Why Does My EA Keep Losing Money? 10 Fixes
Why Does My EA Keep Losing Money?
10 real reasons your Expert Advisor is bleeding your account — and exactly how to fix each one. No fluff, no upsells, just 18 years of experience.
⚡ Quick Answer
Most EAs lose money for fixable reasons: wrong broker, lot size too big, trading through news events, or using a scam EA with fake backtest results. The EA itself is rarely the only problem — it's the setup around it. Go through the 10 reasons below and you'll almost certainly find yours.
In This Article
- Your EA is a scam (curve-fitted or fake results)
- You're trading through high-impact news
- Your lot size is too big
- You picked the wrong broker
- Your spreads are too wide for the strategy
- You keep interfering with the EA's trades
- Your VPS or internet keeps disconnecting
- You're using default settings on the wrong pair or timeframe
- Market conditions changed
- You didn't backtest properly (or at all)
- Quick-fix checklist
- FAQ
If you're reading this, your EA is probably losing money and you're trying to figure out why. Good — because most traders just blame the EA and buy another one. The truth is, the EA is usually only half the problem. The other half is your setup, your broker, your risk settings, and your expectations.
I've been in the forex industry for over 18 years. I've tested hundreds of EAs, sold thousands through CheaperForex, and talked to tens of thousands of traders about what went wrong. These are the 10 most common reasons — ranked by how often I see them.
Your EA Is a Scam (Curve-Fitted or Fake Results)
This is the number one reason. Not because most traders are stupid — because most EA sellers are dishonest. They optimise their EA to look perfect on historical data (curve-fitting), then show you a backtest that looks like a straight line going up. In live trading, it falls apart within weeks.
Red flags: Backtest shows 99%+ win rate. No verified live results on MyFXBook. Developer won't show a live account. The EA was released last month but shows "3 years" of results. Price seems too cheap for a "money printer."
Even MyFXBook results can be faked — some brokers like Weltrade and OnFin allow traders to edit their trade history, making verification worthless.
Only buy EAs with verified live trading results from a reputable broker. Read our complete guide to spotting forex scams — it covers 8 types of scam including fake EAs and manipulated MyFXBook results.
You're Trading Through High-Impact News Events
NFP, FOMC, CPI — these events cause spreads to explode from 0.1 pips to 20+ pips on EUR/USD and up to 5,000+ pips on gold. Your EA doesn't know this is coming. It opens a trade at what it thinks is a normal spread, and instantly you're 50 pips in the red before anything even moves.
Grid EAs are especially vulnerable. They place trades at fixed intervals — but during news, price blows through every grid level in one direction. Martingale EAs double down on each losing trade, turning a bad entry into an account-ending position.
Use our free economic calendar to know when high-impact events are coming. Pause grid, martingale, and scalping EAs 15–30 minutes before and after. Some EAs have built-in news filters — check your settings.
Your Lot Size Is Too Big
This kills more accounts than bad EAs. A trader with a $1,000 account running 0.1 lots on EUR/USD has $1 per pip of risk. That sounds small — but a grid EA with 10 open positions is risking $10 per pip. A 100-pip move against you is your entire account.
Most EA developers provide recommended lot sizes in their documentation — but many traders ignore them or don't understand the maths. "0.01 lots" feels too small. So they bump it up. Then they wonder why a normal pullback wiped them out.
Use our free lot size and risk calculators to work out the correct position size for your account balance. As a rule of thumb: risk no more than 1–2% of your account per trade. If your EA opens multiple positions, divide accordingly.
You Picked the Wrong Broker
Your EA might be excellent — but if your broker has wide spreads, slow execution, or frequent requotes, the EA can't do its job. A scalping EA that targets 5 pips of profit needs a broker with spreads under 1 pip. If your broker is charging 2–3 pips, you've already lost before the trade even starts.
Other broker problems: slippage on stop losses, different price feeds that cause the EA to enter trades at different levels than backtested, and swap rates that eat into profits on overnight positions.
Check your EA developer's recommended brokers. Then read our honest broker reviews — 9 brokers reviewed with zero affiliate links. For most EAs, raw spread / ECN accounts with low commissions perform best.
Your Spreads Are Too Wide for the Strategy
Related to broker choice, but worth its own section. Every EA has a break-even spread — the maximum spread at which the strategy is still profitable. A scalper targeting 8 pips with a 2:1 risk-reward needs spreads under about 1.5 pips to be viable. Go above that and the maths doesn't work.
This is especially brutal during the Asian session when liquidity drops and spreads naturally widen. Your EA keeps trading, but every trade starts further in the red.
Check your average spread on the pairs your EA trades. Most brokers show this in MT5 under View → Market Watch → right-click → Show Columns → Spread. If your spreads are consistently wider than 1–2 pips on major pairs, switch to a raw/ECN account or a better broker.
You Keep Interfering With the EA's Trades
This is more common than you'd think. The EA opens a trade. It goes 10 pips into the red. The trader panics and closes it manually. Then the trade would have hit take profit 20 minutes later. The trader just turned a winning trade into a losing one.
Or worse: the trader moves the stop loss, changes the lot size mid-trade, or manually opens extra positions "to help the EA." Every intervention breaks the strategy's logic.
If you can't watch the EA trade without interfering, put it on a VPS and don't log in. Seriously. The whole point of an EA is that it removes human emotion from trading. If you override it, you're not EA trading — you're emotional trading with extra steps.
Your VPS or Internet Keeps Disconnecting
If your EA runs on your home computer and your WiFi drops for 30 seconds, any open trade without a stop loss is now unmanaged. If the EA uses a trailing stop, that stop only works while MT5 is connected. Disconnect = no trailing stop = potential disaster.
