TL;DR

A prop-firm evaluation fee is spent once. An EA purchase is spent once. The MT5 Strategy Tester report is usually the only evidence you get before either payment. Audit that report first.

Free starting point: How to audit an MT5 backtest (free checklist) and/or the free VerifyEA tier. Escalate to VerifyEA Core ($149) when you need full offline automation (stability/OOS, Monte Carlo, firm presets). Keep journaling and research workflow separate via the $25 Trader Complete Toolkit if useful — it does not replace a report audit.

1. The fee is sunk; the report is a claim

Marketing for challenges and EAs leans on the same object: a smooth equity curve from Strategy Tester. Paying the fee does not buy you a verified edge. It buys you an attempt under rules that are strict about drawdown, not about how pretty the seller’s PDF looked.

Before you pay:

  1. Hold the raw HTML report, not a screenshot.
  2. Confirm modelling quality / tick model, spread, commission, and date range.
  3. Check concentration (few winners carrying profit) and structure (grid recovery, lot escalation, time clustering, identical sequences).
  4. Replay the closed-trade list against the firm’s current rules — and record the date you read those rules.

The longer method write-up lives in How to audit an MT5 backtest before you buy an EA.

2. Why “the backtest passed FTMO rules” is often overstated

Firms publish numbers you can test: profit target, daily loss, overall loss, minimum trading days, best-day caps where they apply. A trade list can be replayed against those numbers.

What the replay often cannot see: intraday floating equity. Many loss limits are equity-based. Closed-trade balance can miss a breach that would have failed the challenge live. On VerifyEA’s framing (product/docs language observed on the free checklist site): verdicts like balance-only pass mean no breach was visible on balance — not that none occurred on equity.

So:

  • FAIL on balance → strong reason not to pay the fee yet.
  • Clean balance replay → you cleared a screen, not a forecast.
  • Missing or conflicting rule values on the firm’s own pages → mark unknown; do not invent the convenient number.

Presets mentioned on the free method page include FTMO (1-step, 2-step), E8, The5ers, FundedNext, Alpha Capital, and Instant Funding — always re-check the live firm page.

3. Structural fingerprints that burn challenge accounts

Even when headline profit factor looks fine, challenge accounts die on path dependency:

  • Grid recovery — lot size roughly doubles after losses until one adverse run ends the account.
  • Lot progression — risk rises while the strategy is already losing.
  • Time clustering — edge lives in one narrow session window that may not match the evaluation calendar.
  • Identical sequences — repeated same symbol/direction/lots/timing; treat as a provenance question.

You can look for these by hand in the deals table (see the free checklist). Offline software exists so you are not doing every mechanical pass in a spreadsheet at 1 a.m. before a checkout timer expires.

4. Decision path (practical)

  1. Seller or your own EA sends an MT5 HTML report → run the free checklist: free checklist (full method host: audit-mt5-backtest.pages.dev). Optional parallel: QuantKit’s 12-check list.
  2. Want automation without upload → free VerifyEA (first 100 trades per product ladder copy on the Whop product page).
  3. Report is large / you need OOS + Monte Carlo + full presets → VerifyEA Core $149; details on the VerifyEA product page (also listed on the QuantKit Whop store).
  4. Still building personal process (journal, prompts, Python) → $25 Complete Toolkit on Gumroad / listed on /products.
  5. Only then pay the EA invoice or the challenge fee — if the open questions from the checklist are closed.

5. Cost context still matters after you pass

If you fund or trade live CFDs, all-in broker cost compounds. QuantKit’s measured EUR/USD round-turn ledger is separate research: broker cost study and graded CSV. Do not mix “cheap broker” with “verified EA.” They are different claims.

Keep reading

Same method — audit the report before cash leaves your account.

Disclaimer: Not financial advice. Not a prediction of challenge pass rates. Firm rules change; verify on the firm’s site and date your notes. QuantKit sells its own tools (Gumroad and Whop links above).