prices. He points to a version of what he calls “group think” which blinded central banks to the unsustainable nature of the trends observed.

At the time, most central bank governors did not accept the argument that asset prices were giving dangerous signals. It would be wrong to suggest, though, that there was one single view within the central banking community. Bill White, the chief economist at the BIS, also argued for a greater focus on credit expansion and asset prices, and did so well before the crisis hit. Surely, he argued, there was a point at which it was possible to identify asset mispricing and bubbles?  Why could interest rate policy not take some account of the risks posed by escalating asset prices, just as it did with other risks to inflation and growth? BIS economists became closely identified with the proposal that the monetary authorities, even those with a tight inflation objective focused on retail prices, should have been prepared to “lean against the wind” of asset price escalation.
White went on to argue that there were stronger grounds, even than in earlier periods, for looking at financial sector developments and their potential to threaten rapid and sustainable output growth, as new indicators which ought to help guide the conduct of monetary policy.  At the same time, there was evidence  that the inflation process had changed markedly in recent years with global measures of capacity utilization exerting a significant degree of influence on domestic inflation.

White concluded that while stable prices bring many benefits, they do not exclude financial sector shocks and that, just as there was a willingness to tolerate the first round effects of negative supply shocks on inflation, there should perhaps be a willingness to tolerate deflation arising from positive supply shocks.  Having accepted that the status quo had delivered a period of rapid growth and low inflation – the so-called “Great Moderation” - White argued that, with a monetary policy focussed on price stability, the endemic procyclical characteristics of the financial system, would only meet with resistance to the extent that they triggered inflationary pressures.  He thought that responding to the subsequent downturn following the bursting of a credit bubble through asymmetrically easier monetary policies was an inadequate response.

Andrew Crockett, then the General Manager of the BIS,  put the argument very clearly in 2003:

            “In a monetary regime in which the central bank’s operational objective is expressed exclusively in terms of short-term inflation, there may be insufficient protection against the build-up of financial imbalances that lies at the root of much of the financial instability we observe. This could be so if the focus on short-term inflation control meant that the authorities did not tighten monetary policy sufficiently pre-emptively to lean against excessive credit expansion and asset price increases. In jargon, if the monetary policy reaction function does not incorporate financial imbalances, the monetary anchor may fail to deliver financial stability. “

And there is no doubt that credit expansion was rapid, especially household credit. The UK and the US observed a dramatic boom, but hats off – or should I say floppy greens off – to the Aussies on this measure. In the UK, mortgage debt went from 50 to 85 per cent of GDP in less then a decade.

The central bank response to this heterodoxy, both before and during the crisis, was robust. Alan Greenspan himself led the counterattack. He challenged every link in the chain of argument. In his view it was not possible to identify when a bubble was inflating, and even if it were possible so to do, a monetary response would be ineffective –“ The notion that a well-timed incremental tightening could have been calibrated to prevent the late 1990s bubble is almost surely an illusion”.  He argued that it would almost certainly be undesirable to attempt to respond, in a way which might constrain markets and hinder the processes of innovation. Instead, central banks should forget about preventive measures and “focus on policies to mitigate the fall-out when it occurs and, hopefully (sic), ease the transition to the next expansion”.  While liberalised capital markets may well be prone to booms and busts (and there is evidence that the incidence of busts is growing) the net result remains positive in terms of economic welfare. Attempting to interfere directly in the process of creative destruction which accompanies these intermittent crises would be unwise and counterproductive. 

These contrasting points of view seem to admit little possibility of accommodation.  Yet there are more recent signs that central banks, faced with the massive value destruction of the last two years, are becoming more pensive about their record. Some are now taking a less strident line. If the bust is so dramatic, bringing economic and social destruction on a massive scale should we not consider whether we might not have done more to avert it?

It is therefore worth picking through the details of the dispute to see if the outline of a new and potentially more effective approach might emerge, one which does not compromise the success achieved in anti-inflation policy, but gives somewhat greater weight to financial stability.

