Monetary Policy and Systemic Risk Prevention - Challenges ahead for Central Banks -
29/09/2006 - Speeches
The Bank of Greece Zolotas Lecture, Athens, 29 September 2006
Monetary Policy and Systemic Risk Prevention - Challenges ahead for Central Banks - by Alexandre Lamfalussy
By way of introduction
The two key mandates of central banking has since long been the preservation of the purchasing power of money and that of systemic stability. Over the past ten years central banks have on the whole been remarkably successful on both counts. Reasonable price stability has been achieved; and systemic stability has been preserved. And this has happened despite an economic environment which, to put it mildly, has not been particularly helpful in producing this result. There has been no shortage of geopolitical risks and acts of violent terrorism; we have had our share of asset price bubbles; the price of oil has, broadly speaking, trebled - not to mention the skyrocketing of a number of other commodity prices; global imbalances, instead of shrinking, have actually widened; and, most important, financial globalisation and the flood of financial innovation have continued unabated. But this success story should not induce complacency. Rather it should incite us to stand back a little and ask some questions about the story's sustainability. I propose to do this to-day in response to the flattering invitation I received from the Governor of the Bank of Greece to deliver this year's Zolotas Lecture, for which I am very grateful. And this is also a way for me to pay tribute to the memory of a great man who deserves our lasting respect.
A short historical reminder: the dotcom bubble
A convenient starting point is to raise the question why the bursting of the US dotcom bubble in 2000 - 2001 and the generalised decline of stock prices which accompanied it failed to lead us into a systemic financial crisis, with potentially dire consequences for the "real" economy of both the United States and the rest of the world.
It is arguable that this could have happened. Consider the following facts, which were highlighted in the 73rd Annual Report of the BIS published in June 2003. From March 2000 to March 2003 the P/E ratio of the US technological stocks dived from 65 to 18. More modestly, that of S&P 500 declined from 38 to 20. This decline implied a decrease in stock market capitalisation which was roughly equivalent to half of the US GDP in 2000. Quite a wealth effect! Yet not one of the large US banks (nor for that matter any of the large banks in the developed countries) got itself into trouble, and after a very short and shallow recession the US economy resumed its "above-the-normal" rate of growth. What happened?
Three converging influences seem to have enhanced the crisis resistance ability of the developed countries' banking systems. Basel I played a key role in allowing the banks to confront the stock market meltdown with a more than comfortable capital base. There had been substantial improvement in the banks' risk management capabilities, of which the prime example was the transfer (via the credit derivatives market) of a substantial portion of the banks ' credit risks to non-bank financial intermediaries, pension funds and other investment vehicles (more about this in a few minutes). Last but surely not least, the speedy and radical decline in the federal funds rate in the United States widened the intermediation margin of the US banks and at the same time allowed them to register sizeable increases in the market value of their Treasury bill and bond portfolios.
But this is not the whole story. It is questionable whether the banks would have been able to cope with the deterioration of their balance sheets had the recession been much deeper and longer. That this did not happen had much to do with the Fed's decision to bring down the targeted federal funds rate from 6.50% in December 2000 to 2% by end-2001 which meant an exceptionally speedy and large monetary stimulus. Further cuts brought this rate down to 1% in 2003, that is, to an inflation-adjusted rate of around minus 1%. The most significant specific effect of this was that the appetite of households for real estate investment was not impaired and prices continued climbing. The bursting of the equity bubble(s) was not accompanied by any meltdown in real estate prices: rather the opposite happened. At the same time there was a radical swing towards a large and growing federal budget deficit, which implied both in size and in speed a fiscal stimulus without a historical precedent.
The reminder of this powerful joint monetary and fiscal stimulus would give the impression that the Fed embarked on a risky monetary policy. Such an impression does not seem to be wholly justified if we take the trouble of recalling the underlying developments in the ''real'' economy of the US at that time. The collapse of the dotcom bubble took place after the longest post - war cyclical upswing - which, however, quite a-typically, did not carry with it the emergence of significant inflationary developments, as measured in terms of price indices (or the implicit GDP deflator), despite the steady decline in the rate of unemployment. Wage moderation was combined, especially from the middle of the 1990s with a remarkable, and largely unexpected, pick - up in the rate of growth of labour productivity. The ''natural rate of unemployment'' appeared to have substantially declined. Whilst at that time there may have been doubts about the nature and the sustainability of fast productivity increases we now know that this was not just a short - term, historical accident: quite a - typically, even during the short recession, the relatively high rate of growth in labour productivity persisted - as it has until to - day with some ups and downs. So something more fundamental seems to have happened in the US economy. Many of us would see the explanation in the joint effect of the deregulation process which started during the 1980s and the way (rather than just the size) in which IT and CT investments have evolved. From a European angle, I look at the unfolding of the US productivity story with some envy.
