Martingale Trading and Grid Strategies: How They Work and Why They Fail
Updated 2026-11-16 · 9 min read
Martingale trading means increasing your position size after every loss, usually by doubling it, so that one winning trade recovers everything lost before it. Grid strategies work in a similar way: they keep adding positions as the price moves against them. Both can produce smooth profits for months, but they carry a hidden risk that one strong trend can wipe out the whole account. This guide uses simple numbers to show how they work and how to spot them in a trading bot's track record.
If you have looked at trading bots, copy-trading platforms or automated strategies, you have probably seen equity curves that rise in an almost perfect straight line. Very often, a martingale trading or grid system sits behind that smoothness. These approaches are not new. They are simple to code and very appealing at first, but the way they manage risk is badly flawed. Below we explain how they work, take you through worked examples with round hypothetical numbers, and set out the warning signs and safer alternatives.
What is martingale trading?
The martingale idea comes from gambling. You bet an amount, and if you lose, you double the next bet. When you eventually win, the payout covers all your previous losses plus a profit equal to your first stake. In trading, the same logic is applied to position size: after each losing trade, the next trade is opened at double the size, often in the same direction.
The reasoning sounds convincing: sooner or later the market must turn in my favour. The trouble is that the strategy needs unlimited capital and unlimited time, and no trader has either. Brokers also set maximum position sizes and margin requirements, so the doubling cannot go on for ever.
A worked martingale example
Imagine a hypothetical account of £2,000. The strategy risks £10 on its first trade and doubles the stake after every loss. Each trade either wins or loses an amount equal to the stake.
| Trade | Stake | Result | Cumulative loss before this trade |
|---|---|---|---|
| 1 | £10 | Loss | £0 |
| 2 | £20 | Loss | £10 |
| 3 | £40 | Loss | £30 |
| 4 | £80 | Loss | £70 |
| 5 | £160 | Loss | £150 |
| 6 | £320 | Loss | £310 |
| 7 | £640 | Loss | £630 |
| 8 | £1,280 | ? | £1,270 |
If any of the first seven trades wins, the account ends the cycle just £10 up. That is the attraction: most cycles do end in a small profit. After seven losses in a row, though, £1,270 has gone and the eighth trade needs a £1,280 stake. Only £730 is left. The strategy cannot carry on, and the trader has lost well over half the account trying to make £10.
Notice how lopsided this is. The potential gain per cycle is fixed at £10, while the potential loss grows exponentially. With a 50/50 outcome on each trade, seven losses in a row has a probability of 1 in 128. That sounds rare, but a bot that runs hundreds of cycles is likely to meet a streak like this sooner or later. Real markets are not independent coin tosses either. Trends can make long losing runs more likely than simple probability would suggest.
What is grid trading?
A grid strategy places a series of buy and/or sell orders at fixed intervals above and below the current price, which builds a grid of levels. Each order usually has a small profit target. When the price swings up and down within a range, orders are repeatedly filled and closed at a profit.
Grid trading is not automatically martingale. A grid with a fixed size per level and a firm overall stop loss is a different thing from one that doubles size at each level and never closes losing positions. In practice, though, many retail grid bots have no stop loss and keep adding exposure as the price moves against them. That gives them the same weakness as martingale: losses can grow without limit while profits stay small.
A worked grid example
Suppose a hypothetical asset trades at 100. A bot buys 1 unit at 100 and then another unit at every 1-point drop (99, 98, 97 and so on). Each unit has a take-profit 1 point above its entry price. There is no stop loss.
- Ranging market: the price moves between 97 and 100 for weeks. Units bought at 97, 98 and 99 are sold repeatedly at 98, 99 and 100. Each round trip earns 1 point. The account grows steadily and the balance chart looks excellent.
- Trending market: the price falls steadily from 100 to 80 without a meaningful bounce. The bot now holds 20 units, bought at 100, 99, 98 and so on down to 81. At a price of 80, the floating loss is 20 + 19 + 18 + … + 1 = 210 points.