Even with a VPS, cheap providers with high latency or frequent restarts can cause missed entries, delayed exits, and orders that get rejected.
Use a forex-optimised VPS located near your broker's server (usually London or New York). Expect to pay $15–30/month for a reliable one. ForexVPS, AccuWeb, and Contabo are popular choices. Make sure latency to your broker is under 5ms if you're running a scalper.
You're Using Default Settings on the Wrong Pair or Timeframe
Most EAs are optimised for specific currency pairs and timeframes. A gold scalper won't work on EUR/USD. A trend-following EA designed for H4 won't work on M1. Using default settings on a pair the developer never tested is like putting diesel in a petrol car.
Even the right pair on the wrong timeframe can produce completely different results. The EA might look for patterns that only exist on M15 — run it on M5 and it's trading noise.
Read the EA's documentation. Every good EA specifies which pairs and timeframes it's designed for. If the EA came with preset files (.set files), use them — they contain the developer's optimised parameters. Don't change settings unless you understand exactly what each parameter does.
Market Conditions Changed
An EA that crushed it in a trending market might bleed in a ranging market. A range-bound EA that averaged 5% per month might get destroyed when a breakout happens. No EA works in every market condition forever.
This is actually normal. The question isn't whether your EA will have losing periods — it will. The question is whether those losing periods are within the expected drawdown range from backtesting, or whether something fundamental has broken.
Check the EA's historical drawdown in backtests. If your live drawdown is within that range, the EA is performing as expected — be patient. If it's significantly worse than historical worst-case, the market may have shifted enough that the strategy is no longer viable. Consider pausing it and reviewing.
You Didn't Backtest Properly (or at All)
Some traders buy an EA, slap it on a live $5,000 account, and pray. No backtest. No demo. No idea what the expected drawdown is. When it drops 15% in the first week, they panic — even though 15% drawdown might be completely normal for that strategy.
Others do backtest, but with the default "Open Prices Only" modelling in MT5, which is useless for scalpers. Or they use 1-year of data and assume it'll work for the next 10 years.
Always backtest with "Every Tick Based on Real Ticks" in MT5 Strategy Tester. Use at least 2–3 years of data. Then run the EA on a demo account for 2–4 weeks before going live. This won't guarantee profits, but it'll set realistic expectations for drawdown and win rate.
Quick-Fix Checklist: Run Through This Before Giving Up on Your EA
- Is the EA from a legitimate developer with verified live results?
- Are you trading through high-impact news events? Check the economic calendar
- Is your lot size appropriate for your account balance? Use the lot size calculator
- Is your broker suitable for EA trading? Check the broker reviews
- Are your spreads low enough for the EA's strategy?
- Have you stopped manually interfering with the EA's trades?
- Is your MT5 platform connected 24/5 (VPS or reliable internet)?
- Are you using the correct currency pair and timeframe?
- Are you using the developer's recommended settings or .set files?
- Did you backtest with real tick data before going live?
- Did you run on demo for at least 2 weeks first?
- Is your current drawdown within the expected range from backtesting?
Related Free Resources
Frequently Asked Questions
How long should I wait before deciding my EA is losing money?
At least 4–8 weeks on a live or demo account. Every EA has losing streaks — what matters is whether the drawdown is within the range shown in backtesting. A 10% drawdown might feel terrible, but if the backtest showed 15% max drawdown, you're actually performing better than expected. Don't judge an EA on a single week.
Should I run my EA on a demo account first?
Absolutely. Run it on demo for at least 2–4 weeks before going live. This lets you verify it's trading correctly, check that your broker's spreads are compatible, and understand the EA's behaviour (how often it trades, typical drawdown, how it handles news events). Demo results won't be identical to live — slippage and spread widening differ — but it's far better than going in blind.
Can a bad broker really cause an EA to lose money?
Yes, significantly. A broker with 2-pip spreads on EUR/USD turns a profitable scalper into a losing one. Slow execution causes slippage on entries and exits. Different price feeds mean your EA sees different patterns than the developer's. This is why most EA developers recommend specific brokers — it's not an affiliate play, it's because the EA was tested and optimised on those price feeds.
How do I know if my EA is a scam?
Check for: verified live results on MyFXBook from a reputable broker (not Weltrade or OnFin), at least 6 months of live trading history, realistic returns (20–50% per year is excellent — anything claiming 100%+ monthly is fake), and a developer who openly shows drawdown. Read our full scam guide for 8 specific scam types and how to verify any EA before buying.
My EA was profitable for months but suddenly started losing. What happened?
Most likely: market conditions changed. A strategy that worked in low-volatility ranging markets may fail when volatility spikes (geopolitical events, central bank surprises). Or your broker changed their liquidity provider, widening typical spreads. Check your spread logs, verify the EA settings haven't been altered, and compare your results with other users of the same EA (developer's community or forums).
Is it normal for an EA to have losing trades?
Completely. Even the best EAs in the world have win rates between 55–75%. A 60% win rate means 4 out of every 10 trades lose. What matters is the risk-reward ratio — if winners average 15 pips and losers average 10 pips with a 60% win rate, you're profitable. Losing trades are not a sign of a broken EA. Losing months, consistently exceeding historical drawdown — those are signs to investigate.
What's the single most important thing I can do to stop my EA losing money?
Use the correct lot size. It's not exciting, but it's the answer. Most blown accounts aren't caused by bad EAs — they're caused by traders using 0.1 lots on a $500 account with a grid EA that opens 15 positions. The maths is simple: use the calculator, keep risk per trade under 1–2% of your balance, and your EA suddenly has room to breathe through its normal drawdown periods.