Regarding central bank policy (and forgetting a minute about the monumental fiscal stimulus) I would be tempted to draw the conclusion that the Fed was well served by the productivity ''surprise'' of the mid - 1990s and by its ability to identify the nature of this surprise, since this surprise contributed powerfully to the lack of inflationary pressure - despite the substantial liquidity creation. It gave time to the Fed to revert to a less accommodating monetary policy, which is what they have been doing over the past two years.
Looking ahead: the challenges of a liquidity ''overhang''
Now let me turn to what I see as potential challenges that would question the wisdom of drawing excessively optimistic conclusions from the dotcom bubble episode.
My main concern is about what I would loosely call world wide excess liquidity or liquidity ''overhang''. I realise that there is no consensus about the definition, and even less about the measurement of excess liquidity, but what carries my conviction is that I am unaware of any possible indicator pointing towards tight liquidity conditions - even now, after the change in the stance of monetary policies. You may compare current real interest rates to ''normal'' levels derived from historical experience, you may look at the levels or changes of interest rate spreads, you may track the rate of growth of variously defined broad money aggregates, you always come to the conclusion that liquidity has been, and is still abundant. What I find even more convincing is the speed and ease with which normality returned to markets after ''credit events'' which in less liquid market conditions could have had longer lasting disruptive effects, including contagion to other markets and outside the United States. A good example of this was given during the Spring of 2005 when the downgrading, to sub-investment status, of the corporate debts of GM and Ford led to only moderate turbulence.
The key question to my mind is whether the high rates of growth of broad money not only in the United States, but also in the Euro area, which characterised the past five - six years, should or should not be regarded with concern. Let me stick my neck out by saying that from the angle of their potential inflation breeding (or at least accommodating) effect I do have some concern, but not a very strong one. The concern, of course, is based on the possibility that the rising energy and commodity prices end up by being incorporated in consumer prices, especially if the business sector regains its pricing power in a fast growing world economy. Contrary to past experiences, this does not seem to have happened on any large scale - so far. Several, relatively novel factors may have played a constraining role in preventing this happening. Under the impact of the explosive export performance of China, and to some extent India, manufacturing corporations in the industrial countries have had to put up with a stiff competitive pressure. A similar pressure can also be observed in services. At the same time, in the industrial countries nominal wage growth has on the whole displayed great restraint. Whatever may be the reason for this - high unemployment in the greater part of Europe, relocation of production facilities (or the fear of it) to low - cost countries, awareness that actual or potential immigration exposes jobs to competition - relatively small nominal wage increases have led to unit labour cost moderation. This happened especially in industries exposed to competitive pressure in the field of internationally traded goods and services - not so much in activities still protected (by inertia or outright protective arrangements).
Whether these countervailing forces will be able to continue to offset the inflation generating impact of energy and commodity prices is of course uncertain. But the outcome depends a lot on the credibility that can be attached to the central banks ' commitment to preserve price stability. Moderating nominal wage increases would become an exceedingly difficult task if market participants were to lose their confidence in the ability of central banks to deliver price stability. I believe that this credibility, gained so painfully during the 1980s and 90s, has so far been preserved. The observation of policy decisions taken by the major central banks over the past years encourages me to believe that this situation will continue to prevail. This explains why my concern is not an overwhelming one.
When it comes, however, to assessing the potential impact of excess liquidity on financial stability my concern becomes an unqualified one. There are several reasons for this. A general concern is that abundant liquidity creates a favourable breeding ground for asset prices bubbles, or if you dislike the use of this word, for allowing asset prices to have rates of increase lastingly outstripping that of consumer prices. The danger here is that such developments affect the main groups of asset prices simultaneously, i.e. both financial assets and real estate. I realise of course that it is exceedingly difficult to ascertain when an asset price boom turns into a bubble, but the fact is that there have been and will be bubbles. And the trouble with bubbles is that, instead of gently deflating, they often tend to collapse, with potentially damaging consequences for systemic stability.
The second, more ''micro'' reason is that abundant liquidity pushes market participants towards excessive risk taking. Anyone who has the opportunity of watching even from a certain distance the generalised, often frantic search for assets that yield a return only a shade higher than the one prevailing for far safer assets, yet carry a substantially higher risk, knows what I am talking about. Most - almost all - indicators point nowadays towards a persistently strong appetite for risk taking. Despite the turnaround in the policy of the Federal Reserve some two years ago, risk premiums are still close to historical lows.
