What is a recovery EA?

A recovery EA is an Expert Advisor that manages positions that already exist instead of looking for new entries. Where a normal strategy EA decides when to buy or sell, a recovery EA decides what to do when a trade goes against you: manual trades you took by hand, positions opened by another EA, or entries fired by a webhook bridge such as a TradingView-to-MT5 alert. It has no opinion on where price will go — it only reacts to your open risk.

The goal is narrow and honest: repair losing positions back to break-even or a small positive exit, instead of letting a single loser sit until it wipes out the week's gains. It does not turn a losing strategy into a winning one. Nothing does. What it can do is stop a handful of open losers from defining your month.

How a recovery cycle works

Implementations differ, but almost every recovery EA on MT5 follows the same five-stage pattern:

  1. Watch and trigger. The EA monitors open positions by symbol, magic number or "manage all". When a position's floating loss crosses a dollar trigger you configure, the cycle starts.
  2. Averaging grid. The EA adds positions around the loser at fixed price or distance intervals, lowering the average entry of the basket. Some tools add in the same direction; others open a mirror grid on the opposite side so the basket profits from the oscillation in either direction.
  3. Basket take-profit. Instead of hoping the original entry comes back, the EA aims the whole basket at a shared exit: break-even plus a small target, measured on combined profit.
  4. Partial closes. When positions overlap, closing slices of the pair banks the bounce and shrinks exposure while the cycle is still working.
  5. Safety layers. A maximum number of grid levels, a hard dollar loss cap that accepts a defined loss instead of holding forever, and often an equity-based stop plus a cooldown before the cycle re-engages.

Advanced tools add modes on top — for example taking over the whole book on a symbol at an equity threshold, or an anti-stall engine that loosens a basket that cannot reach its target. The details vary; the risk shape does not.

The risks, honestly

Recovery EAs get a bad reputation, and most of it is earned by people running them without caps. Before you configure one, understand exactly where the danger lives:

  • Exposure grows while the cycle runs. Averaging means adding to a loser. The basket gets bigger exactly when the trade is going badly, so the floating drawdown accelerates before it shrinks.
  • One-way trends are the killer. Grids are built for ranging markets. A strong trend against the basket walks through every grid level — this is where unrecovered losses come from. If your EA has a trend or impulse filter, leave it on.
  • Swap and commission bleed. Long cycles on high-swap symbols (gold, exotics, indices over the weekend) can eat the small target the basket is aiming at.
  • Margin pressure. Floating loss plus growing lot count pushes your margin level down from both sides. An unrecovered basket can reach stop-out even if the price eventually returns.
  • Prop firm rules measure equity. Daily and overall drawdown are calculated on equity, and equity dips while a cycle is open. A recovery EA on a funded or challenge account can breach a drawdown limit even when price would have recovered. Size triggers far inside the firm's limits, or skip it there entirely.
  • It repairs trades, not strategies. If the underlying strategy keeps producing losers larger than the cycles can recover, the hard cap will keep paying out those losses. The EA makes bad outcomes bounded; it does not make them profitable.

Prudent configuration: a checklist

If you run a recovery EA on MT5 — ours or anyone's — this is the configuration order that keeps it survivable:

  1. Set the hard dollar loss cap first. This is the amount you accept to lose per cycle, decided when you are calm — not negotiated with the market at 3 a.m. If an EA has no cap, do not run it.
  2. Small trigger, small grid. A trigger of a few dollars (or a fraction of one percent of equity) with few grid levels and modest lot multipliers fails slowly enough for you to act.
  3. Cap the maximum grid levels and total lots. Know the worst case in dollars before the first cycle ever starts: max levels × max lots × symbol value.
  4. Prefer a hedging account. Locks and mirror grids need opposite positions on the same symbol. On netting accounts those features disable themselves — know which mode you are on.
  5. Mind the swap. Check triple-swap days and weekend holds on your symbols; high negative swap can quietly turn a break-even target into a losing one.
  6. Respect the news. High-impact events (NFP, CPI, FOMC) produce spikes that walk through grids. A news filter or simply pausing around red-folder events removes the worst single scenario.
  7. Test before you trust. Strategy Tester first, then demo, then minimum live size. Details below.
  8. Keep margin headroom. If a cycle running at full extension would push your margin level anywhere near stop-out, your size is too big. Recovery adds leverage when you least want it.

