Economics is not a science, at least not in the sense that repeated experiments always produce the same results. Thus, economic forecasts are often widely off the mark, particularly at cyclical turning points, with inadequate data, deficient models and random shocks often conspiring to produce unsatisfactory outcomes. Even trickier is the task of assigning probabilities to the risks surrounding forecasts. Indeed, this is so difficult that it is scarcely an exaggeration to say that we face a fundamentally uncertain world – one in which probabilities cannot be calculated – rather than simply a risky one.

Economic history is a useful guide in this respect. The Great Inflation in the 1970s took most commentators and policymakers completely by surprise, as did the pace of disinflation and the subsequent economic recovery after the problem was effectively confronted. Similarly, virtually no one foresaw the Great Depression of the 1930s, or the crises which affected Japan and Southeast Asia in the early and late 1990s, respectively. In fact, each downturn was preceded by a period of non-inflationary growth exuberant enough to lead many commentators to suggest that a “new era” had arrived. Similar surprises can be noted at a more micro level. Around the time of the failure of LTCM in 1998, the firm faced price shocks in various markets that were almost 10 times larger than might reasonably have been expected based on previous history. As a result, its fundamental assumptions – that it was adequately diversified, had ample liquidity and was well capitalised – all proved disastrously wrong.

Of course, many will say that our understanding of economic processes has improved thanks to this experience. Yet this is not such an easy proposition to prove. Consider, for example, the typical way in which central bank economists forecast future inflation using econometric models of how wages and prices interact. To do this accurately, at least five questions have to be answered correctly. What is the best way to measure excess capacity in the domestic economy? What is the trend rate of growth of productivity? Are foreign influences limited to import prices alone? Are wages driven by forward-looking price expectations, or by past price developments? If expectations are important, are they influenced by the credibility of central banks or by something else, like actual or even perceived inflation? Each of these questions is currently highly contentious. And when we turn to other economic variables, the degree of disagreement about many equally fundamental issues is just as great.

Indeed, in the light of massive and ongoing structural changes, it is not hard to argue that our understanding of economic processes may even be less today than it was in the past. On the real side of the economy, a combination of technological progress and globalisation has revolutionised production. On the financial side, new players, new instruments and new attitudes have proven equally revolutionary. And on the monetary side, increasingly independent central banks have changed dramatically in terms of both how they act and how they communicate with the public. In the midst of all this change, could anyone seriously contend that it is business as usual?

There is, moreover, a special uncertainty in the area of monetary policy. While the commitment of central bankers to the pursuit of price stability has never been stronger, the role played by money and credit is being increasingly debated, against the backdrop of the uncertainty about the inflation process referred to above. For some central banks, and indeed many leading academics, neither money nor credit is thought to play any useful role in the conduct of monetary policy. For others, in contrast, the too rapid growth of such aggregates could be either a harbinger of inflation or the sign of a financially driven boom-bust cycle with its own unwelcome characteristics.

Against this background, neither central banks nor the markets are likely to be infallible in their judgments. This has important implications. The implication for markets is that they must continue to do their own independent thinking. Simply looking into the mirror of the central banks’ convictions could well prove a dangerous strategy. The implication for policymakers is that they should continue to work on improving the resilience of the system to inevitable but unexpected shocks.

Topics of current concern to policymakers

The consensus forecast for the global economy, which is obtained through a poll of economists, anticipates that recent high levels of growth will continue, that global inflation will stay quite subdued, and that global current account imbalances will gradually moderate. With respect to financial markets, the consensus forecast for 2007 is that long rates will stay around current levels. Evidently, and appropriately, this forecast implicitly assumes that there will be no major geopolitical disruptions and no disturbances in the financial sector significant enough to affect the real economy.

