Also called Genetic algorithm

Evolutionary search

Seeding a population of settings, keeping the strongest, then mutating and recombining them over generations to home in on what works.

Updated 2026-09-03

An evolutionary search starts with a population of random parameter sets, keeps the ones that performed best, combines and mutates them, and repeats. It borrows the mechanism of natural selection to explore a search space too large to enumerate.

It is the most powerful of the three optimisers here and by some distance the most dangerous. It is extremely good at finding whatever is best in the data you gave it, and a backtest's historical noise is part of that data. It will find the noise, because the noise is there and it is looking hard.

If you use it, hold back an out of sample period before you start and look at it exactly once. An evolutionary search on a single sample, judged on that same sample, is not an optimisation. It is a very sophisticated way of describing the past.

See also

  • Grid search — Testing every combination of the settings you chose. Exhaustive and certain, and the count explodes as soon as you add a third or fourth knob.
  • Random search — Sampling combinations at random instead of testing all of them. Cheap coverage of a space too large to walk, at the cost of possibly missing the best corner.
  • Curve fitting — Tuning a strategy until it memorises one stretch of history. It produces beautiful backtests and no future returns, which is why results are held out and re-tested.

← See this one in the deck

Educational content, not investment advice. Angie explains how a valuation is built so you can judge it yourself. What you do with that is your call.

IsiofiaIsiofia