To do so it is useful to parse the argument into a series of issues on which different points of view are advanced. On the first two, the germs of a revised consensus can be identified. Divergences open up on the later questions.

            Q. 1. Should central banks target asset prices?

Here there is a large measure of agreement. The locus classicus of the case for taking more account of asset prices in the setting of monetary policy, by Cecchetti et al, says “It is important to emphasise that… we are recommending that while [central banks] might react to asset price misalignments, they must not target them”. So the argument is not about adjusting the definition of inflation on an ad hoc basis as asset prices fluctuate, it is about how far decision-makers should take account of asset price misalignments in setting interest rates, and in determining the appropriate inflation forecast horizon. Advocates of what is now called ‘leaning against the wind’ believe that central banks seeking to smooth output and inflation can do so more successfully if they set rates with an eye towards asset prices in general, and misalignments in particular.

            Q. 2. Should the measure of inflation targeted include an element of asset price, and particularly house price inflation?

This is a crucial question as, if asset price changes are not incorporated in the measure of inflation which the authorities are enjoined to stabilise, they may express well-founded anxieties about asset price inflation from time to time, but lack a framework within which to respond to them in an effective way. 

The current definition of inflation used in the UK and in EMU excludes any element of housing costs. In the UK the target rate was changed from the RPI, which did include an element of imputed house rental, to the CPI, on the model of the Harmonised Index of Consumer Prices used in the euro area, in 2003.

Since then, the Bank of England has changed tack, and the Governor has now accepted that it would be preferable to change to a measure which did incorporate an element of housing costs. The question is in the hands of, or perhaps in the long grass at Eurostat, as it has been for some years. Mervyn King himself has expressed public frustration with the lack of urgency displayed.

So there appears to be an emerging consensus on this point. But just how sizeable an element of housing costs should be incorporated in  the target rate is likely to prove far more controversial. In the US, the index includes an estimate of the price of owner-occupation based on a survey of rental costs. Stephen Cecchetti has calculated the impact on US inflation were that element to be replaced by an index of home sales prices.  The effect is dramatic. Over five years from 2000 recomputed inflation would have been three quarters of a per cent a year higher than on the CPI index used. In the UK, the effect would have been even greater. The recomputed RPI including house price inflation would have been between 2 and 4 per cent a year higher over the last decade.

It is unlikely that a recomputation on that scale is what King has in mind. And one might question how useful it would be to respecify the index in this way. Would the Bank of England have kept interest rates significantly higher for some time on these grounds? Only if they were persuaded that there was indeed a serious misalignment. And on that question opinion is divided. So while some readjustment of the index might be helpful as a signalling mechanism, it is highly unlikely that the adjustment would be anything like as dramatic as Cecchetti’s calculations imply.

The third question is where things get more difficult.

            Q. 3. Is it possible to identify serious asset price misalignments, and are they of legitimate concern to monetary policy-makers?

Unless it is possible to know with some degree of confidence that asset prices are in unsustainable territory, it is not possible to know how, if at all, to respond to them.

Some say that it is quite impossible to know, ex ante, whether or not a bubble has inflated. Ben Bernanke, before he joined the Federal Reserve, was a sceptic, noting that “Advocates of bubbles would probably be forced to admit that it is difficult or impossible to identify any particular episode conclusively as a bubble, even after the fact”. He went on, nonetheless, to acknowledge that “episodes of irrational exuberance in financial markets are certainly a logical possibility, and one about which at least some central bankers are evidently concerned”. Greenspan’s position on this point is curious. When he made his celebrated comment about irrational exuberance in December 1996 he appeared to have reason to believe that prices (at that time the principal concern was with the level of the Dow Jones Index) were misaligned. Perhaps the subsequent experience – even after the falls of 2008 the Index is still higher than it was when Greenspan issued his gypsy’s warning – caused him to rethink. He now argues that it is impossible know when misalignments occur, and that central banks should limit themselves to a focus on policies which mitigate the fallout when it occurs.