As a near-term proposition, a forecast that says the future will be a lot like the past has much to recommend it. Indeed, looking closely at forecast errors in recent years, one might conclude that there are grounds for even greater optimism. Real growth has, on the whole, been stronger than expected, while inflation has generally stayed in line with predictions, despite sharp increases in commodity prices in the last year or so. Long-term interest rates have also consistently come in below anticipated levels. Since it is well known that forecast errors often display a significant degree of persistence, one might with some confidence expect the good news to continue. Only with respect to global trade imbalances have the actual outturns been markedly worse than expected, but even here, as noted in the Introduction to this Annual Report, there are some signs of improvement.

Yet it is not difficult to identify uncertainties that could conceivably cause this near-term forecast to come unstuck, or that could result in less welcome outcomes over a longer horizon. Below, various areas of concern are identified and analysed separately, although they could well be interdependent. As will be described in the following section, those who are more focused on such interdependencies tend to see accommodative financial conditions as the causal thread linking these areas of concern together.

A first uncertainty has to do with the possible resurgence of global inflation and, potentially, inflation expectations. Estimates of capacity gaps in most of the major industrial countries indicate that they are approaching or have reached the limits of their potential. Disinflation pressures originating in emerging market economies also seem to be easing in the wake of sometimes extraordinary domestic growth rates. In China, in particular, it has become increasingly clear over the last few months that measures to slow the economy have not been effective so far. Partially as a result, continued strong increases in global energy and other commodity prices show no signs of abating, and questions are being raised about the ongoing capacity of companies to offset these higher costs through savings elsewhere. Finally, the fact that monetary and credit aggregates have also been growing very rapidly, not least in countries such as China that use foreign exchange intervention to resist currency appreciation, is a further worrisome sign for many.

Given the still pivotal role of the United States in the world economy, the possible inflationary impact of cyclically rising wages and declining productivity growth is a source of near-term uncertainty. In addition, two medium-term considerations need to be taken into account. The ratio of house prices to rents is at an all-time high. Unless house prices fall significantly, a renormalisation would imply a prospective rise in rents which would feed directly into the measured CPI. Moreover, if global trade imbalances need to be resolved, a further and perhaps substantial decline in the dollar might also be part of the adjustment process. To date, shrinking foreign margins, allied with productivity increases, have sufficed to keep exchange rate pass-through to a minimum in the United States. Whether this will continue remains to be seen.

Viewed in this light, the recent slowing in the US economy must be judged welcome. Yet, as 2006 wore on, concerns began to mount that this might turn into rather too much of a good thing. The attention of financial markets first focused on the US subprime mortgage market, but the underlying issue is much broader. The household saving rate in the United States fell for a time into negative territory, as sluggish wage growth failed to provide adequate support for a sharp increase in consumer spending and residential investment. Easy credit terms, especially in the mortgage market, encouraged both higher debt levels and higher house prices. The latter, in turn, provided both the collateral to justify more lending, and the perception of increased wealth to justify more spending.

The concern is that this might all reverse. Debt service levels are already elevated and mortgage rates might rise further. House prices only need to stop rising (indeed, this may already have happened) to slow both the recourse to credit and the sense of confidence arising from increases in wealth. Moreover, when cuts in construction jobs begin to match the much larger fall in housing starts to date, then wage income, job security and confidence could be further affected. Were corporate fixed investment, already inexplicably weak given high profits and low financing costs, to retreat as well, then the stage might be set for a more significant and perhaps unwelcome deceleration in US growth.

If this is the risk, it must also be recorded that the ratio of US household debt to income has been creeping up for decades without seriously compromising consumer confidence. Consumer spending could also get a second wind from faster wage increases. The wage share is secularly low and might rebound. Moreover, the United States is at that “late cycle” stage, when unemployment is low and compensation normally tends to rise. However, this possibility could have undesirable implications as well. As noted above, inflation pressures might increase, or, if wage demands instead cut into profit margins, stock market expectations could be disappointed, with possible implications for both asset prices and corporate investment.

Were the US economy to slow substantially, the crucial question would be how others might be affected. On the one hand, domestic demand has recently gained strength in the euro area and Japan, as well as in a number of emerging market economies. Furthermore, unlike the IT-related slowdown around the turn of the century, there has not been a synchronised industrial boom that might suddenly collapse at the global level. Support for continued global growth is also provided by the falling share of exports to the United States, in a context of surging world trade overall.

