Interview: Claudio Borio

The BIS has been at the forefront of some very stimulating research suggesting that monetary policy should perhaps pay more attention to asset price developments. Do you think that some of these insights have trickled down to the policy-making process, or are these arguments still too controversial?

The arguments are still controversial, and rightly so: the issues involved leave ample room for reasonable people to disagree. At the same time, the way they are often discussed in the popular press and by some observers clouds the key questions and overstates the differences that exist within the central banking community. All central banks agree that, in performing their monetary policy functions, asset prices should be taken into account to the extent that they affect the ultimate objectives, usually couched in terms of inflation and, less explicitly, output. In this sense, there is no question of "targeting" asset prices.

The possible disagreement concerns whether central banks should lean against booms in asset prices which appear to be unsustainable by tightening policy even if near-term inflation is under control. In research with some colleagues at the BIS, we have argued that such a response may be desirable, at least if the boom in asset prices coexists with an equivalent boom in credit, since this can signal the build-up of a "financial imbalance". Historically, such imbalances have tended to herald economic weakness, unwelcome disinflation and possibly financial strains further down the road as they unwind.

In recent years, arguments of this type have had some resonance within the central banking community. Central banks have become more keenly aware of the nature of the trade-offs involved. And a number of them have in different ways adjusted their frameworks so as to allow for the option of responding to perceived imbalances in this way. Some have even justified policy actions along these lines. The communication problems involved, however, remain daunting.

Looking at the kind of financial disruptions we saw in recent months, do you think central bankers have the analytical tools at their disposal needed to assess the impact on prices and the real economy? Or do these disruptions mean that the models "break down", if only temporarily?

By construction, the mainstream tools developed by central banks are best suited to assess economic developments in "normal times", when financial disruptions of the kind we have seen in recent months do not play a role. This is especially true of standard macro-models, which are not well suited to take systematically into account the multifaceted links between the financial system and the macroeconomy.

There is still much we have to learn about these interactions and the potentially large non-linearities involved, through which apparently small events can have disproportionate effects. This inevitably means that judgement and "soft" information have to play a larger role than in normal circumstances. I expect that this knowledge gap will narrow. A lot of analytical work is going on at central banks to link more systematically the financial and real spheres of the economy, largely as a result of the focus on financial stability. The development of macro-stress tests is just one such example. And as experience with financial disruptions accumulates, central banks will be able better to assess risks and hence to calibrate responses.

Even if the models were to prove useful, do you think there is a fundamental problem with respect to reliable, timely data on short-term financial market developments?

There is no dearth of data on financial prices; indeed, as a result of rapid financial innovation and the creation of new markets, we have never had as much data as we have now. This allows us to get a better feel for how market participants see the future and for how their risk appetite evolves.

By contrast, data on quantities are much harder to come by. For example, one factor exacerbating the current disruption was the lack of public information concerning what was contained in individual structured products and, above all, concerning the ultimate holders of the risks. The uncertainty associated with this opaqueness acted as a powerful catalyst for the generalised retrenchment from risk taking.

Here, I am personally less optimistic. While current gaps may well be narrowed, as the financial system continues to evolve, new ones are likely to develop. Policymakers will find it hard to keep up. Often, it takes a crisis to generate a sufficient consensus to gather costly data.

In the good times, my experience suggests that the incentives are not strong enough. Having said that, how much data you need depends on the purpose. In particular, before the current disruption, there was sufficient information available to indicate that risk-taking was likely excessive, although the precise nature and timing of the disruption were impossible to predict. Policymakers did issue warnings to that effect. To my mind, the issue is not primarily one of availability of information, but of the lens through which that information is examined.

"Moral hazard" is a phrase that has been mentioned a lot in recent months. What are the key considerations in distinguishing between a moral hazard-inducing bail-out - or the perception thereof - and prudent intervention by central banks and regulatory bodies?

This is a question that deserves a longer answer than can I can give you here. The trade-off between limiting, on the one hand, the short-run damage of financial turmoil and, on the other hand, the risk of inducing imprudent behaviour in the future is a very familiar one to the authorities in charge of addressing banking crises. In that context, the long-standing guiding principle to limit moral hazard has been to hold the shareholders, managers and, to the extent possible, creditors responsible and to make them suffer losses.

The recent turmoil has highlighted two novel aspects, which have been emerging as the nature of financial crises has been evolving. First, regulatory responsibilities are more diffused when the problems originate in markets, as opposed to specific institutions, and take initially the form of a generalised liquidity crunch. Second, the risk of unwittingly encouraging moral hazard can also arise in the performance of monetary policy functions, through the setting of interest rates. This is because, in practice, policies designed to address inflation and output objectives need to respond to financial disturbances, thereby partly insulating market participants from the effects of excessive risk taking.