My third reason is that both the possible emergence of ''global'' bubbles and the persistent appetite for risk taking has been encouraged by the perception that while central banks have displayed a remarkable ability to prevent that specific ''credit events'' or other kinds of crisis manifestations turn into a full blown systemic crisis, they did precious little to discourage bubbles from arising. Whether the perception of this asymmetry - which carries the risk of entertaining moral hazard - is justified or not is debatable. But it underlines the extraordinarily difficult balancing act that central banks are supposed to perform between, on the one hand, preventing the emergence of (what I have just called) a ''full blown systemic crisis'' and, on the other, refusing to bail out too many imprudent market participants either indirectly (by pumping liquidity into the system) or directly, by providing emergency liquidity assistance to specific banks.
A number of events happened since the late 1980s which have made this balancing act even more difficult. At the risk of oversimplification I would argue that in earlier times many central bankers regarded asset price bubbles as a sort of natural by-product of inflationary developments. This view implied that bubbles can be avoided, or at least contained, if inflation remains under control. The first serious doubt about the general validity of this belief emerged at the time of soaring real estate and equity prices in Japan in the second half of the 1980s. The P/E ratio of equities which stood at around 35 (an already quite high level) in 1984 - 85 peaked at 72 in 1987, and after an initial decline following the collapse of US equity prices in October that year, regained another high at 70 (which induced me to talk about their ''levitation'' in one of the BIS reports). At the same time, real estate prices more than trebled between early 1985 and end - 1988. This stands in sharp contrast with the fact that during the second half of the 1980s the average yearly rate of consumer price inflation was less than 2%. The second major event was the development of the dotcom bubble in the US between 1997 and early 2000 - again, at a time when inflation was well under control. The third, recent example is provided by the sharp rise of US house prices over the past three years which has radically outpaced the relatively modest increase in consumer prices (although since the beginning of this year this discrepancy seems to have come to an end). All three examples suggest that central bankers should not remain indifferent to such developments, even in circumstances when inflation appears to be reasonably well under control. This applies in particular to situations where the high level of indebtedness of households and/or corporations would be liable to induce radical expenditure cutting on their part in response to diving asset values - an expenditure cutting which would be exacerbated by the reluctance of banks to grant credits.
Aggravating factors: global imbalances and financial globalisation
Reducing world - wide excess liquidity, without putting at risk the growth of the world economy, would amount to a major challenge for the central banking community in any circumstances. It is becoming a truly difficult task to perform when we take into account that this will have to be done in a world environment characterised by growing imbalances and new developments in financial globalisation. This is a world very different from the one in which the dotcom bubble developed and eventually collapsed.
The financial imbalances are sufficiently well known to allow me to be brief. The core manifestation of these imbalances is the twin deficit of the United States: a current account deficit of around 6% and a huge domestic net saving deficit which itself is the result of large net borrowing both by the government and the household sector. While none of us can know how long this can go on, we would, I suspect, all agree that it cannot go on for ever. The problem is aggravated by the fact that one of the major international counterparts of the US deficit is the Chinese surplus - a bizarre situation which boils down to the fact that a (still) very poor country finances the wealthiest country in the world. The key question is how to implement policies which would lead to a gentle unwinding of these imbalances - that is, to soft - landing. I believe that we would more or less agree on the main features of such policies. But the danger is that adjustment will have to imply changes in exchange rates and interest rates, and no one can know in advance whether we can expect gradual price movements rather than abrupt ones. Given the size of these imbalances, abrupt shifts in global asset allocation can play havoc with our financial system.
The possibility that this might happen is enhanced by a number of novel features of our globalised financial system, of which I would mention just three.
One is the simple fact that China has by now become a major player in international financial markets. The most striking fact in this respect is that as a result of the spectacular growth of its foreign exchange reserves it is playing a significant role in the US Treasury bill and bond market and could also influence exchange rate developments for the reserve currencies. At the same time it operates a banking system in which bank balance sheets are growing at a rate at which it is close to impossible to practice prudent credit policies. The amount of non performing bank loans is subject of much guessing, but there is little doubt that a simple deceleration of growth would be liable to produce bank failures. This goes a long way towards explaining the reticence of the Chinese authorities to engineer an exchange rate adjustment which at the same time would seem to be an essential component of a policy package setting in motion the unwinding of the global current account imbalances. What really matters is that the Chinese figures have by now reached a level at which domestic financial disturbance in China might have disruptive consequences for global financial stability.
