"The central tendencies of market-based forecasts are at least as accurate, and in fact somewhat superior to, the 'consensus' or 'survey' forecast derived by taking the average estimate from a survey of forecasters."
Economic derivatives, which invite investors to purchase options based on macroeconomic activities, have been traded for less than four years. The payoff from these options depends on macroeconomic outcomes, such as growth in GDP and non-farm payrolls, inflation, the international trade balance, retail sales, and business confidence. In this market, "digital" or "binary" options are traded, which means that traders purchase a security that is worth $1 if, for instance, monthly employment growth was between 100,000 and 125,000 jobs - otherwise, the security is worth nothing. The prices of these options provide market-based measures of investors' expectations about the likelihood of various outcomes. These active "macro markets" (see www.economicderivatives.com for more information) potentially allow for better allocation of risk and for enhanced investor protection against macroeconomic risks.
In Macroeconomic Derivatives: An Initial Analysis of Market-Based Macro Forecasts, Uncertainty, and Risk (NBER Working Paper No. 11929), Refet Gurkaynak and Justin Wolfers analyze the data derived from the first few years of this new market, focusing on the forecasts for non-farm payrolls, initial unemployment claims, retail trade, and business confidence. Their main finding is that the central tendencies of market-based forecasts are at least as accurate, and in fact somewhat superior to, the "consensus" or "survey" forecast derived by taking the average estimate from a survey of forecasters. In addition, the authors report that financial market responses to economic news are better captured by market-based expectations than by the survey-based measures; again, this suggests that the financial market outcomes are better at capturing investor expectations. Further, these researchers find that some behavioral anomalies present in survey-based expectations, such as predictable forecast errors, are notably absent from market-based forecasts.
Gurkaynak and Wolfers note that the economic derivatives market establishes prices for options based on numerous and varied outcomes. Thus, the economic derivatives market allows researchers to derive not only a single "best" forecast, but also a measure of the uncertainty around such a forecast. Previously researchers had analyzed data on disagreement among forecasters, hypothesizing that disagreement was a reasonable proxy for uncertainty. While uncertainty actually measures the likelihood and extent to which the economic outcome might differ from the central estimate, disagreement only measures how much the central estimates offered by different forecasters differ. Gurkaynak and Wolfers compare the measure of uncertainty that is implicit in the economic derivatives data to the measure of disagreement that can be extracted from the survey data. They find that, while there is some correlation between the two, on a release-by-release basis, disagreement is not a particularly good proxy for uncertainty. P>
Beyond capturing uncertainty, economic derivatives provide detailed information on the market's assessed likelihood of a full range of outcomes occurring. Historically, it has been quite rare to find such "density forecasts." Gurkaynak and Wolfers proceed to analyze their data in terms of the efficacy of these option prices as density, or probability, forecasts. If the price of an option paying $1 if a specific economic outcome occurs is twenty cents, does this suggest that the chance of the outcome occurring is 20 percent? Their findings suggest that the answer is yes, and that economic derivatives yield efficient density forecasts, which they note is a rarity.
When applied to market-based measures, the researchers' density-forecast-efficiency tests jointly test efficient pricing and the absence of risk premiums. Yet it might seem reasonable that risk aversion would lead investors to bid up the prices of particular options, so as to insure against particularly bad outcomes; this would lead a risk premium to drive a wedge between prices and probabilities. The researchers' finding that economic-derivatives-based densities are efficient thus indicates that risk premiums in this market are probably small. This also allows them to investigate the degree to which the pricing of economic derivatives can be used to estimate investors' risk aversion.
The fact that risk premiums in this market are generally small also brings home the point that while these markets currently provide some protection against "event risk" -- the possibility that a portfolio's value may change sharply when economic data are released -- until these markets are expanded to allow taking positions on longer-term outcomes, they do not provide much protection against macroeconomic downturns.
By using the institutional structure of economic derivatives to study risk and risk aversion, Gurkaynak and Wolfers surmise that economic derivatives are promising instruments for economists who would like to consider the relationship between investor's beliefs, risk attitudes, and asset prices. They conclude by noting that their paper is "an initial exploration [that] showed that economic derivatives correctly capture subjective beliefs and provided some applications of this information. Having these subjective probabilities will facilitate future research to study how expectations are formed and how they relate to actions, as well as to analyze agents' responses to occurrence of events of different prior subjective probabilities.
-- Matt Nesvisky