Friday, 24 July 2009

Correlation of the Week: Eclipses and the economy

In the paper Dark Omens in the Sky: Do Superstitious Beliefs Affect Investment Decisions?, Gabriele Lepori from Copenhagen Business School has correlated the occurrence of solar and lunar eclipses with four American stock indices: the Dow Jones Industrial Average, the S&P 500, the New York Stock Exchange Composite, and the Dow Jones Composite Average.

This is possibly the most fun, and most mathematically rigorous, Correlation of the Week we have had. I love it!

Lepori's idea was to test whether superstitious practises could be picked up in the stock-market. His theory was that individuals are more likely to resort to superstitious practices when operating in environments dominated by uncertainty and high stakes. The stock market is therefore an ideal place to test this theory.

Eclipses are regarded as "unlucky" in many Western and Asian societies, and as they are worldwide events occurring over a short period of time, the effects of such unlucky events should be seen in stock trading - if there is an effect, that is.

Lepori took the dates of all 362 lunar and solar eclipses that had been visible anywhere in the world between 1928 and 2008, and computed the four stock indicies on the day of the eclipse. The results suggest that in the three days around the date of an eclipse, three of the four stock indices exhibited statistically significant lower-than-average returns. If an eclipse took place on a weekday (when the stockmarkets were open), its effect was larger than if it occurred on a weekend. And the greater the magnitude of the eclipse, the more likely it would influence stock returns!

On the days following the eclipse, Lepori showed that markets reverse the eclipse-related dip, suggesting that the market realised the drop was irrational. Here is a perfect opportunity for astute investors to make some money - buy stocks at the eclipse maximum and sell a few days later! Or sell a few days before the eclipse and buy back at its maximum. This is known as arbitrage. Indeed, Lepori found that an investor who had bought the Dow Jones Industrial Average at the end of 1928 would have multiplied their money 37 times by now. However, one who sold before each eclipse and bought back straight after would have multiplied their money by 55.

Lepori's theory is consistent with the idea that at the time of an eclipse, there is less buying pressure coming from the superstitious. Trading volume also decreases. These patterns are inconsistent with Efficient Market Theory, which states that traded assets reflect all known information, and instantly change to reflect new information. This means that it is impossible to consistently outperform the market by using any information that the market already knows - eclipses are perfectly predictable events and so even if they did influence stock prices, this should already be built into the price.

I recommend a read of the original paper - check out the mathematical rigour and the lengths to which Lepori went to control for other variables such as day of the week, media coverage of the eclipse and even the weather! Fantastic stuff!

I'm not sure whether this gives me a more positive or negative opinion of the stock market. On one hand, it shows how complex the market is and why we must employ smart people (and pay them loads) to understand it. On the other hand, this study shows that completely irrational beliefs and behaviours can influence how we price things. This surely undermines the system. In the modern globalised world, everything is governed by finance, and allowing irrational thought to influence world affairs scares me! But maybe that shouldn't surprise me - wars are started over personal disagreements, politics is dominated by petty arguments and some people still doubt evolution - I guess the world just ain't rational!