Set per-trade risk at 0.5% to 1.0% of account balance, with the lower end reserved for accounts under a strict daily loss limit or a volatile symbol. Build your maximum drawdown buffer 1.5 to 2 percentage points inside whatever the firm publishes. None of that matters, though, without one non-negotiable piece: an equity-based kill switch that flattens every open position and blocks new trades the instant your internal thresholds are hit.
TL;DR:
- Setting per-trade risk at 0.5% to 1.0% is advisable, especially for accounts with lower daily loss limits or volatile symbols like gold and exotic pairs.
- Internal daily stops and drawdown buffers should be set 1.5 to 2 percentage points inside the firm's limits to account for slippage and floating losses.
- Use a strict cap of three to five open positions, with automatic reduction if floating losses reach half of daily risk, to control overall exposure.
- An equity-based kill switch is essential for real-time risk management, combined with disconnect protection and detailed logging for transparency.
- Test EAs over at least a month in a setup that mirrors the firm's conditions before risking live funds, focusing on robustness during spreads widening and fast market moves.
Table of Contents
- What EA Risk Settings Do Prop Firms Actually Require?
- How Do You Calculate Lot Size From Percent Risk?
- How Do You Build Account-Level Safety Into an EA?
- How Do You Test an EA Before Risking a Challenge Fee?
- What Broker and Platform Details Break EA Assumptions?
- Setup Checklist: Test One Month, Then Pay for the Challenge
- Why Conservative Settings Beat Aggressive Ones on a Challenge
- How Fundedaxe Fits an EA-First Trading Plan
- Sources
- FAQ
What EA Risk Settings Do Prop Firms Actually Require?
Prop firms rarely dictate exact EA parameters. They publish outer boundaries, usually a daily loss limit and a maximum drawdown limit, and leave the internal math to you. That gap between the firm's rule and your EA's actual behavior is where most challenge attempts fail, not because the strategy was bad, but because the risk settings never accounted for slippage, correlated trades, or a losing streak hitting on the same day news drops.
The settings below aren't arbitrary. They're built around one idea: your EA should breach its own internal limits well before it ever gets close to the firm's limits.
Per-trade risk: 0.5% or 1.0%?
Most practitioners cap per-trade risk between 0.5% and 1.0% of account balance, precisely so one losing streak doesn't wipe out an entire evaluation. Use 0.5% per trade when:
- The firm's daily loss limit sits at 4% or below.
- Your EA trades more than three positions concurrently.
- You're running the strategy on a symbol with historically wide spread swings (gold, exotic pairs).
Anything riskier than 1.0% per trade on a prop challenge is asking for a single bad afternoon to end the attempt.
Daily stop and drawdown buffer logic
Your EA's internal daily stop should trigger 1.5 to 2 percentage points before that, giving you a cushion for slippage on the closing trade and any floating loss that hasn't settled yet.
Total drawdown works the same way. If the firm's max drawdown limit is at a common industry level, set your EA's hard stop to be a couple percentage points inside it to maintain a safety buffer. That 2-point gap has saved more challenge attempts than any entry signal ever will.
Position limits and exposure caps
Concurrent open positions need a hard ceiling, not a soft guideline. A reasonable starting point:
- Cap total open positions at 3 to 5, depending on correlation between symbols traded.
- Set an aggregated exposure cap so that if all open trades hit stop-loss simultaneously, total loss never exceeds your daily stop.
- Reduce position count automatically once floating loss crosses 50% of the daily stop threshold.
Stop-loss distance, news filters, and weekend rules
Stop distance should be volatility-aware, not fixed in pips.
Speaking of news: block new entries 15 to 30 minutes before and after high-impact releases, and flatten any position still open 10 minutes before a red-folder event if your strategy isn't specifically built to trade the volatility. Holding positions through weekends or major news introduces gap risk that can breach a daily or total drawdown limit before the market even opens, regardless of how conservative your other settings are. Close everything by Friday 3:00 PM server time unless the firm explicitly permits weekend holding and your strategy has been stress-tested for gap scenarios.