Each successful round trip earned 1 point, so this single trend has erased the equivalent of 210 winning trades, and the loss keeps growing if the price falls further. If the bot also increases size at each level (1 unit, then 2, then 4), the floating loss grows far faster still. This is the classic martingale grid.
Martingale vs grid vs fixed-risk trading: a comparison
| Feature | Martingale | Grid (no stop loss) | Fixed risk per trade |
|---|---|---|---|
| Position size after a loss | Doubles (or multiplies) | Adds more positions as price moves against it | Stays the same percentage of the account |
| Typical win rate | Very high | Very high | Varies; often lower |
| Equity curve in normal conditions | Very smooth | Very smooth | Uneven, with visible drawdowns |
| Maximum loss | Potentially the whole account | Potentially the whole account | Limited and known in advance per trade |
| Main danger | A long losing streak | A strong, sustained trend | Gradual losses if the strategy has no edge |
| Use of stop loss | Rarely | Often none | Always |
Why martingale and grid strategies look great for months
The most dangerous feature of these systems is that they do not look dangerous. Several things combine to make them seem attractive.
A very high win rate
Because losing positions are held open or covered by larger trades, almost every closed cycle ends in profit. A track record showing a 95% or higher win rate can feel reassuring. Usually it means the losses have not been realised yet, not that they never happened.
Balance and equity tell different stories
Trading platforms track two key figures. Balance reflects closed trades only. Equity includes the profit or loss on positions that are still open. A grid bot can show a balance rising steadily while its equity falls sharply because of losing trades it has not closed. Many performance charts show only the balance, which hides the real risk.
Markets often range, until they don't
Many markets spend a lot of time moving sideways, and that is exactly the environment where grid and martingale systems do well. A strong trend can follow a central bank decision, an unexpected economic release, a geopolitical event or a sudden shift in sentiment. When it comes, the strategy has no mechanism for accepting a small loss and moving on.
Survivorship bias
The track records you see are the ones that are still alive. Systems that blew up are often deleted, restarted under a new name or simply forgotten. A sample made up only of survivors makes the approach look more reliable than it is.
How one strong trend can wipe out an account
The failure is usually sudden rather than gradual. As the price moves further against the open positions, three things happen at once:
- Floating losses grow faster than linearly, because more positions (or larger positions) are being added the further the price moves.
- Margin usage rises as each new position needs more margin, so free margin shrinks.
- The broker steps in: once equity falls below the required margin level, a margin call or stop-out closes positions automatically, often at the worst point of the move.
Months of small, steady gains can disappear within a few days, or even a few hours. Because the losses come all at once, the trader rarely has time to react. Overnight price gaps make things worse: a market can open far beyond the last grid level, which turns floating losses into realised losses at a price nobody chose.
How to spot martingale or grid trading in a bot's track record
You do not need access to a bot's code to recognise these patterns. Look for the following warning signs in any track record or performance report:
- An unusually smooth equity curve. Real trading involves losing periods. A line that rises with almost no dips over many months deserves closer inspection.
- Many open trades at the same time, especially on the same instrument and in the same direction at different entry prices.
- Increasing lot sizes in a sequence of trades, for example 0.01, 0.02, 0.04, 0.08.
- A huge gap between balance and equity, or a large maximum floating drawdown compared with closed-trade drawdown.
- A very high win rate with a small average win and a very large average loss. A handful of enormous losses against hundreds of tiny wins is a typical signature.
- No stop losses on individual trades.
- Losing trades held for weeks while winners close in minutes or hours.
- A short track record, or one that began recently after an earlier account was closed.
If the provider shows only balance growth and will not share equity drawdown, open-position history or lot sizes, treat that as a red flag in itself.
Safer alternatives to martingale trading
No trading approach is risk-free. Some methods, however, keep losses limited and predictable rather than open-ended. These principles form the basis of conventional risk management.