Others strongly dispute the argument that misalignments cannot be identified. Cecchetti argues that there are long-run measures which can help to identify mispricing. In the equity markets extravagant price-earnings ratios which could not be explained by changes in dividend policies, or a fall in the equity risk premium, were a powerful leading indicator of trouble in the dot com boom. More recently a dramatic fall in the risk premium on high yield bonds, which ran for a couple of years at something close to half its long-run average, was a strong sign of mispricing there. In the case of housing there are price/earnings ratios, and perhaps more importantly price/rental income ratios, which point to the likelihood of downward shifts. The growth of credit aggregates may also be helpful in identifying unsustainable asset price increases. These indicators cannot be used as automatic triggers, and misalignments may persist for some time (as they did in the US and UK housing markets). But, as Wadwhani contends, the uncertainties are no greater than in many other areas where the monetary authority has to take a view on the basis of highly uncertain assumptions: “it is not obvious to me that it is any easier to estimate the output gap than to identify bubbles”, he says.

Bernanke’s attitude is now somewhat more nuanced. He has accepted that there can be circumstances in which asset price fluctuations are of legitimate concern to policy-makers, and that there can be sources of non-fundamental price fluctuations which can be identified.

And he accepts, furthermore, that booms and busts in asset prices have important effects on the real economy. His concern is not so much the so-called “wealth effect”, whereby household spending is depressed by falling net worth, but the balance sheet channel. Firms and households use assets as collateral for borrowing. As asset prices fall the value of that collateral falls, generates an unplanned increase in leverage and impedes access to further credit. Financial firms “which must maintain an adequate ratio of capital to assets, can be deterred from lending…by declines in the value of the assets they hold”. This perfectly describes the painful adjustment process of 2007-2008, which was highly resistant to post-event changes in interest rates. It is therefore surprising that Bernanke was not more concerned in 2006 by what he saw and more ready to act. Of course he inherited a policy stance from his predecessor which was inimical to this kind of analysis. But it also seems that he, along with his colleagues in the other major central banks, was not persuaded that interest rate changes were an appropriate response to a house price bubble, even one which showed every sign of reaching alarming proportions, which leads to the next question one on which opinions differ starkly.

            Q. 4. Even if we can identify misalignments, and believe that some price adjustment is bound to occur, is it right to use interest rates to try to moderate the expansion and bring forward the adjustment?

It is striking how often in this debate, which may seem arcane to many observers, the opposing sides caricature each others positions in the worst tradition of political point-scoring. So advocates of the use of the interest rate weapon prefer to use the non-threatening metaphor of “leaning against the wind”, suggesting nothing more hazardous than a brisk walk in a warm overcoat on a breezy morning. Those who resist it typically raise the stakes by talking of the risks of trying to “prick bubbles”, which sounds an inherently more hazardous activity, likely to be accompanied by loud noises and collateral damage. So Wadwhani, for example, talks of a gentle “tilt” in policy to reduce the risks of further inflation of an emerging price bubble, while Bernanke argues that there is no such thing as “safe popping”, and that “bubbles can normally be arrested only by an increase in interest rates sharp enough to materially slow the whole economy”.
In his view, and the point is repeated regularly by his colleagues in the US and elsewhere, the scale of interest rate changes needed to make a significant impact on a price bubble, whether in the equity or property markets, would be so large as to threaten the health of the economy overall, and inflict greater damage on economic welfare  than a policy of benign neglect , followed by aggressive easing to mitigate the adverse consequences of the crash if and when it comes. The RBA has advanced similar arguments recently.

But there is evidence that it can be done.

 Lars Heikensten, the former Governor of the Swedish Riksbank, wrote

“With house prices increasing drastically…on a few occasions in 2004-05 the Riksbank did for that reason not follow a strict inflation-targeting rule. We “leaned against the wind “, in the sense that we did not take rates down as quickly as we could have done considering the outlook for inflation alone…We explicitly referred to asset prices in our published minutes, press releases and speeches”.

