Stock markets' fear of falling

The sharp drop in the world's stock markets on August 9, after BNP Paribas announced that it would freeze three of its funds, is just one more example of the markets' recent downward instability or asymmetry.

That is, the markets have been more vulnerable to sudden large drops than they have been to sudden large increases.

Daily stock price changes for the 100-business-day period ending August 3 were unusually negatively skewed in Argentina, Australia, Brazil, Canada, China, France, Germany, India, Japan, Korea, Mexico, the US and the UK.

In the US, for example, the Standard and Poor's 500 index in last month's recorded six days of declines and only three days of increases amounting to more than one per cent.

In June, the index dropped more than one per cent on four days, and gained more than one per cent on two days. Going back further, there was a gigantic one-day drop on February 27 of 3.5 per cent, and no sharp rebound.

The February 27 decline began with an 8.8 per cent one-day drop in the Shanghai Composite, following news that the Chinese government might tax capital gains more aggressively. This news should have been relevant only to China, but the drop there fueled declines worldwide. For example, the Bovespa in Brazil fell 6.6 per cent on February 27, and the BSE 30 in India fell four per cent the next day. The subsequent recovery was slow and incremental.

In the US, the skew has been so negative only three other times since 1960: at the time of the 6.7 per cent drop on May 28, 1962, the record-shattering 20.5 per cent plummet on October 19, 1987, and the 6.1 per cent decline on October 13, 1989.

Stock markets' unusually negative skew is not inconsistent with booming price growth in recent years. The markets have broken all-time records, come close to doing so, or at least done very well since 2003 ( the case in Japan) by making up for the big drops incrementally, in a succession of smaller increases.

Nor is the negative skew inconsistent with the fact that world stock markets have been relatively quiet for most of this year. With the conspicuous exception of China and the less conspicuous exception of Australia, all have had low standard deviations of daily returns for the 100-business-day period ending August 3 when compared with the norm for the country. The February 27 drop in US stock prices was only the 31st biggest one-day drop since 1950. But all of the other 30 drops occurred at times when stock prices were much more volatile. Thus, the February 27 drop really stands out, as do other recent one-day drops.

Indeed, one of the big puzzles of the US stock market recently has been low price volatility since around 2004, amid the most volatile earnings growth ever seen. Five-year real earnings growth on the S&P 500 set an all-time record in the period ending in the first quarter of 2007, at 192 per cent. Before that, between the third quarter of 2000 and the first quarter of 2002, real S&P 500 earnings fell 55 per cent - the biggest-ever decline since the index was created in 1957.

One would think that market prices should be volatile as investors try to absorb what this earnings volatility means. But we have learned time and again that stock markets are driven more by psychology than by reasoning about fundamentals.

Is psychology somehow behind the pervasive negative skew in recent months? Maybe we should ask why the skew is so negative. Should we regard it as just chance, or, when combined with record-high prices, as a symptom of some instability?

The adage in the bull market of the 1920s was "one step down, two steps up, again and again." The updated adage for the recent bull market is "one big step down, then three little steps up, again and again," so far at least. No one is looking for a sudden surge, and volatility is reduced by the absence of sharp up-movements.

But big negative returns have an unfortunate psychological impact on markets. People still talk about October 28, 1929, or October 19, 1987. Big drops get their attention, and this primes some people to be attentive for them in the future, and to be ready to sell if another one comes.

In fact, willingness to support the market after a sudden drop may be declining. The "buy-on-dips stock market confidence index" that we compile at the Yale School of Management has been falling gradually since 2001, and has fallen especially far lately.

The index is the share of people who answered "increase" to the question, "If the Dow dropped three per cent tomorrow, I would guess that the day after tomorrow the Dow would: Increase? Decrease? Stay the Same?" In 2001, 72 per cent of institutional investors and 74 per cent of individual investors chose "increase".

By May 2007, only 48 per cent of institutional investors and 59 per cent of individual investors chose "increase".

Perhaps the buy-on-dips confidence index has slipped lately because of negative news concerning credit markets, notably the US sub-prime mortgage market, which has increased anxiety about the fundamental soundness of the economy.

But something more may be at work. Everyone knows that markets have been booming, and everyone knows that other people know that a correction is always a possibility.

So there may be an underlying sensitivity to price drops, which could fuel a succession of downward price changes, amplifying public concerns about problems in the economy and heralding a profound change in investor sentiment.

The author is professor of Economics at Yale University, chief economist at MacroMarkets LLC, which he co-founded and author of Irrational Exuberance and The New Financial Order: Risk in the 21st Century.

© Project Syndicate, 2007.
www.project-syndicate.org