Deciphering the determinants of stock market volatility

• Pages: 132-134

Page 132

Since the dawn of modern stock exchanges in early seventeenth-century Europe, periodic bouts of stock price volatility have puzzled economists and noneconomists alike. Commenting on one such episode-the boom and bust of the famed South Sea company shares in 1720 England-Sir Isaac Newton is said to have remarked, "I can measure the motion of bodies, but I cannot measure human folly." An important strand of current research focuses on the role of country location and industry affiliation in determining stock price movements. In a new IMF Working Paper, Luis Catão of the IMF's Research Department and Allan Timmerman of the University of California, San Diego, gauge these effects.

The extent to which country and industry factors influence stock returns has important practical implications.

If stock returns are primarily determined by country location, investment risk can be reduced by holding an internationally diversified portfolio. If industry factors are paramount, crossing national borders will yield meager gains, and risk diversification can more easily be accomplished by investing in different industries in a given country. Catão and Timmerman use a new methodology (see box, page 133) to examine these relationships and find that the benefit of exiting from national stock markets and buying foreign stocks declines dramatically during periods of high global volatility, such as the current one.

Their study also supports the so-called geography or cultural view of financial markets-that is, that stocks tend to move together more tightly across English-speaking countries (and notably so between the United Kingdom and the United States) and much of continental Europe, but much less so between, say, Japan and the United States. This is true during periods of both high and low volatility and implies that being purely Anglophile or Europhile is not beneficial, at least when it comes to diversifying equity risk.

Nature of volatility

Applying their methodology to a global data set of monthly returns for about 4,000 firms from 1973 to 2002, Catão and Timmerman find overwhelming evidence of two distinct and well-defined "states" of high and low volatility. Overall market volatility (countryand industry-specific volatilities plus a "global" volatility component common to all firms) is highest around the time of the first oil shock (1973-75), the second oil shock...