The information contained in asset prices is useful to central banks for several reasons. Asset prices may embody information about macroeconomic aggregates that is more accurate and up-to-date than what is currently available to policy-makers. Aberrations in some asset prices may indicate imperfections or manipulations relevant for banking and financial market surveillance. And asset prices reflect market participants’ expectations about the future.

In a recent CEPR Discussion Paper, Paul Söderlind and Lars Svensson survey some of the latest techniques for extracting market expectations about interest rates, exchange rates and inflation from asset prices. These techniques, which have become feasible as a result of deeper and more developed financial markets, are increasingly used by central banks for monetary policy purposes.

Traditionally, interest rates have been used to extract expected averages of future interest rates, exchange rates and inflation. For example, a two-year interest rate can be interpreted as the expected average of short interest rates over the next two years plus a risk (or term) premium.

More recently, these methods have been refined to take advantage of implied forward interest rates in order to extract expected future time-paths. For example, by combining a one-year and a two-year interest rate, the expected average of short interest rates over the next twelve months can be separated from the expected average of short interest rates (plus risk premia) over the twelve months starting a year ahead.

Very recently, methods have been designed to extract the whole probability distribution of market expectations from a set of option prices. While a long interest rate reflects the value that the market expects the interest rate to be at some time in the future, options can reveal how uncertain market participants are in their guess, and what probability they attach to a decrease of more than x% or an increase of more than y%.

A call option gives the holder the right (but not the obligation) to buy a one-year bond, for example, for a pre-specified strike price at the expiry date of the option one year ahead. The pay-off of the option is zero if the bond price on that day turns out to be lower than the strike price. But it is positive if the bond price is higher than the strike price.

The option is therefore a bet on the interest rate being lower than that implied by the strike price. The price of this bet reveals something about the probability the market attaches to low interest rates. Options are typically traded for a set of different strike prices. They can be combined to trace out most of the probability distribution of the one-year interest rate one year from now.

These techniques seem capable of shedding light on many issues of concern to central banks. Market expectations for short interest rates over the next six to nine months, say, can be seen as market expectations of future monetary policy, when a short interest rate is used as the instrument of policy.

With such short horizons, the risk premia are likely to be relatively small, so forward rate curves for settlements up to six to nine months could be useful. Estimated probability distributions within the same horizon will also be useful for indicating the degree of market uncertainty about monetary policy in the near future.

In small open economies with floating exchange rates, the exchange rate channel is an important part of the monetary transmission mechanism. Indeed, some central banks measure the impact of monetary policy through monetary conditions indices (MCIs), weighted combinations of interest rates and exchange rates.

Then it is natural to extract exchange rate expectations for the same horizon, and use those together with interest rate expectations to construct expected future MCIs. Again, estimated exchange rate distributions will indicate the uncertainty in the market’s exchange rate expectations. Together with the distribution of interest rate expectations, these allow estimates of the distribution of future MCIs.

In the medium and long term, say beyond two to three years, interest rate expectations, especially if there are sufficiently liquid indexed bonds, will be useful as indicators of inflation expectations. These are always of interest to central banks, particularly if the bank has an explicit inflation target.

In these circumstances, the difference between inflation expectations and the inflation target is an obvious measure of the (inverse) credibility of the monetary policy regime. Forward rates from nominal and indexed bonds should contain useful information.

For these horizons, estimated exchange rate expectations can also, to some extent, be interpreted as indicating the degree of credibility of the monetary regime. A large expected depreciation will obviously tend to be associated with high inflation expectations and low credibility.

For fixed exchange rate regimes or exchange rate target zones, exchange rate expectations for all horizons are of particular relevance as indicators of the credibility of the regime.

This article reviews research reported in ‘New Techniques to Extract Market Expectations from Financial Instruments’, CEPR Discussion Paper No. 1556 (January 1997) by Paul Söderlind and Lars E O Svensson. Söderlind is at the Stockholm School of Economics and a Research Affiliate in CEPR’s International Macroeconomics programme; Svensson is at the Institute for International Economic Studies in Stockholm and a Research Fellow in CEPR’s International Macroeconomics programme.

RELATED BOOKS FROM CEPR

Leonardo Leiderman and Lars E O Svensson (eds), Inflation Targets, CEPR (1995)