The Riksbank believe that their actions did have a helpful effect on the expansion of asset prices in Sweden, though they did not avoid a fall in 2008. The ECB, too, maintains that it takes asset prices into account in the monetary pillar of its analysis.
So there is a clear fault-line here within the central banking fraternity, though in public both sides are generally keen to play it down. The experience of 2007-08 strengthened the hands of those who favour LATW, and the language used by Governors has begun to change, but the Greenspan tendency is not down and out. Two months ago Mervyn King argued that while maintaining price stability “did not prevent a recession induced by a financial crisis”, nevertheless “ diverting monetary policy from its goal of price stability risks making the economy less stable and the financial system no more so”. This suggests that the Bank of England will not in practice allow its monetary policy to be influenced by its new financial stability mandate, legislated in 2009.

Both sides are agreed, though, that interest rates are not the only weapon that can be used. Even central banks which paid little attention to bank regulation before the crisis have begun to take more interest in capital requirements.  That leads to the fifth and last question.

            Q. 5. Should we try to find and use mechanisms other than interest rates to moderate extravagant credit expansion and associated asset price bubbles?

Almost all central bankers would now answer yes, in principle, to this question, whether or not they believe that interest rate changes should also be used to that end. Indeed some seem to argue that while it is not possible for central banks as monetary authorities to spot a bubble, bank regulators should certainly be able to do so. A more sophisticated version of the argument, is that the burden of proof on bank regulators is not as high as it is on monetary policy-makers. (Regulators contest the point, highlighting that regulatory decisions are in some countries subject to judicial appeals, unlike interest rate moves).The certainty needed to raise interest rates, given the likely impact on economic activity as well as on asset prices directly, is higher than is needed to justify a precautionary rise in capital ratios for banks active in the markets whose behaviour is of concern. So regulators can afford to take a chance, now and again, with fewer downside risks. Macro-prudential adjustments to capital ratios are therefore a preferable response to bubbles.

But raising capital ratios as a precautionary move will also affect the economy as a whole, unless regulators use some kind of sectoral approach (raising capital in relation to mortgage finance only, for example- likely to be a politically unpopular move). Any form of credit rationing will feed through into the  price of credit through interest rate changes, which will themselves affect the monetary stance. So a macro-prudential mechanism is not an easy option, and should be considered alongside the interest rate decision, not apart from it.

So, how do I conclude on the “leaning against the wind” arguments?

In my view central banks must pay more attention to asset price bubbles than they have in the recent past. The output costs of bubbles are large. The IMF estimate that equity market busts are typically associated with a 4 per cent GDP loss, while sharp house price adjustments generate output losses twice as large.

I am not persuaded by the argument that bubbles cannot be identified ex ante. Of course assessing price misalignments is not an exact science, but nor are many other aspects of monetary policy. All interest rate judgements are based an assessment of risk, and the consideration of a range of possible outcomes. There are indicators which can be used to identify potential misalignments.

Furthermore, I  believe that asset prices should be identified as an explicit factor in the consideration of policy.  An element of house prices should be re-incorporated in the index as soon as possible. That would help to explain policy.
It is not enough to say that an inflation target regime can, without amendment, take full account of asset price moves. Inflation forecasts are usually based on assessments of likely inflation at a horizon of two years. It is in principle possible to imagine a regime which looks at longer horizons, to allow the possible long term impact of asset price misalignments to feature explicitly in decision-making, but to do so may complicate the exposition of policy more than would a statement that explains that on occasion policy will be influenced by asset prices and credit expansion, and why. A policy reformulation along these lines does not necessarily pose any significant threat to the clarity of inflation targeting, which is a legitimate central bank concern.