Who a recovery EA fits (and who should skip it)

A recovery manager makes sense for traders whose entries are fine but exits are not: manual traders who hate closing at a loss and let losers run, or algorithmic traders whose entry system fires webhooks but does not manage the open trade afterwards. It also fits traders who already cap their risk and want the losers handled systematically instead of emotionally.

It makes no sense on accounts without margin headroom, on prop-firm challenges with tight daily drawdown rules, or for anyone hoping it will rescue a strategy that simply loses. A recovery EA changes when you take a loss and how much it costs; it does not change whether the underlying strategy has an edge.

How to test one safely

Stage 1 — Strategy Tester. Many recovery EAs include a tester helper that auto-opens a losing trade every N bars so you can watch complete cycles without writing your own losing strategy. Fix the direction (buys only, then sells only) and a fixed date range so runs are reproducible, and watch what the equity curve does around the cap.

Stage 2 — Demo. A few weeks of real ticks on a demo account, with the same config you intend to run live. This is where you discover swap, spread and slippage behaviour the tester rounds away.

Stage 3 — Live, minimum size. Smallest lot your broker allows, hard cap on, one symbol. Scale only after you have watched real cycles resolve. If you trade prop-firm rules, revisit the drawdown math before this step — our guide to backtesting prop firm rules covers how to measure a tool like this against daily loss limits.

FAQ

A recovery EA is a position manager, not an entry strategy. It watches trades you already have open (manual, other EAs, or webhook entries) and, when one loses more than a dollar trigger you set, it starts a recovery cycle: an averaging grid, a basket take-profit and partial closes designed to bring the position back to break-even or better.
They share the same core math: both add to losing positions, so losses can compound. The difference is scope and control. A martingale bot opens and compounds trades on its own; a recovery EA only reacts to positions that already exist and can be capped with a hard dollar loss cap, a maximum number of grid levels and an equity stop. Without those caps, a recovery EA behaves exactly like a martingale.
Yes. The averaging cycle increases exposure while it runs, a strong one-way move against the basket can stack a large drawdown, and long cycles bleed swap and commission. Every serious recovery EA ships with a hard loss cap that accepts a defined loss instead of holding to the end — that cap is the feature that keeps the tool survivable, and you should always configure it.
They can, but they are the highest-risk tool to run on one. Prop firms measure daily and overall drawdown on equity, and a recovery cycle in progress pushes equity down before it pulls it back. If your daily loss limit is close, an open cycle can breach it even when price later recovers. If you run one on a funded or challenge account, size the trigger and hard cap far inside the firm's limits.
Hedging accounts are recommended. Features like locking the loss with an opposite position or running a mirror grid on the other side both need opposite positions on the same symbol, which netting accounts cannot hold. Good EAs detect the account mode and disable those features automatically on netting, but the recovery toolbox is wider on hedging.
In three stages. First, the Strategy Tester: many recovery EAs include a test mode that auto-opens losing trades so you can watch complete cycles; fix the direction and the date range for reproducible runs. Second, a demo account for a few weeks of real ticks. Third, live with the smallest size and the hard cap on. Never take an untested grid straight to a funded account.

Disclaimer: SignalForge AI is an order-execution and position-management software maker. We do not provide investment advice, trading signals, or recommendations. Trading involves substantial risk of loss. Profitability depends entirely on your own strategy and decisions. Past performance does not guarantee future results. Trade only with capital you can afford to lose.