Tested backtesting fidelity across 4 options platforms with the same iron condor
Proud of how rigorous I made this test because the conclusion is messier than I expected. Built the same iron condor strategy in four platforms, ran the same window (Jan 2023 to Sep 2024), same SPX, same wing widths (10-delta short, 5 wide), same DTE entry (45), same 50% profit close. Ran the equity curves and trade-by-trade output side by side.
Setup:
OptionBots: bar-level fill simulation, integrated build-test-deploy
Option Alpha: bar-level fill simulation, deeper trade history available
tradeSteward: tick-level (1-second) fill simulation, backtest-focused
OptionStrat: analytical backtester, less full-strategy oriented
Results across the same strategy, same window:
OptionBots
Total return: 18.3%
Win rate: 76%
Average winner: $142
Average loser: -$485
Max drawdown: -7.2%
Option Alpha
Total return: 19.1%
Win rate: 77%
Average winner: $138
Average loser: -$471
Max drawdown: -7.0%
tradeSteward
Total return: 16.4%
Win rate: 73%
Average winner: $135
Average loser: -$498
Max drawdown: -8.1%
OptionStrat
Total return: 20.7%
Win rate: 78%
Average winner: $144
Average loser: -$462
Max drawdown: -6.8%
Key finding: the spread across platforms is 4.3% on total return for the same exact strategy. That's bigger than I expected. tradeSteward came in lowest, OptionStrat highest, OptionBots and Option Alpha in the middle and within 1% of each other.
What's actually causing the spread:
tradeSteward's tick-level simulation catches stop-outs and intra-bar adverse moves the bar-level platforms read as wins. This makes its backtest more conservative and probably more realistic.
OptionStrat is the most generous because its fill assumptions don't model adverse intraday slippage as aggressively as the others.
OptionBots and Option Alpha land in the middle. Bar-level resolution that handles open-high-low-close adversely enough to be reasonable but not as conservative as tick-level.
Practical takeaways:
If you're picking a platform for live execution, OptionBots and Option Alpha are roughly equivalent on backtest fidelity for most strategies. Decide on platform UX, broker fit, and pricing, not on backtest accuracy.
If you're doing pure research where intraday timing matters (tight stops, scalping options, 0DTE single-direction trades), tradeSteward's tick-level resolution is the right tool. Run your final backtest there before committing the strategy.
Don't trust any single platform's backtest. The 4.3% spread across these four for the same exact strategy is the lesson. Backtest in two and average if you can.
What I'm doing personally: live execution stays on OptionBots, final research backtests run on tradeSteward, periodic sanity checks on Option Alpha. That covers UX, fidelity, and second-opinion. NFA.