On the other hand, in both Germany and Japan the revival of domestic demand has been overwhelmingly in the form of corporate investment, itself driven by strong export demand. The same can be said for China, where the growth rate of the economy has been characterised by Premier Wen Jiabao as “unstable, unbalanced, uncoordinated and unsustainable”. Moreover, even without a synchronised business cycle, boardroom confidence globally might be affected by a sharp US downturn. And while direct exports to the United States might have fallen relative to global totals, a major component of the latter has been imports for assembly in China. In this regard, the indirect exposure of many countries in Asia to slower US growth might still be significant. Finally, it is notable that the United States is by no means alone in its dependence on debt-fuelled consumption, with some countries even having substantially negative household saving rates. This provides a further channel for possible contagion.

To near-term uncertainties about inflation and growth must be added a number of medium-term concerns, not least persistent and substantial global trade imbalances. Does this constitute a problem, requiring a policy response to lower the possibility of large and perhaps abrupt movements in exchange rates? Or, rather, can we assume that the capital inflows needed to finance such deficits will be available on not significantly different terms for the foreseeable future?

Countries with large trade deficits are generally those where domestic demand has been growing relatively fast, and where interest rates are relatively high in consequence. In principle, such countries should also have depreciating currencies. This would allow external deficits to be reduced over time as domestic demand began to ease under the influence of higher rates. Unfortunately, in practice, relatively high interest rates often induce private capital inflows of such a magnitude as to cause the exchange rate to appreciate rather than depreciate, and to raise domestic asset prices, which leads to more spending rather than less. Both these developments will cause the trade deficit to worsen further. This process was very much part of the story in the United States prior to 2001, and the second element of it continues today. Moreover, in recent years there have been a number of variations on this “carry trade” theme, with still more dramatic effects on smaller economies like New Zealand and a number of countries in central and eastern Europe. In many of these countries, including some where fundamentals have significantly improved, fears have been rising that a sudden reversal of such capital flows might significantly complicate macroeconomic management.

The US trade deficit is of a very special nature, largely because of the dollar’s role as a reserve currency. Thus, the significant reduction of private sector capital inflows after 2001 was counterbalanced by inflows from the public sector, leading to only a gradual decline in the value of the dollar. This has had the advantage of being quite manageable, but the disadvantage is that there has been no discernible reduction in the US trade deficit. When we add to this the gradual movement of the service account into deficit, and the growing size of the external debt position, the dollar clearly remains vulnerable to a sudden loss of private sector confidence, and presumably associated increases in risk premia in financial markets. While to some degree this would be welcome, as part of the external adjustment process, it could at the same time aggravate both near-term inflation pressures and the risks of a more serious downturn.

The reliability of public sector inflows has also become more uncertain, for at least two reasons. First, countries outside the United States might now be increasingly inclined to reduce intervention and let their currencies rise. Reasons for this might include a desire to limit the losses arising from an evergrowing currency exposure. But, likely to be of greater importance, there is mounting evidence of the domestic distortions associated with both currency intervention and easy monetary policies whose effect has been to hold down exchange rates. Authorities in China, Japan and some commodity-producing countries have already publicly expressed strong concern about excessive capital investments, and possible resource misallocations, in their respective countries. And, as noted above, in a number of countries inflationary pressures are rising and sterilisation seems to be becoming increasingly difficult.

The second potential threat is that holders of large portfolios of reserves might begin to reduce the proportion of new reserves held in US dollars. On the one hand, the principle of uncovered interest parity implies that, over sufficiently long time horizons, returns will not be increased by such a strategy. This would argue against the rebalancing of portfolios still primarily held in dollars. On the other hand, the variance of such returns, measured in domestic currency, could be reduced if the currency composition of the reserve portfolio were chosen with this end in mind. Whether concerns about the variability of returns would provide sufficient motivation for a significant reduction in the proportion of dollar holdings remains an open question. So too does the issue of whether such official actions would materially affect exchange rates, barring widespread imitation by the private sector. What is more certain is that, as reserve managers increasingly focus on maximising returns, they will be attracted to the currencies of countries that give them ready access to equity and other instruments that allow them to do so.

