Suppose you are optimizing a trading strategy, and it trades too frequently for your tastes. One way to reduce the amount of trading is to try adjusting the shortest and longest average trade span parameters found under the Optimization tab. Set the shortest parameter higher to reduce trades, and set the highest parameter lower to increase trading. The genetic algorithm is gently nudged in the proper direction.

However, setting average trade spans doesn’t always work well, especially if your strategy just doesn’t lend itself to change, no matter what the GA does. The GA isn’t FORCED to reject solutions that don’t meet the criteria, only encouraged to do so.

So here is another thing you can try that often will work even better. To reduce trading, increase commissions. To increase trading, reduce commissions. If the cost of making a trade is higher, the profit will go down as more trades are made, so the genetic algorithm will favor less trades. It doesn’t really matter that those may not be the costs you actually pay, because our goal is to get buy signals in the valleys and sell signals at the peaks. Whether or not we realistically represent actual profit is not nearly as important. The whole process of predicting the market is an estimation task anyway, any you shouldn’t be worried about great accuracy or precision.