From Backtest to Funded Account: The Validation Checklist Prop Traders Need Before Paying
Challenge fees are $200–$1,000+ per attempt, and ~90% of attempts fail. This checklist — run before you pay, not after — is the cheapest money in prop trading.
A prop challenge fee is $200–$1,000+ per attempt, depending on account size. About 90% of attempts fail. That means the average trader pays several fees before ever seeing a funded account — or never does. The cheapest money in prop trading is the validation you run before you pay, not after.
Here is the checklist, in order. Skip a step and you are paying to discover what that step would have told you.
1. Recompute from raw predictions
If your strategy's numbers come from a claimed-accuracy column, a vendor's screenshot, or a backtest you cannot reproduce from raw data, stop. Recompute accuracy from the raw predictions yourself. If the recomputed number is much lower than the claimed one, you just saved a challenge fee.
2. Beat the dumb baselines
Always-bullish and no-change, on the same data. If your model cannot beat both, it adds nothing — it is riding the base rate of up-moves. On a bare CSV with no price history, those two baselines are the honest ones.
3. Check walk-forward folds, not the average
A single overall accuracy hides decay. Demand fold-by-fold results, earliest to latest. A model decaying from 80% to 56% across folds is a dying edge with a flattering average — and it will die faster under challenge pressure.
4. Count the trials and correct for them
How many configurations did you try before this one? If the answer is “hundreds” and no multiple-testing correction was applied, the headline number is partly luck. Apply a Bonferroni-style correction, or at least discount the result honestly.
5. Model costs and slippage
Backtest with realistic spread, slippage, commissions, and taxes. A strategy whose edge survives only at zero friction does not survive.
6. Backtest against the firm's exact rules
Daily drawdown from intraday peaks (5% at FTMO), total drawdown (10%), profit target (10% phase 1, 5% phase 2), minimum trading days. Run the strategy through those constraints historically. If it breaches the daily cap in backtest, it will breach it live — with a fee attached.
7. Size for the worst streak, not the average
Set risk per trade at 0.5–1%, so a normal four-loss streak stays inside the daily cap. Then set a personal daily stop below the firm's cap (stop at 3% down, not 5%).
8. Forward-test on paper first
Two to four weeks of live paper trading under the same rules. Backtesting tells you if a strategy could have worked; forward testing tells you if it actually functions — with real data flow, real fills, and real psychology.
9. Only then, pay
One challenge, sized to the account you can afford to lose the fee on, with the validated strategy and the validated risk parameters.
Most traders do this list backwards: pay first, discover the failures with money. The traders who pass are usually the ones who did the boring work first. Validation is not the opposite of ambition. It is the cheapest form of it.
BuildIT Index runs steps 1–5 mechanically on any prediction file: recompute accuracy, beat baselines, check folds, count trials, model costs, return a verdict with an exit code.
Validation is not the opposite of ambition. It is the cheapest form of it.
This article is educational, not financial advice. Past results do not predict future returns.