A final set of medium-term uncertainties has to do with potential vulnerabilities in financial markets and possible knock-on effects on financial institutions. As noted in the Introduction, the prices of virtually all assets have been trending upwards, almost without interruption, since the middle of 2003. For some commentators, it is not hard to find plausible reasons why individual asset price increases are justified and therefore more likely to be sustainable. For example, the very low risk spreads on sovereign issues are consistent with clear improvements in governance and macroeconomic policies in many countries. Comparably low spreads on high-risk corporates reflect high profits and very low default rates in recent years. Unusually low term premia could be the result of the absence of volatility in the major macro variables for some time. The prices of commodities and fine art reflect new sources of demand from newly emerging markets. And the increase in house prices, which has now become almost a global phenomenon, can be ascribed to lower longterm mortgage rates.

Yet it could also be suggested, consistent with the inherent difficulty of making longer-term valuations, that the market reaction to good news might have become irrationally exuberant. There seems to be a natural tendency in markets for past successes to lead to more risk-taking, more leverage, more funding, higher prices, more collateral and, in turn, more risk-taking. One manifestation of this, over the last few years, has been that the intermittent periods of financial volatility have become progressively shorter. Apparently, the observed resilience of markets to successive shocks has increasingly encouraged the view that lower prices constitute a buying opportunity. The danger with such endogenous market processes is that they can, indeed must, eventually go into reverse if the fundamentals have been overpriced. Moreover, should liquidity dry up and correlations among asset prices rise, the concern would be that prices might also overshoot on the downside. Such cycles have been seen many times in the past.

The obvious question is: who might be hurt by such a turn of events? The big investment and commercial banks seem very well capitalised, and many have been making record profits. Their attention to risk management issues has also been unprecedented. Yet some sources of concern must already have been identified by the markets, since the spreads on credit default swaps for some of the best known names have recently been elevated in comparison to the levels that would be normal given their credit ratings. One area of concern is market risk and leverage. Balance sheets have grown significantly. Moreover, value-at-risk measures have stayed constant even though measured volatility has fallen substantially. Another possible worry, linked to the “originate and distribute” strategy, is that originators might be stuck with a warehouse of depreciating assets in turbulent times. The fact that banks are now increasingly providing bridge equity, along with bridge loans, to support the still growing number of corporate mergers and acquisitions, is not a good sign. A closely related concern is the possibility that banks have, either intentionally or inadvertently, retained a significant degree of credit risk on their books.

Assuming that the big banks have managed to distribute more widely the risks inherent in the loans they have made, who now holds these risks, and can they manage them adequately? The honest answer is that we do not know. Much of the risk is embodied in various forms of asset-backed securities of growing complexity and opacity. They have been purchased by a wide range of smaller banks, pension funds, insurance companies, hedge funds, other funds and even individuals, who have been encouraged to invest by the generally high ratings given to these instruments. Unfortunately, the ratings reflect only expected credit losses, and not the unusually high probability of tail events that could have large effects on market values. Hedge funds might be most exposed, since many have tended to specialise in purchases of the riskiest sorts of these instruments, and their inherent leverage can in consequence be very high.

It is not, by definition, possible to put all these uncertainties together and arrive at a prediction. Rather, if one believes that a range of possible developments could all interact in various ways, such interactions could form the basis of a thousand stories. Yet it must be noted that behind each set of concerns lurks the common factor of the highly accommodating financial conditions noted in the Introduction. While this observation need not call into question the consensus forecast as such, it should at least serve to remind us that tail events affecting the global economy might at some point have much higher costs than is commonly supposed.

The full report is available at http://www.bis.org/publ/arpdf/ar2007e.htm