Objectives and challenges of monetary policy

Speech by Mr Malcolm D Knight, General Manager of the BIS, at the Conference on Inflation Targeting, Magyar Nemzeti Bank, 18-19 January 2007.

Abstract

While an inflation targeting framework has served many central banks well over the last decade or so, at least three challenges remain. First, how can one gauge underlying inflationary pressures when inflation has been low and stable for an extended period? Second, how should the possible emergence of boom-bust cycles be dealt with within an inflation targeting framework? Third, what complications can arise in such a framework from expansionary fiscal policies?

Full speech

Let me first thank the organisers for the invitation to address this conference at Magyar Nemzeti Bank. While the titles of the next two sessions take it for granted that an inflation targeting framework is appropriate for Hungary, this session addresses a broader theme. Thus it allows me to paint upon a somewhat larger canvas and raise some possible issues and qualifications. Indeed, I want to highlight three sets of challenges, some old and some new, faced by those charged with the conduct of monetary policy in today's world. Naturally, I hope that the points I make today will resonate with all the central bankers from the various countries assembled here, but I particularly hope that they will be of some relevance to our Hungarian colleagues, because it is important to note that they are operating against a macroeconomic backdrop of twin deficits, fiscal and external, that adds an extra degree of complexity to the conduct of their monetary policy.

Let me begin by making it very clear that I do not intend to contest the entirely justified view, now almost the conventional wisdom, that the principal objective of monetary policy should be the pursuit of price stability. In practice, this means that central banks should pursue the goal of keeping inflation low, stable and predictable (say near 2%) over the one- to two-year policy horizon implied by perceptions about the lags in the monetary transmission mechanism. There is no need for me to list the many microeconomic, macroeconomic and social reasons for wanting to do this. Suffice it to say that high inflation poses a threat to sustained economic growth, and also to social stability. We have seen the proof of this in many countries over the centuries, not least in the industrial countries during the 1970s and early 1980s. The experience of high inflation was not pleasant at the time, but it also left a legacy of debt, resource misallocation and volatile expectations that took years to reverse. As central bankers, we can take pride in having both reduced inflation and kept it down over the last two decades.

Yet, this said, it also bears recognition that achieving and maintaining a state of low inflation might not - in and of itself - be sufficient to ensure good economic performance consistently over long periods. Even with low inflation, financial stability problems can occur. There was, for example, no significant inflation in Southeast Asia before the crisis of the late 1990s. Going back a little further, inflation was low in Japan in the 1980s. And going back much further, there was not inflation but mild deflation in the United States in the 1920s. Yet each of these extended periods of reasonable price stability was followed by a severe and sometimes long-lasting economic setback that had a major impact on both growth and unemployment. And so the question is whether monetary policymakers need, at least sometimes, to try to mitigate these problems as well.

I know that some will be tempted to say that there is little to be learned from these episodes of financial instability in the distant past, and that the world has changed enormously in the interval. But a closer look at these changes actually leads one to a quite different conclusion. The continued process of liberalisation and globalisation, affecting both the real and the financial sectors, has produced a world that has more in common with the pre-1930 period than the heavily regulated era that followed it for many years. It is not inconceivable that today's more efficient economy might also be inherently less stable, since there is now much freer rein for rational exuberance to turn into irrational exuberance. At the very least, prudent policymakers should keep an open mind about the possible limitations of their chosen policy regime, recognising how frequently in the past fashionable regimes have had to be abandoned because they simply failed to deliver the goods.

The first set of challenges faced by central bankers today is a by-product of success. The fact is that both inflation and its volatility have in many countries been reduced to low levels. One problem this poses is econometric. In such an environment, traditional empirical relationships between capacity "gaps" and inflation break down. How then is a central bank to gauge whether monetary policy might be stimulating demand to a point that risks unleashing a sudden shift in either inflation or inflation expectations? Indeed, given the stability of both series over such a long period, might monetary policy not even be inclined to test such limits? The fact that global growth has been at record levels over the last three years, while real interest rates have been maintained at remarkably low levels and credit is growing rapidly almost everywhere, could imply that a problem is indeed building up.

Another problem associated with low inflation and low inflation expectations has to do with uncertainty as to the cause of this phenomenon. If it is primarily due to confidence in the monetary regime, looking forward, then this happy state of affairs might be more likely to continue. But, conversely, low inflation could have been due primarily to a series of positive supply side shocks (China, India, the Wal-Mart effect etc), with expectations being formed in a backward-looking way. In this case, should these shocks cease or even just moderate, then the degree to which demand had been allowed to rise to unsustainable levels would more quickly become evident in rising inflationary pressures.

The possibility of low inflation reflecting the influence of positive supply shocks also raises another important question: should central banks worry about unusually low inflation, or even deflation, arising from such a source? More specifically, does it always make sense to respond to a moderate slowing of growth with a substantial monetary accommodation when the side effect of such a relaxation could itself entail risks?

This question brings me directly to the second set of challenges which face central bankers today, the increased likelihood of "boom-bust" conditions emerging in liberalised economies with their associated problems of financial instability. Indeed, all of the historical episodes I have just referred to were characterised by initial "New Era" optimism, rapid credit expansion, strong increases in asset prices and significant effects on the allocation of real capital. At the present juncture, there seem to be two schools of thought as to how to respond. Unfortunately, both have their shortcomings.

