Most prop-firm challenges are not failed by the strategy. They are failed by the risk plumbing around the strategy — a single trade that runs past the daily-loss line, a martingale recovery sequence that detonates on a news spike, an EA that keeps trading through Friday rollover into the weekend gap. As a prop firm EA developer who has built MetaTrader Expert Advisors since 2017, my job is to make the EA respect your firm's rules as hard constraints, not suggestions.
Let me be precise about what this page is, because this is a YMYL topic and you deserve straight talk: I do not sell a "challenge-passing EA." No honest developer can. What I build is a compliant risk architecture — code that mathematically cannot breach the limits you give it — plus verifiable backtests so you can see how the logic behaves before a single dollar of your evaluation fee is at risk.
Why most EAs fail prop-firm challenges
I have audited and rebuilt enough broken EAs to see the same failure modes repeat. If you understand these, you understand exactly what I'm protecting you against:
- No equity-based drawdown guard. Most firms measure daily loss and max drawdown on equity (including floating/open P&L), not closed balance. An EA that only checks closed trades will happily sit on a -6% floating position while you have already breached a 5% daily limit. The guard has to read equity in real time.
- Martingale and aggressive grid. Doubling into losers produces a beautiful equity curve right up until the trade that wipes the account. It also violates the terms of several firms outright. I do not ship it — more on that below.
- Over-optimized curve-fitting. An EA tuned to print perfect results on three years of one symbol is memorizing history, not finding edge. It dies in forward conditions during your evaluation.
- Ignoring news and weekend rules. Holding through high-impact news or over the weekend is restricted or banned at several firms. No filter means an automatic, avoidable breach.
- Trailing-stop and execution assumptions that don't survive real spread/slippage. A backtest on ideal ticks is not live performance — I'll say that more than once because it's the single most expensive misunderstanding in this business.
Common prop-firm rules and how the EA enforces each
Rules vary by firm and account size, and firms change them — always verify the current rulebook with your firm before you start an evaluation. The table below shows the typical constraint categories and the concrete mechanism I build into the .mq5 to honor each one.
| Prop-firm rule (typical) | What it means | How the EA enforces it |
|---|---|---|
| Daily loss limit (e.g. 4–5%) | Equity drop from the day's starting balance/equity | Records the daily anchor at server-day rollover; on every tick checks (anchor − equity); closes all positions and disables new entries for the day when the threshold is approached. |
| Max drawdown (e.g. 8–10%) | Trailing or static cap from peak/initial balance | Tracks running equity peak (or fixed initial, per firm model); hard kill-switch closes trades and halts the EA before the cap is hit. |
| Minimum trading days (e.g. 3–5) | Must trade on N distinct days | Counts distinct days with at least one valid trade; prevents the EA from "completing" the target too fast and skipping the requirement. |
| News restriction | No trading around high-impact events | Economic-calendar / time-window filter blocks entries (and optionally flattens) inside a configurable buffer before/after flagged events. |
| Weekend / overnight holding | Restricted or banned at some firms | Time-based flatten before Friday close and/or block of new positions inside a no-hold window. |
| Per-trade risk & stop-loss | Every position must have defined risk | Hard stop-loss attached on every order; lot size derived from a risk-% input so no trade exceeds your configured exposure. |
| Consistency / lot rules | No single oversized day or trade dominating profit | Caps max lots per trade and (where required) per-day exposure to keep profit distribution within firm tolerances. |
The point of building these as code-level constraints is that they fire whether or not the strategy is having a good day, whether or not you are at your desk. A risk rule you have to remember to follow is a risk rule you will eventually forget. Here is the shape of the equity guard in plain MQL5 — the single most important piece of a prop EA:
// On every tick, enforce equity-based daily loss + max drawdown
double equity = AccountInfoDouble(ACCOUNT_EQUITY);
double dayStart = DailyAnchorEquity(); // captured at server-day rollover
double peak = RunningEquityPeak(); // for trailing-DD firms
double dailyLossPct = (dayStart - equity) / dayStart * 100.0;
double ddFromPeak = (peak - equity) / peak * 100.0;
if(dailyLossPct >= InpDailyLossBuffer || ddFromPeak >= InpMaxDDBuffer)
{
CloseAllPositions(); // flatten floating risk immediately
BlockNewEntriesToday(); // no re-entry until next session
Print("Risk guard tripped — trading halted for compliance.");
}
Notice it reads ACCOUNT_EQUITY, not balance — that single choice is what separates an EA that survives a bad open trade from one that breaches while you watch. Inputs like InpDailyLossBuffer are deliberately set inside the firm's hard line so the EA shuts down with margin to spare, not at the exact threshold where slippage can still push you over.
What I refuse to build
This is where being a solo developer who actually trades matters — I will tell you no when an agency would take the money. I do not ship:
- Aggressive martingale or recovery-grid blow-up systems. They violate several firms' terms and, more importantly, they are designed to eventually lose everything.
- Latency / HFT arbitrage and tick-scalping designed to exploit feed delays. These are explicitly banned and will get a funded account terminated and profits voided.
- Multi-account coordination or copy schemes meant to game evaluation rules.
If your goal requires one of these, I am the wrong developer — and any developer who agrees to build it is selling you a banned account, not an edge. I would rather lose the job than hand you a system that gets your funded capital clawed back.
