08-12-2026, 08:50 PM
A strategy that ignores trading friction is usually measuring an optimistic fantasy. Model costs before comparing ideas.
A basic estimate is:
Start conservatively. Use a round-trip commission, half-spread on entry and exit, and slippage that increases with volatility or order size relative to typical volume. For less-liquid instruments, test several adverse scenarios instead of one fixed number.
Also report turnover. A small edge traded too frequently can disappear after costs. Re-run the backtest with 1x, 1.5x, and 2x your baseline cost estimate; a robust idea should not collapse immediately.
How do you stress-test cost assumptions in your research?
A basic estimate is:
Code:
net_return = gross_return - commission - spread_cost - slippage - fundingStart conservatively. Use a round-trip commission, half-spread on entry and exit, and slippage that increases with volatility or order size relative to typical volume. For less-liquid instruments, test several adverse scenarios instead of one fixed number.
Also report turnover. A small edge traded too frequently can disappear after costs. Re-run the backtest with 1x, 1.5x, and 2x your baseline cost estimate; a robust idea should not collapse immediately.
How do you stress-test cost assumptions in your research?
Building small AI tools and automations in Bengaluru. Notes on what ships, what fails, and what I'm learning along the way.