EA Losing Money? What Actually Happened This Week (And What to Do)
My Trading

EA Losing Money? What Actually Happened This Week (And What to Do)

29 March 2026, 15:30
Diego Arribas Lopez
Diego Arribas Lopez
0
20

Your EA is red. Your account is down. And right now, every instinct you have is screaming at you to do something — disable it, close the trades, pull the money, find a different EA, start over.

Stop.

Because the biggest losses in EA trading almost never come from the market. They come from what you do in the 48 hours after the market moves against you. The panicked decisions. The revenge trades. The "let me just switch to this other EA I saw on MQL5 with a perfect backtest." That is where accounts actually die.

And here is the part nobody in this industry will tell you right now: every single EA vendor who sold you a dream of consistent returns knew this week was possible. The ones who showed you a pretty equity curve and said "look at these results" — they knew that curve included weeks exactly like this one. They just did not show you what it felt like to live through them with real money.

This is not a sales pitch. There is no product link coming. This is a framework for the decision you are about to make — because that decision matters more than the drawdown itself.

What Actually Happened This Week (No Jargon, No Sugarcoating)

Geopolitical conflict escalated faster than anyone expected. When that happens, institutional money moves first and moves hard. Billions get repositioned in hours based on risk models that have absolutely nothing to do with your support levels, your fibonacci retracements, or your EA's carefully optimized parameters.

Your EA did not see this coming. Neither did mine. Neither did anyone's.

The result:

  • Gold went insane. Not a clean rally — a spike, a pullback, another spike. Stop losses on both sides got liquidated like they were not even there.
  • Major pairs whipsawed. Moves that normally take days happened in hours. Spreads widened to levels that turned profitable strategies into losers on every single entry.
  • Correlations broke. Pairs that normally move together diverged. Portfolio diversification — the one thing supposed to protect you — temporarily stopped working.
  • Volatility multiplied 2x to 4x. Your EA was calibrated for a 300-pip daily range. Gold moved 1,000. The math stopped making sense.

This is called a regime change. The market's underlying behavior shifted so dramatically that the rules your EA was built on temporarily do not apply. Not broken. Not invalid. Temporarily suspended by forces bigger than any retail trading strategy can account for.

Your EA Did Not "Break." You Were Sold an Incomplete Picture.

Here is where it gets uncomfortable.

If you are feeling betrayed by your EA right now — that is valid. But the EA is probably not the problem. The problem is how most vendors sell EAs: with a backtest curve that only tells the story they want you to see.

That 45-degree equity curve they showed you? It included drawdowns like this one. But on a historical chart, a 3-week drawdown looks like a tiny dip before the line continues up and to the right. It looks manageable. It looks like nothing.

Living through it with real money is a completely different experience. And nobody prepared you for that because preparing you for it does not sell products.

This is the uncomfortable truth about EA trading that the industry hates: the backtest always looks easy. The live experience never is. Not because the strategy is bad — but because a chart cannot simulate the feeling of watching your money disappear while the news tells you the world might be falling apart.

If you have been through this before — bought an EA, hit a drawdown, panicked, turned it off, watched it recover without you, bought another EA, repeated the cycle — you are not stupid. You were given tools without context. Results without preparation. A product without the ugly part of the manual. We explained why this happens systematically in Why Perfect EA Backtests Lose Money Live.

The 48-Hour Rule (The Only Rule That Matters Right Now)

Do not make any permanent changes to your EA setup within 48 hours of a market shock.

This is not patience-as-virtue motivational nonsense. It is neuroscience.

Watching real money disappear triggers the same neurological response as physical threat. Your prefrontal cortex — the part that weighs evidence and makes rational decisions — is being overridden by your amygdala. You are literally operating from the fight-or-flight part of your brain. Every decision you make right now is compromised by biology.

The initial shock is also almost always the worst part. Markets typically overshoot on the first reaction and then find some equilibrium within 24-48 hours. Making permanent decisions based on the peak of an overreaction is how a 10% drawdown becomes a 40% catastrophe.

What you can do right now:

  • Reduce position sizes by 50-75% (smaller bets, same strategy)
  • Enable wider spread filters if your EA supports them
  • Monitor without touching anything structural
  • Write down what you observe — this data is gold later

What will absolutely make things worse:

  • Killing the EA entirely and closing all positions at a loss
  • Switching to that "better EA" you saw on MQL5 with the beautiful backtest (it is probably another trap)
  • Doubling position sizes to "recover faster" — this is how accounts blow up, every single time
  • Changing fundamental settings like timeframes or pairs based on three bad days

3 Questions Before You Touch Anything

After 48 hours, when your brain is working again, answer these. Not before.

Is this drawdown within the documented range?

Every legitimate EA has a documented maximum drawdown — from backtesting, forward testing, or live history. Check it. Not what the vendor said in the sales pitch. What the verified data shows.

If the vendor claimed 8% max drawdown and you are at 7% — it feels terrible, but it is within range. You accepted this risk when you started. The question is whether you accepted it for real or only in theory.