This does not mean that I reject the case for regulatory policies which try to work more directly on asset prices themselves. The argument for more explicitly counter-cyclical adjustments of capital ratios is well-made, though it is interesting that central banks have rediscovered the link between monetary policy analysis and banking regulation only in the aftermath of the crisis. There was little sign of such interest in the years of expansion. The case developed by White, Borio and others at the BIS was given little support by the Fed or other major monetary authorities. It is hard to avoid the conclusion that the new focus on regulatory policy is something of a smokescreen to conceal other policy errors, especially in monetary policy. Furthermore, the second capital accord, Basel 2, with its generally pro-cyclical bias, was in fact developed in the early years of the century with the central banks firmly in charge of the process through their chairmanship, and oversight of the Basel Committee, though it is often presented as a purely regulatory intervention.

So I argue that, in future, monetary policy decision-makers in central banks should pay more attention to the creation of credit, both within and without the banks, in reaching decisions on interest rates. Steve Roach has suggested that one way of doing so would be to give central banks a statutory financial stability objective.  That is an idea worth considering. It has indeed been implemented in the UK, though it is not clear how much it will influence policy there, if at all.

I recognise that the interest-rate weapon is powerful and blunt. There will be circumstances in which it will be more appropriate to act directly on the expansionary appetites of banks themselves, through adjusting capital requirements.  We have to recognise, though, that the application of additional capital requirements for macroeconomic reasons, not directly related to the risk positions of individual banks, is not straightforward, and will require new participants and new processes. Furthermore,  there would be feedback effects to interest rates and hence monetary policy.

If it is accepted that, in future, central bankers and supervisors should try to incorporate an understanding of the state of asset and credit markets in their decisions on capital requirements, with the aim of “leaning against the wind” of unsustainable booms, then we need to consider by whom, and on what grounds, decisions to require more capital would be made. The answer to that question depends on precisely what the ambition is. One can imagine prudential supervisory interventions designed to “protect banks from the cycle”, or which are designed to reduce the contribution to instability by the financial sector: “protecting the cycle from the banks” . At the extreme would be attempts to seek to moderate the business cycle itself, by inducing banks to lean against it. In our view the ambition should be the more limited one of dampening the impact of booms and busts in the banking system on the wider economy, which points to a lead role for regulators, supported by central banks, where they are separate.

For competitive reasons it would be essential for a framework for these decisions, if not the decisions themselves, to be made on an international basis – it would be seen as unreasonable, and contrary to the spirit of the Basel process, for a bank in one country to be put at a disadvantage by its domestic supervisor imposing a macro-prudential adjustment when others were not doing so, unless it is clear that the economic circumstances are very different.

Basel 2 provides a way of thinking about how such a framework might work.  Pillar I sets a basic capital requirement for the bank, based on an assessment of loss probabilities related to its own portfolio of loan assets. Pillar II of the Accord provides for decisions on individual capital requirements to be set, bank by bank, by reference to the idiosyncratic risks of their credit portfolios and the quality of their management. But there is no place in it at present for a supervisor to take a view that, however good the quality of a bank’s book might appear, risk spreads have narrowed too far, and asset prices have escalated in a way which will be unsustainable for borrowers.

Adjusting capital requirements to the state of the economic cycle will be very difficult. The decisions to impose higher capital requirements would require courage on the part of economists and supervisors. But if the assessments were agreed by central banks from a broad range of countries they would have authority and would give line supervisors the cover they need to impose unpopular “taxes” on their banks. We need to recognise that, at bottom, a macro-prudential tool is a tax, and one whose cost would largely be passed through to borrowers and savers. Now would be the right time to introduce a system of this kind, when the memories of the crisis are fresh and the wounds still raw.

At the G20 summit in April, Heads of Government agreed in principle that macroprudential measure should be agreed, though it is not wholly clear that they knew quite what they were asking for. There is much work to do before a workable system, but it should not blind us to the need to reintegrate fiancial market analysis, credit and asset prices into the monetary policy regime. If it is conceived as a substitute for a broader approach to monetary policy, which takes more account of financial stability, then it will inevitably disappoint.

So could central banks have done more to head off the crisis? Yes indeed. And since financial markets are in the words of the song  ‘forever blowing bubbles’ they need to do better next time round.