On the one hand, the European Central Bank and the Bank of Japan appear to judge that some preventive action might be required. The ECB has its "two-pillar" approach, with the monetary pillar increasingly being interpreted as an indicator of when problems are building up. The Bank of Japan has described its "two-perspective" approach which combines a more traditional concern for near-term inflationary pressures with a longer-term concern to avoid repeating the mistakes of the 1980s. In a similar vein, a number of other central banks have announced that they reserve the right, in unusual circumstances, to lengthen the horizon of their policy forecast to incorporate the full effects of their policies acting through balance sheet and other non-traditional channels. The Japanese authorities, and perhaps others, have also been encouraged to pursue preventive measures by their experience of the difficulties of "pushing on the monetary string" after a boom has turned to bust.

But there are some practical problems with this approach. How can one reliably identify when imbalances are building up enough to merit a monetary tightening? The point has been made repeatedly about the difficulties of targeting asset prices, though it presumably has less weight when a number of the relevant indicators (credit, money, various asset prices, investment etc) are simultaneously behaving in a threatening manner. What degree of monetary tightening might be required to moderate the growing euphoria, and might it do great harm to particular sectors of the economy? In particular, what might happen to the exchange rate in a small open economy like that of Hungary, and what might be the implications for the current account? And if inflation were initially under control, and tighter policy pushed it below the targeted levels, how difficult would this be to explain to the public?

In the face of such difficulties, a number of central banks intent on preventive action have actually turned to the use of supervisory instruments to promote credit restraint. But, of course, in many countries central banks do not have the authority to use such instruments, and regulations routinely invite evasion. Nevertheless, this issue of monetary and regulatory cooperation in resisting emerging imbalances needs more attention, as does the international dimension of such problems. This is particularly the case in central and eastern Europe, where most of the credit is being granted by commercial banks headquartered in western Europe but the costs of macroeconomic instability will be borne by those further east.

On the other hand, there is a second school of thought. This school says that, in light of all the shortcomings I have just noted, preventive measures should be avoided and reliance should instead be put on vigorous monetary easing as soon as a downturn begins. In fact, it could be contended that this approach has been used repeatedly over the last two decades and has effectively limited the impact of a large number of financial disturbances. As for the Japanese contention that this policy did not work in Japan, the counterargument would be the assertion that the Japanese policy response was in fact too slow and too timid.

But this second approach also has its shortcomings. Contrary to the assertion I have just referred to, it might not work. It is hard to believe that the 15-year period of sub-par growth in Japan had its origins in an initially too timid monetary reflation. In contrast, it is not so hard to believe that the zero lower bound for policy interest rates proved an important constraint, and that high debt levels and excessive capital built up in the boom years subsequently became a long-lasting burden on the economy. Moreover, recourse to extremely low interest rates can itself create further imbalances in both the real and financial sectors of the economy. Think of the stultifying impact of "zombie" companies in Japan, the increased risk-taking in financial markets encouraged by the need to "search for yield", and the effect of low interest rates on household saving rates in many countries.

So what is it to be, the first approach of tightening monetary policy to forestall financial imbalances even though inflation is broadly on target? Or the second - letting the boom run its course as long as inflation remains on track and then loosening monetary policy aggressively as soon as the downturn begins? The analytical challenge posed by this question is underlined by recognising that this question was essentially the one debated by Keynes and Hayek in the 1930s. The debate goes on, and hopefully the supporting research as well.

And, finally, there is a third challenge commonly faced by monetary policymakers pursuing price stability, that of an unsupportive fiscal environment, often accompanied by an external current account deficit. It seems to me that this corresponds to the current situation in Hungary and it poses a formidable complication for the conduct of monetary policy. The famous article on "Some unpleasant monetarist arithmetic" by Wallace and Sargent provided an analytical proof of how unrelenting fiscal stimulus would eventually force monetary accommodation. But more recent experience, particularly in small open economies, throws an interesting light on the dangerous processes through which this might occur.

In a world of highly mobile capital flows, fiscal stimulus can initially result in a much stronger exchange rate even though the preceding monetary tightening might have been quite moderate. This configuration encourages domestic asset prices to rise, while increased domestic absorption and the stronger exchange rate cause the trade account to deteriorate. But eventually, as stocks of both government and external debt rise, financial markets will impose a risk premium on further borrowing, perhaps brutally. In this situation, the exchange rate can become unstable on the downside, while longer-term bond rates become unstable on the upside.

Evidently, the central bank can resist one of these tendencies only at the risk of aggravating the other, with the upshot often being a steeper depreciation and higher initial inflation followed by recession. In Hungary and a number of other countries in central and eastern Europe, currency depreciation might also induce a domestic demand contraction, rather than trade-promoted expansion, due to the large proportion of household and enterprise debts denominated in foreign currency. Depreciation in such circumstances raises debt servicing burdens, which can reduce the scope for domestic spending. Given these unpleasant prospects, all one can suggest is that governments get their fiscal houses in order as quickly as possible. Even a credible plan to do so soon would provide important support to market sentiment.

While we could all think of many other challenges to the conduct of monetary policy, let me stop at these three: keeping inflation low, stable and predictable; avoiding the episodes of financial instability that occur periodically even when inflation is under control; and the special challenge of continuing to make good monetary policy in the face of an unsupportive fiscal environment. I believe each of these challenges provides enough stimulus to analytical work to keep all of us occupied for some time.

Many thanks for your attention.