Trading Costs: 3 Hidden Traps That Kill Good Strategies
In our tests, trading costs turned positive edges into losses in three different ways. Here are the numbers, the broker spreads we measured and a simple way to compute cost per trade in R.

In our tests, trading costs turned positive edges into losses in three different ways. Here are the numbers, the broker spreads we measured and a simple way to compute cost per trade in R.
A seller’s backtest tells you what the seller wants you to see. Here are the 10 checks we run in every independent trading bot review, the red flags we look for, and what our report contains.
Overnight financing looks tiny on a statement, yet in our tests it turned every monthly trend-following setting negative. Here is how swap fees work, how to measure them in R, and why long holds suffer most.
We turned Smart Money Concepts and Malaysian SnR into mechanical rules and tested 34 settings on 7 markets with real broker spreads. None came close to our pass bar, and an honest look at limit fills erased the best result.
Our Asian-session mean reversion test made money in all 12 settings before costs and lost in all 12 after the spread. Here is the arithmetic behind it, the spreads we measured, and what would need to change.
We tested 39 classic strategies, 25 suggested improvements and 806 rule settings on broker data after costs. Nothing passed. These are the seven lessons we took from it.
Most backtests grade their own homework. These are the four out-of-sample testing rules we use at TradeLab Research, what they caught, and a checklist you can use today.
A good backtest can still be pure luck. Here is the coin-flip luck test we run on every strategy, the trend-day rule it exposed, and five steps to run one on your own trades.
Seven large studies, from European regulators to Taiwan’s day traders, show how few retail traders win. We add our own 806-setting search and explain why costs decide most of the outcome.
Breakouts, taker flow, funding rates and weekday effects: we tested four ideas for a crypto trading strategy with real costs and luck tests. None passed, and the reasons are worth knowing.
Send us your bot or strategy and get an independent report that uses your broker's costs.