The paper looks at how superstitions affect financial markets by examining limit orders (an offer to trade shares at a given price) in the Taiwanese stock market. The authors measured the strength of traders’ superstitions by the regularity with which they chose prices ending in eight, and how strenuously they avoided the number four. Institutional investors and foreigners were generally unbiased in this regard, with no numerical pattern in their prices. Local individuals, however, were more than 50% more likely to place an order ending in an eight than in a four.

Such irrationality costs money: the daily returns of the most superstitious quintile of investors were 0.03% lower than the least superstitious on trades ending in an eight. That equates to an annual shortfall of—by coincidence—8.8% a year, according to The Economist’s calculations. Avoiding four was less costly, but still diminished returns by around 2% a year. The cost of the bias extended beyond trades involving the numbers four and eight. An increase in superstitious beliefs of one standard deviation led to a 0.02% decrease in daily returns (4% over the course of the year, another coincidence). This suggests that superstitious beliefs are associated with other biases which lower profits for traders more broadly. Some of these are also numerical, such as investors’ preference for round numbers. The paper also alludes to superstitious traders having a “general cognitive disability in financial decision making”, a diplomatic way of saying they are nitwits.

Small investors are known to succumb to overconfidence, spurious extrapolation and a preference for the familiar. These biases make them an easy target for algorithmic traders, or “quants”, who exploit such predictable behavior.

Fortunately for Taiwanese day-traders, the researchers find that digit prejudice abates over time. Perhaps tired of losing money, investors tend to ditch their superstitions as they gain experience.