A copyable starting setfile
For a $100,000 account with a 5% daily limit and 10% total drawdown:
- Risk per trade: 0.5%
- Daily stop: set to a conservative percentage that is safely below typical firm daily loss limits
- Max drawdown stop: set below your firm's published maximum drawdown limit by a couple percentage points
- Max concurrent positions: a small number such as three to five to limit exposure
- Stop distance: 1.5x ATR(14) on H1
- News filter: 20 minutes before/after high-impact events
- Friday close: close positions by Friday afternoon in server time unless weekend holding is explicitly allowed and stress-tested
Adjust the percentages down, never up, as you add position count or trade correlated pairs.
How Do You Calculate Lot Size From Percent Risk?
The formula stays the same no matter the account size: account balance times risk percent equals your dollar risk, and dollar risk divided by stop distance (converted to account currency) gives you lot size.
Here's the exact sequence:
- Find dollar risk. Account balance × risk % = dollar risk. A $50,000 account risking 0.5% puts $250 on the line per trade.
- Convert stop distance to dollars. If your stop is 30 pips on EUR/USD with a standard lot worth $10 per pip, that's $300 per lot for a 30-pip stop.
- Divide dollar risk by dollar stop value per lot. $250 ÷ $300 = 0.83 lots. Round down to 0.80 lots to stay inside your risk budget.
Worked examples across account sizes
| Account Size | Risk % | Dollar Risk | Stop Distance | Lot Size (approx.) |
|---|---|---|---|---|
| $100,000 | 0.5% | $50 | 25 pips | 0.20 lots |
| $50,000 | 0.5% | $250 | 30 pips ($10/pip on 1 lot) | 0.83 lots |
| $100,000 | 1.0% | $1,000 | 40 pips ($10/pip on 1 lot) | 2.5 lots |
Treat the firm's daily drawdown limit, not the advertised balance, as your real trading capital. On a $100,000 account with a 5% daily limit, you're effectively working with $5,000 of usable risk per day, and every open position draws from that pool.
Aggregating risk across multiple EAs
If you run more than one EA on the same account, or mix manual trades with automated ones, dollar risk has to be tracked centrally, not per strategy. Build a shared risk ledger, even a simple spreadsheet that logs open positions and their worst-case loss, so you always know the combined number before the market does.
A basic spreadsheet with columns for symbol, lot size, stop distance, and dollar risk covers most single-account setups. For anything more complex, code the aggregation directly into the EA using a global variable or a shared file that every running instance checks before opening a new trade.
How Do You Build Account-Level Safety Into an EA?
An equity-based kill switch that immediately stops trading and closes positions when risk limits are reached is essential for managing prop firm challenges. Regulatory guidance on algorithmic trading treats an immediate disable mechanism, one that can cancel working orders and halt new entries, as a baseline requirement for automated systems, not an optional extra.
The kill switch has to check equity, not balance. Balance only updates when a trade closes; equity reflects floating losses tick by tick. An EA that monitors balance alone can sail past its daily stop while three open positions are underwater, because the account "looks fine" until those trades close. Equity-based monitoring catches the breach in real time, before it becomes irreversible.
Structure the safety layer in tiers:
- Warning threshold: Trigger an alert (push notification, email, or terminal popup) at 80% of your internal daily stop.
- Soft stop: Block new trade entries once the daily stop is hit, but let existing positions run their own stop-loss.
- Hard stop: Close all open positions immediately and disable the EA for the remainder of the trading day once total drawdown buffer is breached.
- Per-order caps: Reject any single order whose risk exceeds your per-trade percentage, even if a bug or bad input tries to push it through.
- Per-symbol exposure limits: Cap total lots on any one symbol so a correlated move can't multiply your loss beyond the plan.
Cancel-On-Disconnect functionality, which cancels working orders automatically on disconnection, is an important safeguard to prevent unintended exposure rather than sit exposed with no one watching them. Industry best practices for automated systems specifically call out cancel-on-disconnect alongside pre-trade limits and intraday position caps as core safeguards, and a prop challenge is exactly the environment where a silent disconnect can turn a passing account into a failed one overnight.