Fixed risk per trade
Instead of increasing size after a loss, risk the same small percentage of the account on every trade. This is often called fixed fractional position sizing. For example, with a hypothetical £10,000 account and a 1% risk rule, each trade risks £100. After a loss, the account is £9,900 and the next trade risks £99. Position size shrinks a little in losing periods and grows a little in winning ones.
Compare what ten losses in a row would do. Under martingale starting at £10, the tenth trade alone would need a £5,120 stake, and the cumulative losses would be £10,230 by the end of that streak. Under a 1% fixed-risk rule, ten consecutive losses cost roughly 9.6% of the account. That is painful, but survivable.
A stop loss on every trade
A stop loss defines in advance the point at which a trade is wrong and should be closed. It turns an unknown, potentially unlimited loss into a known one. Stop losses are not perfect. Fast markets and gaps can cause slippage beyond the stop level. Even so, they are far safer than holding losing positions in the hope of a reversal.
One position at a time (or strictly limited exposure)
Limiting the number of open positions, or the total risk across all of them, stops exposure from quietly piling up. A simple rule such as never more than one open trade per instrument rules out the stacking behaviour that sinks grid systems.
Accept a lower win rate
Strategies with sound risk management often win fewer than half of their trades but aim for winners larger than losers. That makes for a less comfortable equity curve, with visible drawdowns. It is, however, a more honest picture of risk than a curve that hides losses inside open positions.
Questions to ask before using any automated strategy
Whether you are considering a bot, a signal service or a managed automated strategy, it is worth asking direct questions. Does it use a stop loss on every trade? Does it ever increase position size after a loss? What is the maximum number of open positions? What was the largest equity drawdown, not just balance drawdown? Can I see the full trade history, including lot sizes?
It also helps to understand where your money is held. With some services, such as Cortex Automation, the funds stay in the client's own broker account, so you can watch open positions, equity and margin in real time. That kind of visibility makes it easier to check for yourself how risk is being managed, whichever provider you use.
Key takeaways
- Martingale trading increases position size after losses so that one win recovers them. It produces small, frequent gains and rare but potentially catastrophic losses.
- Grid strategies without a stop loss share the same weakness: they collect small profits in ranging markets and can build very large losses in a trend.
- A smooth equity curve, a very high win rate, many simultaneous open trades and a large floating drawdown are classic warning signs.
- Fixed risk per trade, a stop loss on every position and limited exposure keep losses bounded and predictable.
Risk warning: trading leveraged products such as forex and CFDs carries a high level of risk and you can lose more than expected, including all of your capital. This article is general information for educational purposes only and is not financial advice. The examples use hypothetical numbers and do not represent any real strategy or result. Past performance does not guarantee future results. Consider your own circumstances and seek independent advice if you are unsure.
Frequently asked questions
Does martingale trading work?
It can produce small, frequent profits for a while, but it relies on unlimited capital to survive losing streaks. In real accounts, with limited funds and broker margin requirements, a long enough losing run or strong trend can cause losses far larger than all the previous gains combined.
Is grid trading the same as martingale?
Not exactly. A grid places orders at fixed price intervals, while martingale increases size after losses. However, many grid bots add positions without a stop loss or increase size at each level. That gives them the same open-ended risk as martingale.
Why do martingale bots show such high win rates?
Losing positions are usually kept open or recovered by larger follow-up trades, so most closed cycles end in a small profit. The win rate looks high because losses stay unrealised until a strong move forces them to be closed, often all at once.
How can I tell if a trading bot uses martingale?
Look for lot sizes that increase after losses, many open trades on the same instrument at different prices, no stop losses, a very smooth balance curve and a large gap between balance and equity or a high floating drawdown.
What is a safer alternative to martingale trading?
Common alternatives are risking a fixed small percentage of the account per trade, placing a stop loss on every position and limiting the number of open trades. These do not remove risk, but they keep each loss limited and known in advance.
What is floating drawdown?
Floating drawdown is the unrealised loss on positions that are still open. It shows up in the account's equity but not in its balance. It is the key figure for judging the real risk of grid and martingale strategies.
Want to see how an automated system works, with the money in your own account? See results and simulator →