What you actually get — deliverables
This is BOFU work, so the deliverable list is concrete. Every engagement ships:
- Commented
.mq5(or.mq4) source code you own — not a locked, ex5-only black box. You can read it, audit it, and change inputs. - A full risk-input panel — daily-loss %, max-DD %, risk-per-trade %, max lots, news buffer, session/weekend windows, min-trading-day logic — so the same EA adapts to FTMO, FundedNext, The5ers, Topstep or MyFundedFX by changing numbers, not rewriting code.
- Verifiable backtests on real tick/spread settings, with the assumptions stated openly — a backtest is a hypothesis, not a promise.
- Plain-English documentation mapping each input to the firm rule it protects.
- Milestone-based payment, NDA on request, and you work directly with me — no agency layer between you and the person writing the code.
If you already have a TradingView strategy you trust, I can convert that Pine Script strategy into a compliant MT5 EA rather than starting from scratch — the entry/exit logic carries over and I wrap it in the prop-firm risk layer. If you have an existing EA that keeps breaching, that's an audit-and-retrofit job: I add the missing equity guard, news filter and stop-loss enforcement to your current code. See the broader MT5 EA development service for scope and the deeper breakdown of why EAs fail prop-firm challenges for the engineering detail behind this page.
How an engagement runs
I keep it boring on purpose, because predictability is the whole value proposition with funded capital on the line:
- Scope call. Which firm(s), which account model (one-step, two-step, static vs. trailing DD), your strategy logic or your existing EA, and the exact rulebook you're targeting.
- Risk architecture first. I build and demonstrate the compliance layer — the guards, filters, and sizing — before tuning entries, so you can see the safety net working in the Strategy Tester.
- Backtest & review. You get results on realistic spread/slippage settings with the assumptions written down, plus a forward/demo check.
- Delivery. Commented source, input docs, and a walkthrough so you can run it yourself.
You can see real broker-automation systems running in my video proof gallery and browse delivered projects on the work page. If you want a developer rather than a service to assess a specific idea, the hire-a-developer page covers how direct engagements work.
One last honest framing, because it's the most important thing on this page. A prop-firm evaluation has two halves: a strategy with genuine edge, and risk discipline that never blinks. I can build you the second half to a standard you cannot reliably hit by hand — code that flattens at the daily line every single time, attaches a stop to every order, and refuses to trade into news or the weekend. What I cannot do — what nobody can — is guarantee the first half produces a winning month inside your evaluation window. Markets don't sign contracts. If a developer promises you a pass, walk away. What I'll commit to in writing is a compliant, auditable risk architecture and verifiable backtests — the part of the equation that is actually engineering, done properly, in source code you own. Send me your firm and your strategy and I'll tell you honestly whether it's a fit before you spend a rupee, dollar, or euro on the build.
Tell me which prop firm and account model you're targeting and what your strategy or existing EA does. I'll tell you honestly whether it's a fit, then build a compliant risk architecture with verifiable backtests and editable .mq5 source you own.
FAQ
Can you build an EA to pass my FTMO challenge?
I can build an EA engineered to respect FTMO's rules — hard stop-loss on every trade, equity-based daily-loss and max-drawdown shutdown, news filter, and minimum-trading-day logic. What I cannot do is guarantee it passes, because passing also depends on whether the strategy has real edge in your evaluation window. I deliver a compliant risk architecture and verifiable backtests, not a promised result.
Will an automated EA get my funded account banned?
Not if it's built to the firm's terms. Accounts get banned for banned behaviors — latency/HFT arbitrage, prohibited martingale, multi-account coordination, tick-scalping that violates terms — which I refuse to build. A properly compliant EA that uses normal entries, hard stops, and rule-based risk control is exactly what the firms expect automated traders to run.
Do you guarantee I pass the challenge?
No, and you should distrust anyone who does. Passing depends on live market behavior during your evaluation, which no developer controls. What I guarantee is a compliant, auditable risk architecture — code that enforces your firm's daily-loss, drawdown, news, and stop-loss rules as hard constraints — plus backtests with the assumptions stated openly.
Is martingale allowed on prop firms?
Several firms restrict or prohibit aggressive martingale and recovery-grid systems, and even where it's tolerated it's a blow-up risk by design. I don't build it. If you want a system whose survival depends on never hitting a long losing streak, I'm the wrong developer — I build EAs with fixed, defined risk per trade instead.
Which prop firms do you support (FTMO, FundedNext, The5ers, Topstep, MyFundedFX)?
All of them and others — the EA's compliance layer is driven by inputs (daily-loss %, max-DD %, news buffer, session windows, min-days), so the same code adapts to FTMO, FundedNext, The5ers, Topstep, or MyFundedFX by changing numbers. You give me the current rulebook for your firm and account model, and I configure and test against it. Always verify the rules with your firm yourself, since they change.
Can you make my existing EA prop-firm compliant?
Yes — that's an audit-and-retrofit job. I add the missing pieces most EAs lack: an equity-based drawdown guard, hard stop-loss enforcement on every order, a news/weekend filter, and risk-percent lot sizing. You get the upgraded, commented source back so you own and can audit the changes.
Do you deliver editable source code?
Yes. You get commented .mq5 (or .mq4) source you own outright — not a locked ex5 black box — plus a documented risk-input panel and plain-English docs mapping each input to the firm rule it protects. NDA on request, milestone-based payment, and you work directly with me.