In effect, this is a new form of "time inconsistency", akin to the one central banks had been used to in the context of inflation. In the years ahead, central banks will have to confront these questions head on and develop a set of principles to resolve them effectively. I suspect that a more pre-emptive use of monetary policy in the face of the build-up of financial imbalances may be part of the answer.

A few months ago leading central bankers were singing the praises of rapidly expanding markets for structured products, saying they spread risk rather than increase it in a systemic manner. Do you think recent events have fundamentally challenged this view?

Central bankers rightly point to the potential benefits of financial innovation in general, including those of structured products. At the same time, as I noted before, they have also pointed to the potential risks involved, notably complexity and opaqueness. Beyond nuances in tone, I believe there is a consensus concerning the long-run benefits of such financial innovations. The challenge is how to maximise those benefits while at the same time minimising the attendant risks, particularly during the transition phase, until products and markets become well established - part of the furniture, so to speak.

Looking ahead, when a crisis occurs, the main risk is that of an overreaction, especially under mounting political pressure. This risk has been magnified in the current turbulence, given that retail investors and borrowers have been so heavily affected. The challenge is to put the current innovations on a sounder footing, by articulating a system of adequate checks and balances while avoiding an excessive regulatory backlash. We should avoid throwing the baby out with the bath water.

Central banks invest an increasing portion of their resources on the research function, yet research output can take years to become relevant to policymaking. Moreover research is typically a public good, which can be produced elsewhere. And finally, PhDs in economics can be expensive to hire and retain. Given that central banks' budgets and spending patterns are always scrutinised closely and can be a political hot potato, how would you justify this increasing focus on research?

There are three reasons why research is an integral part of an effective central bank. First, research is a method. All policymakers' decisions are based on some model or vision of the world, whether consciously or not. A research function helps to inform those decisions, by seeking to accumulate information in a systematic, rigorous and structured way. Doing research within the central bank allows it to be best directed to the questions that policymakers care about. Experience indicates that the relevance of research in academia can sometimes be limited, because of the different incentives that influence priorities there.

Second, research is a communication tool. Internally, academic training allows staff to distill from academia the key policy insights and to tailor them to central bank requirements. Above all, externally, a research function allows central banks to be more effective in justifying their decisions vis-à-vis an increasingly demanding and sophisticated external audience, in which academics are particularly prominent, either directly or by influencing the terms of the debate. It can help to gain intellectual credibility. Finally, research is a motivational tool. It is instrumental in attracting and retaining some of the best talent. Economists with PhDs may be expensive, but these academic titles are among the most effective screening devices.

Several trends have put a premium on an own research function. The increased specialisation and compartmentalisation of academia sometimes works against the relevance of some of the work produced there. For example, finance, macroeconomics and the economics of financial crises have been proceeding quite independently of each other, but, as we were discussing before, understanding the world calls for a merger of their different perspectives. The increasing importance of transparency and accountability highlight the need for central banks to engage in the public debate. And the increased sophistication of jobs within central banks, allied with the general trend towards more time spent in higher education, mean that attracting PhDs has become more important. This is why we see an increasing number of governors and members of policy boards have PhDs and strong academic backgrounds.

Given that there are complementary - and even competitive - institutions generating research, in what areas and topics can central bank research departments add the most value? How important is research collaboration with universities and international financial institutions to a central bank's research department?

Central bank research adds most value when it focused on those questions that central banks care about but others are slow to pick up. A policy focus is essential. We should not forget, for instance, that the basic analytics of inflation targeting were initially developed by central banks. Likewise, central banks have taken the lead in work that seeks to integrate the financial and real sides of the economy in macro-models. Collaboration with universities is very important. It can help get academics interested in the questions central banks are interested in, thereby acting as a catalyst for more relevant work there. It can greatly improve the quality of the research at central banks. And it is can help identify, attract and retain high-calibre staff.

What can central banks do to ensure that they attract the very best graduates, who are likely to also be interested in careers in academia and the private sector?

This is a major challenge, especially at times when the financial sector is booming. The room for manoeuvre in terms of remuneration is very limited: central banks cannot compete with the financial sector or even parts of academia, such as business schools. The only possibility is to provide a stimulating environment in which to work. Ultimately, this means that the people that will be attracted to central banks - and be most successful - need to be personally motivated by a strong desire to contribute to policy. As they emerge from university, it is not common to find people with such a clear vision. This is also why a research function is important, as it provides an ideal environment in which to acquire a better taste for policy.