More on the Cosmic Connection

[Guest guest]

Seven souls have boldly gone where no economists have gone before,

producing no fewer than three new papers

quantifying extra-terrestrial influences at work in world stockmarkets.

David Hirshleifer, of Ohio State University, and

Tyler Shumway, of the University of Michigan, have taken a long look at

the sun. Across 26 stockmarkets, in 1982-97,

they found that, on days with more sunshine in the morning than usual,

shares generated above-average gains, and lower

gains on unusually cloudy days. After taking sunshine into account,

other weather conditions such as snow or rain turned

out to have no impact on share returns.

 

The effect of the sun seems to have been quite big. For shares traded in

New York, for instance, annualised returns on

perfectly sunny days averaged 24.8%, compared with 8.7% on perfectly

cloudy days. Moreover, unlike some stockmarket

“anomalies” discovered by economists, investing by the sun would have

been more profitable than simply investing in the

market index, even after subtracting trading costs. Alas, this does not

guarantee profitable trading in the future. Even

assuming the effect really exists, investors, now they have been alerted

to it, will probably arbitrage away any excess

returns to be had by exploiting it.

 

The other two papers, both published by the University of Michigan, look

a little closer to home, at the moon. Kathy

Yuan, Lu Zheng and Qiaoqiao Zhu examined share returns in 48 countries

until the end of July 2001, from start dates as

far back as 1965. They found that, on average, daily returns were much

higher around new moons than full moons, when

investors are presumably sprouting fangs and howling rather than buying

shares. Returns were 8.3% lower in weeks with a

full moon than ones with a new moon. The lunar effect held in 43 of the

48 countries, and investing by it would have been

profitable even after trading costs.

 

Ilia Dichev and Troy Janes reach a similar conclusion from a study of

100 years of American share prices, and at least 15

years' data in each of 24 other countries. Daily returns around new

moons were roughly double those around full moons.

The lunar effect is strongest outside America. All seven economists

insist that their results are not the product of

data-mining—crunching the numbers in search of any interesting pattern

they can find. Rather, the results are the latest

manifestation of behavioural finance, which seeks to show how

psychological patterns explain apparent market

inefficiencies.

 

Still, although it is widely believed that the phases of the moon affect

behaviour, psychologists have yet convincingly to

prove it. The behavioural economists reckon this may be because

psychologists have focused on trying to link the moon

to extreme behavioural problems in a few disturbed people, rather than

to more humdrum lunacies affecting humanity as a

whole, including a bias against shares around full moons.