Over 30 economic announcements are released world-wide on a typical business day. Some market participants suggest that these announcements are irrelevant to investors, as the markets learn to expect and price in the announcements well ahead of their release. Some companies accumulate and distribute economists’ forecasts, aggregated into “consensus” formats, in an effort to predict the announcements ahead of their release. Are these forecasts taking away the power of the announcements to move the markets? This article examines the issue.

First, a bit of economic theory. According to the Theory of Rational Expectations, economic announcements are largely incorporated in market prices by the time of the announcement. Only the “surprise” component, the announcement’s deviation from its forecast, is expected to move the markets on the day of the announcement and thereafter. The surprise is inevitably assumed to be small, and its impact on the markets to be negligible.

In practice, however, economists’ forecasts are substantially different than the actual values. Consider the following statistic: from January 2009 to present, the consensus forecasts of changes in major macroeconomic figures collected by a popular economic news distributor (not Reuters), were 42% different than their realized value. Even the direction of the change (up or down) indicated by the consensus forecast was correct in only 65% of cases. This correctness number drops further down for forecasts of forward-looking economic indicators, such as Consumer Sentiment. In the 2009-2010 period, only 55% of consensus forecasts of the impending changes in the U.S. Consumer Sentiment fell in the direction of the change actually announced. The average magnitude of the announced change well exceeded 100% of the economists’ consensus forecast.

How should investors to react to these changes? One route is to create independent models that are capable of more accurately forecasting economic activity—a labor-intensive and time-consuming exercise. Another way is to use history to estimate the response of securities to various announcements, wait until just after the announcement for the direction and magnitude of the economic change, and then invest with confidence.

Consider the historical quantitative impact of three major announcements to be released for November 2010: changes in the U.S. Money Supply, Monthly Changes in the U.S. Import Prices, and Consumer Sentiment Figures. The U.S. Money Supply figures are to be released on Nov. 12, Nov. 18, and Nov. 26, at 4:30 PM EST. The announcements document changes in the U.S. money available in circulation.

Changes in the U.S. Money Supply tend to affect a broad selection of U.S.
equities, fixed income and foreign exchange instruments. As a case study, consider equities forming the U.S. Consumer Discretionary sector ETF (ticker XLY). For every $1 billion in the money supply increase, XLY on average rises $0.01 on the day of the announcement (a typical money supply announcement registers a change of $17 billion, resulting in a roughly 0.5% directional gain in XLY).

A release of the Monthly Changes in the U.S. Import Prices is expected on Nov. 10, 2010, at 8:30 AM EST. As does the U.S. Money Supply, changes in the U.S. Import Prices affect the U.S. markets at large. In particular, Dow Jones Industrial index rises at least 50 points following a 1% rise in the U.S. Import Prices with probability of over 90% for the three weeks following the change announcement.

Last, but not least, changes in the U.S. Consumer Sentiment also significantly drive the markets. Historically, for every 2 point rise in Consumer Sentiment, the S&P 500 ETF (ticker SPY) rises 0.2% or about 16 cents per share on the day of the announcement with 94% probability relative to previous close. For every 2 point drop in Consumer Sentiment, SPY falls about 16 cents per share on the day of the announcement with 94% probability relative to previous closing price. (Relative to the same day’s opening price, SPY rises 20 cents on the day of the announcement with a 70% probability—while some news gets incorporated into SPY overnight, still lots of trading room for SPY just after the announcement is released). Holding SPY 2 to 4 days after the announcement results in approximately the same average gain, but the gain becomes less certain as time passes. The upcoming Consumer Sentiment figures are expected on Nov. 12 and Nov. 24, at 9:55 AM.

Irene Aldridge is a quantitative portfolio manager at ABLE Alpha Trading, LTD. She is also a co-author of the Quant Investor’s Almanac 2011: A Roadmap to Investing (Wiley, 2011), an up-to-the-minute compendium of economic announcements upcoming in 2011, and a summary of responses of various equities, foreign exchange and other securities to different economic announcements.

This entry was posted on Friday, November 5th, 2010 at 1:48 pm and is filed under Quant Speak. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.