Pro Tip: Log every kill-switch trigger with a timestamp, the equity value at trigger, and which threshold fired. When you review a failed evaluation later, that log tells you whether the strategy failed or the safety system did its job.
Persistence matters too. If your terminal restarts or the VPS reboots, the EA needs to reload its daily loss counter and open-position state from a saved file, not start fresh and assume zero risk has been taken. A kill switch that resets on restart isn't a kill switch, it's a loophole waiting to be found the hard way.
How Do You Test an EA Before Risking a Challenge Fee?
Test EAs extensively in a demo environment that replicates the firm's conditions before entering paid evaluations to catch errors early, or you're paying to find bugs that a demo account would have caught for free.
Start with a backtest sample large enough to mean something. Two hundred to five hundred trades is the range practitioners recommend to avoid drawing conclusions from a lucky (or unlucky) short stretch. Fewer than 100 trades tells you almost nothing about how the strategy behaves across different market regimes.
Once the backtest clears that bar, move to a forward or demo test that copies the firm's setup exactly:
- Same broker or a broker with matching spread and execution characteristics.
- Same symbols, including any suffix or naming convention the firm's platform uses.
- Same leverage cap (many firms limit to 1:100).
- Same server time zone, so news filters and Friday-close logic trigger at the correct real-world time.
- Same account size and drawdown limits programmed into the EA's internal stops.
Run this for at least a month before committing challenge fees. Markets are tightly interconnected systems, and treating an EA as a set-and-forget tool rather than something requiring periodic stress testing is one of the more common ways traders get blindsided by a scenario their backtest never modeled.
Stress test execution specifically, not just strategy logic:
- Simulate spread widening during a news window and confirm the EA doesn't open a trade it wouldn't take at normal spread.
- Check slippage tolerance on stop-loss orders during fast market moves.
- Force two positions to close simultaneously and verify the kill switch calculates equity correctly mid-execution.
Every time you update the EA, even a minor parameter tweak, rerun a shortened regression test before touching a live or evaluation account. Build a release checklist: backtest results, forward-test log, kill-switch trigger test, and a rollback version saved separately. Skipping that step because "it's a small change" is how a working setup turns into a failed challenge.
What Broker and Platform Details Break EA Assumptions?
Symbol naming is the most common silent killer. A broker labeling gold as "XAUUSD" while your EA was coded for "GOLD" or a firm platform adding a suffix like ".raw" will cause the EA to simply not trade, or worse, miscalculate lot size against the wrong contract specification. Check contract size, tick value, and symbol name against the firm's exact server before running anything live.
Spread and swap behavior shift during low liquidity, particularly in the hour after rollover and during major news. An EA with a fixed stop distance calculated for average spread can get stopped out immediately when spread triples during a data release. Build in a spread filter that pauses entries when current spread exceeds a multiple of the recent average.
Weekend and Monday gap risk deserves its own line item. A position held into Friday close can open Monday 50 to 100 pips away from where it closed, instantly consuming a daily or total drawdown buffer that took weeks to build. Closing by Friday afternoon, server time, removes that risk entirely for strategies not specifically designed to hold through the gap.
Forward testing on the firm's actual evaluation platform is important because demo and live execution conditions can differ significantly. Some prop platforms report execution slippage on evaluation accounts running 20% to 40% higher during high-volatility windows than what a standard demo shows, which is exactly why forward-testing on the firm's actual platform matters more than testing on a generic demo elsewhere.
Operational hygiene rounds this out:
- Keep a running log of every trade, kill-switch event, and connectivity drop.
- Set alerts for disconnects, not just for trades.
- Save a rollback version of every EA setfile before deploying a change.
Setup Checklist: Test One Month, Then Pay for the Challenge
- Read the firm's rules document line by line. Note the exact daily loss limit, total drawdown limit, and whether it's calculated on balance or equity.
