A Classic Case of "Data Snooping" for Classroom Discussion

Data snooping (mistaking spurious statistical relationships for genuine ones) is an important and dangerous by-product of financial analysis. However, data snooping is a difficult concept to explain to students of financial economics because, by its very nature, it is difficult to illustrate by example (a strong statistical relationship between complex financial variables is difficult to refute). To overcome this pedagogical difficulty, I present an example of data snooping where one variable is non-financial: I show that near both new moon and full moon, stock market volatility is higher and stock market returns are lower than away from the new or full moon. The simple and off-beat nature of this example enables substantial classroom discussion. INTRODUCTION