The second fact is that the flow of innovation affecting international capital markets - far from calming down - has accelerated. Take the case of credit derivatives, which (as I have mentioned earlier) may well have played a beneficial role in allowing banks to withstand the shock of the 2000 - 2003 meltdown of equity prices. But with the currently prevailing astronomical credit derivatives figures we have to raise the question whether these risk transfers are a source of efficiency and strength for our globalised markets or, on the contrary, a source of fragility. Just as in any insurance scheme, when risk averse market participants transfer their risk to willing risk takers the market gains in stability as long as the latter know what they are doing and are capable of rightly assessing their ability to absorb shocks. Whether these two conditions are satisfied or not is a moot point, and my own instinctive answer is that we just don ' t know. My reason for doubt is that insurance in financial markets is based on the assessment of covariations (or the opposite), for which we possess a data base going back only a limited number of years. As we have learnt from the LTCM crisis in 1998, any of these observed relations may easily change under the influence of new innovation or political shocks: in our world, financial globalisation is not a ''steady state'', but a process involving new countries, new products, new market participants. But this is not the whole story. The generalised use of derivatives has made our markets extraordinarily opaque, by which I mean that it has become impossible to identify, with any degree of certainty, the interlinkages between various market segments. Gone are the good old days of the early 1980s when, by looking at the cross border asset and liability banking figures collected by the BIS, central banks gained a very useful insight into such interlinkages. These figures are still useful, but their importance has been dwarfed by the links created through markets for derivatives.
The third, more specific fact is the spectacular come - back and growth of the hedge fund industry - a striking example of how financial innovation may affect the structure and the modus operandi of financial markets. Regarding these funds, I very much share what the June 2006 issue of the ECB's Financial Stability Review said about the danger of herding which may have come about as a sort of by - product of the aggressive hunt for yields: ''One indication of this herding, or so - called crowding of trades, is that hedge fund returns, both within and across different investment strategies, became increasingly correlated after mid - 2003. In addition, an area of growing concern has been the exponential growth of CRT markets in which hedge funds - institutions which tend to be rather opaque about their activities - are known to have become increasingly active''. I have also noted with interest the efforts deployed by investment banks to become, or to preserve their position as, prime brokers to hedge funds - a highly lucrative position - which they try to achieve by offering the full range of services requested by hedge funds. This could include, I presume, granting loans to hedge funds. I only hope that the banks in question possess the appropriate information on the hedge funds ' ability to survive during times of stress, and will act accordingly.
Finally, let me add a concern which has to do with globalisation in general, but could have, indirectly, implications for financial stability. These are the accelerating changes in industrial, and therefore employment structure. Even if we believe that the transfer of labour from (unskilled) labour - intensive industries hit by global competition to high productivity sectors will yield ''globally'' beneficial results, the process of transfer is likely to be painful - so painful, indeed, that it could lead to social upheaval, ending up with political backlash. To avoid this happening, great will be the temptation to resort to protectionism, which is a sure recipe for aggravating economic, and therefore financial distress.
The main concern I wanted to share with you in this short lecture is that world wide excess liquidity may store up trouble for us in the future. First, it may trigger or at least accommodate inflationary developments. But given the awareness of the directly concerned major central banks of such potential (or perhaps already observable) developments, I trust that they will take this into account in their decision making process. I also believe that that there are countervailing influences in the ''real'' economy, mostly arising out of pressures on prices and wages emanating from globalisation. For these two reasons my concern in this field is somewhat moderated.
But this moderation does not apply to my concern that excess liquidity represents a genuine, although unquantifiable, danger for the stability of our financial system. It does so because (a) it provides a favourable breeding ground for developing bubbles in markets for asset and commodity prices, (b) it erodes risk awareness and therefore encourages careless risk taking, and (c) because the story of its emergence over the past five years may have given - rightly or wrongly - the impression to market participants that central banks are more active in trying to protect them when the bubbles burst than to prevent the emergence of bubbles in the first place. In addition, we have to bear in mind that liquidity abundance coincides with (d) global current account imbalances, the unwinding of which may go hand in hand with interest rate and exchange rate turbulence, and (e) the accelerated pace of financial innovation which enhances the opaqueness of financial markets.
Bearing in mind that in the event of financial crisis manifestations central banks would inevitably find themselves in the ''first line'', they would be well advised to gradually absorb global excess liquidity and cooperate among themselves and with governments with a view of implementing policy decisions which would contribute to unwinding global imbalances. They should also actively participate in the elaboration of a regulatory framework which would (a) enhance the ability of financial institutions, and of banks in particular, to resist financial turbulence and (b) strengthen the payments and securities settlement systems – which otherwise could be instrumental in transforming specific crisis manifestations into systemic crisis. Quite an assignment!
It is however possible, although (hopefully) perhaps not that likely, that despite these precautionary moves a crisis will erupt with systemic ramifications. In such a situation central banks would be called upon to pump liquidity into the system and in extreme cases provide emergency liquidity assistance to particular institutions. As already mentioned, central banks should respond to such requests with utmost care: there is a genuine danger that speedy and large scale intervention would breed moral hazard. Be that as it may, situations might nevertheless arise when these two kinds of central bank intervention would be warranted. To take the right decision, central banks need information, and to gain such information they need a ''real time'' insight into the working of the financial system as a whole, and of banks in particular. The organisation of the channels through which such insight is obtained is an extremely difficult task. My concluding concern is that those central banks which do not have an operational involvement in bank supervision would find it hard to gather the appropriate information.