Fees and slippage: why costs change the verdict

A fee is a known quantity. Whatever your exchange charges, you can look it up and enter it once.

Slippage is not knowable in advance. It depends on your order size, how liquid the pair is, and how fast the market was moving at that moment. The default of zero is therefore the optimistic case rather than a neutral one, and on a thin pair or a violent move it is optimistic by a lot. Setting it to 0.05% or 0.1% and re-running is a quick way to see whether a result depends on getting perfect fills.

A trade that gained 0.15% before costs closes down after 0.1% a side, and the win rate counts it as a loss. That is the correct answer: the money is the money.

Because the charge is per trade, the effect scales with how often a strategy trades. A rule taking four positions a year barely notices it. The same rule on a faster timeframe might take hundreds, and at that point the cost setting moves the result as much as the rule does. Neither outcome is decided in advance, which is the reason to test the same rule at more than one speed rather than assume it carries over.

One detail about the buy-and-hold comparison: it pays an entry-side fee too. A completely flat market therefore benchmarks a fraction below zero rather than exactly at it.