- Configure the EA's internal stops to sit inside those limits. Use the 1.5 to 2 point buffer discussed earlier for both daily and total drawdown.
- Load the firm-specific setfile, matching symbol names, leverage, and account currency to the exact evaluation environment.
- Run a full month of demo trading under those exact conditions, logging every trade, every kill-switch trigger, and every day the internal stop activated.
- Check pass criteria before risking money: no kill-switch triggers in the final two weeks, drawdown staying comfortably inside the internal buffer, and a trade count consistent with your backtest's sample size.
- Save the setfile, the trade log, and a rollback copy somewhere you can reference if the live challenge behaves differently than the demo.
Pro Tip: If your demo month includes even one week where the internal daily stop triggered twice, don't take the paid challenge yet. That's the strategy telling you the risk settings are still too loose for that account size.
Why Conservative Settings Beat Aggressive Ones on a Challenge
Traders consistently underprice one thing: the value of not failing.
The discipline that matters most isn't the initial setfile, it's what happens after. Watch equity daily, not weekly. Pause the EA the moment market conditions shift outside what your backtest covered, a new central bank policy stance, a change in typical volatility, a broker changing its spread model. Some prop firms offer account structures with no time limit on any evaluation phase, static drawdown rules instead of a trailing calculation that punishes early gains, and support for algorithmic trading rather than restricting EAs to a workaround. Those conditions reward exactly the conservative, well-tested approach outlined here rather than punishing patience with an artificial clock.
— Jean
How Fundedaxe Fits an EA-First Trading Plan
Some prop firms allow algorithmic trading on evaluations, sometimes with no time limit on any phase, so an EA that needs a month of live-conditions data to prove itself doesn't get rushed into a rule violation. Leverage options and swap conditions vary by firm; some offer leverage up to 100:1 and swap-free options for overnight holding strategies, and static drawdown rules mean an early winning streak doesn't shrink your buffer the way a trailing drawdown calculation would elsewhere.
If you'd rather not risk a full challenge fee finding out whether your setfile behaves the same on a live evaluation server as it did in backtesting, start with the Free1K simulated trial account, no card, no deposit, just a real environment to confirm your kill switch and lot-sizing math work as designed. When you're ready to commit, Pay After Pass lets you start an evaluation for $9.99 and only pay the remaining $515.01 base fee once your EA actually passes, instead of putting the full amount up front on a strategy you haven't stress-tested under real firm rules. Compare account sizes and step structures on the package comparison page before you pick your evaluation.
Sources
A handful of sources go deeper than any single article can on the regulatory and technical side of automated trading risk:
- How to Size Your Prop Trades
- FINRA Notice 15-09 (algorithmic trading supervisory guidance)
- FIA best practices for automated trading risk controls (2024)
- HRO and normal accident theory applied to automated trading systems
- Weekend gap problem: why holding positions over the weekend increases risk
FAQ
What Is the Best EA for Prop Firms?
There's no single "best" EA. What matters is whether the EA's risk settings, per-trade percentage, daily stop, and kill switch, are configured conservatively enough to survive a full evaluation month without breaching firm limits. A mediocre strategy with disciplined risk settings outperforms an aggressive one that risks disqualification on its first bad week.
Is 3% Risk Per Trade Good?
Practitioners generally recommend keeping per-trade risk between 0.5% and 1.0% specifically so a short losing streak doesn't consume the firm's daily or total drawdown limit.
What Trading Strategy Has a 90% Win Rate?
Win rate alone tells you little without knowing the risk-to-reward ratio and how the strategy performs across at least 200 to 500 trades.
What Is the 3-5-7 Rule in Trading?
Definitions of the 3-5 rule vary across trading communities, and it isn't a formally standardized risk framework.
Does Fundedaxe Allow EAs on Its Evaluations?
Yes, Fundedaxe permits algorithmic trading and EAs on every account size and evaluation type, with no time limit on any phase to complete testing. Static drawdown rules and leverage up to 1:100 make it straightforward to configure the internal risk settings covered throughout this guide.