← Research logMachine Learning
KILLED · overfit by design
#

Evolving populations of indicator-threshold rules with a genetic algorithm discovers profitable combinations no human would hand-craft.

Fitness-proportional selection over generations:

$$ P(\text{rule}_i) = \frac{\phi_i}{\sum_j \phi_j},\quad \phi=\text{in-sample Sharpe} $$

GA over indicator/threshold/logic genes, fitness = in-sample Sharpe, 200 generations, then hold-out test.

Best in-sample Sharpe4.1
Same rule out-of-sample−0.3
Multiple-testing inflationextreme
KILLED
Evolving toward in-sample Sharpe is industrial-scale data snooping: searching millions of rule combinations guarantees a spectacular in-sample winner that is pure overfit. Out-of-sample it is negative. The GA optimizes the one thing you must not optimize directly.
If your search procedure maximizes in-sample performance over a huge rule space, your "discovery" is the maximum of noise. Significance must be penalized by the number of trials.

We publish the failures too.

This is one of 100+ documented hypotheses. Browse the full lab notebook, or see the strategies that survived into production.