Passing a prop firm evaluation with an algo bot is different from passing manually. The rules are the same, but bots introduce specific failure modes that human traders rarely encounter. This guide covers the full process: choosing the right firm, configuring your bot for evaluation constraints, and avoiding the most common automated trading pitfalls.
Not all firms that allow automation are equal for your specific strategy. Before paying for an evaluation, answer these questions about your bot:
This is where most algo traders fail. They test their bot with one set of risk parameters and run the evaluation with the same parameters — without accounting for the evaluation's specific drawdown limits.
Calculate your bot's maximum expected drawdown from backtests, then size down until that drawdown is 60-70% of the firm's limit. This leaves buffer for live market conditions being worse than historical conditions.
Example: Apex $50K evaluation has $2,500 trailing drawdown. If your bot's backtested max drawdown is $1,800 at standard sizing, reduce size by 30% so your expected max drawdown is around $1,260. You now have meaningful buffer.
If you're at a firm with trailing drawdown (Apex, Topstep, MFFU), your bot needs to be aware that open profits reduce your effective buffer. Consider these adjustments:
Most evaluations require you to reach a profit target. For bots, this means you need enough trading days with the right market conditions. A few practical considerations:
One of the biggest advantages of algo trading at prop firms is the ability to run the same bot across multiple evaluation accounts simultaneously. Firms like Apex explicitly allow this.
Instead of buying one $250 evaluation, buy five $50 evaluations. Run your bot across all five accounts with account synchronization. The bot trades identically on all accounts. When some accounts pass and some don't (due to timing of entry on each account), you're effectively spreading the statistical variance of your evaluation outcomes.
This is mathematically more efficient than single large evaluations when your bot has consistent but not perfect expectancy.
Backtests often use fill prices that are better than live execution. In fast markets, your bot may get filled worse than expected. Always add a slippage buffer to your backtested results before sizing for a live evaluation.
If your bot loses connection to the broker API mid-position, you may end up with an open position and no active management. Set up alerts for disconnection and ensure your broker has appropriate stop-loss orders resting on the exchange.
A bot that looks great on historical data may fail on live data if the backtest was overfitted. Use out-of-sample data for validation before trusting a bot with a real evaluation.
Some firms restrict trading around major news events. Make sure your bot either avoids these windows or is configured to not trade during restricted times. An automated position opened during a prohibited window can result in account violation.
For most algo traders running their first prop firm evaluation:
Compare evaluation rules, costs, and automation policies across all five firms.
View Full Comparison Table →