P-Hacking

P-hacking, or data dredging, is the misuse of data analysis to find patterns in data that can be presented as statistically significant, when in fact there is no real effect. This is often done by testing many different variables or hypotheses until one yields a favorable p-value, without adjusting for the multiple comparisons made.

In finance, this is a dangerous practice because it creates the illusion of a trading strategy with a strong statistical edge. In reality, these results are just the product of chance.

P-hacking leads to the deployment of strategies that fail immediately upon live trading. It is a significant ethical and practical concern in quantitative research and academic finance.