If you are significantly beyond the documented range — 15% when the max was supposed to be 8% — that is a genuine problem. But even then, check whether the documentation accounted for black swan events or only "normal" conditions.

Is the EA behaving consistently with its own logic?

A losing EA and a malfunctioning EA are two very different animals. Check the MT5 journal:

  • Is it opening and closing trades at expected times?
  • Are position sizes matching your configuration?
  • Are stops and take profits placed where they should be?

If yes to all — the EA is working exactly as designed. It is just losing because the market conditions are hostile to its strategy. That is temporary. A malfunctioning EA placing random trades or ignoring its own rules is a different story entirely.

Has the market changed, or has the EA changed?

If the market changed (geopolitical shock, regime shift) but the EA is behaving consistently → the problem is temporary. Markets return to normal. The question is whether your account survives until they do.

If the EA changed (broken update, API issues, vendor went silent) → the problem may be permanent.

In almost every geopolitical crisis, the answer is: the market changed, the EA did not. That means patience, not panic. AI-integrated EAs like DoIt Alpha Pulse AI have an advantage here — they can read market context and reduce trade frequency during hostile conditions, which is exactly what happened during the last volatility spike. But even traditional EAs that are working correctly deserve the benefit of the doubt during a regime change.

When "Turn It Off" Is the Right Call

Sometimes it is. If the EA is exceeding documented drawdown by a wide margin, if the vendor has disappeared (which tells you everything about who you bought from), if you are on a funded account approaching limits — turning off is not panic. It is arithmetic. If you are running EAs on a funded account, the funded account settings guide covers the specific rules that matter.

But do it correctly:

  1. Disable new trades first — not the EA entirely. Keep it managing existing positions.
  2. Do not close all positions in a panic. Some may be near their stops already. Manually closing them just locks in losses that were about to be contained.
  3. Write down why you turned it off and what would need to happen for you to turn it back on. Because the number one mistake after disabling is never restarting. The crisis passes, the EA would have recovered, but you never turned it back on because "what if it happens again." And then you watch from the sideline as the equity curve climbs without you.
  4. Set a review date. Not "when I feel ready." An actual calendar date. One week. Numbers, not feelings.

The Hardest Truth in EA Trading

The equity curve you admired before buying your EA included drawdowns exactly like this one. You just did not feel them, because they were data points on a chart, not your rent money on the screen.

Every EA that has ever been genuinely profitable over 12 months has had weeks where it lost money. Some have had months. The traders who make money with EAs are not the ones who found a perfect system that never loses. Those systems do not exist, and the vendors who claim otherwise are lying to your face.

The traders who succeed are the ones who can tell the difference between "this is a normal drawdown in a legitimate strategy during abnormal conditions" and "this is genuinely broken." They have a framework. They use data instead of fear. And they reduce risk instead of eliminating exposure.

You now have that framework. The question is whether you will use it — or whether the panic will win again.

One more thing: if this drawdown exposed that you were running a single EA on a single pair with no diversification — that is not just a risk management issue, it is a structural one. A single EA is not a trading plan. A portfolio of uncorrelated strategies handles regime changes better because when one strategy suffers, another may be flat or positive. If you want to start building that diversification without spending anything, the free USDJPY strategy module is a zero-cost starting point.

And if the crisis also showed you that your broker's spreads blew out during peak volatility — that matters more than most traders think. Brokers with deep institutional liquidity like IC Markets or Pepperstone maintain tighter spreads when conditions get hostile. During a crisis is when you discover whether your broker's "raw spreads" claim is real. If you are scaling capital, Axi Select offers a funded trading model without the artificial daily loss limits that make crashes even more dangerous on traditional prop firms.

Frequently Asked Questions

How do I know if my EA is broken vs experiencing normal drawdown?

Check three things: (1) Is the current drawdown within the vendor's documented maximum? (2) Is the EA executing trades consistently with its configured logic — same entries, exits, and position sizes as before? (3) Did the drawdown start during or immediately after an unusual market event? If the drawdown is within range, the logic is consistent, and the timing aligns with a market shock — it is almost certainly normal drawdown, not a broken EA. If any of those checks fail, investigate further before continuing.

Should I close all open trades immediately during a market crash?

Almost never. Closing all positions during a crash crystallizes your losses at the worst possible moment — you absorb 100% of the drawdown and participate in 0% of the recovery. Instead, disable new trades while letting existing positions be managed by the EA's built-in stop losses and take profits. The exception: if you are on a funded account approaching your daily loss limit, closing to preserve the account may be necessary. But that is account preservation, not trading strategy.

How long should I wait before restarting my EA after a crisis?

Wait until volatility returns to within 1.5x of its pre-event average. Compare current daily ranges to the 30-day average before the event. When daily ranges normalize, spreads tighten back to normal, and correlations resume their typical patterns — the regime change is likely over. This usually takes one to three weeks after a major geopolitical event, though severity matters. Do not use "it feels calmer" as your metric. Use actual numbers.

Resources