Martingale, grid, and averaging-down systems share one feature: they win very often. A typical equity curve rises smoothly, with small gains almost every week and shallow drawdowns. That is exactly the profile that looks best on a signal provider’s profile page: high win rate, steady monthly returns, an impressive profit factor in the early months.
The catch is the shape of the risk. These systems collect many small wins and carry a rare, very large loss. Most of the time you only see the wins. They are not scams, and many are run honestly, but their risk is hidden in the tail, and the tail eventually arrives.
How the main systems work
Martingale
After every losing trade, the next trade is made larger, usually doubled. The idea: when a trade finally wins, it recovers all previous losses plus one original unit of profit.
Example with a $10 starting stake and a 50/50 outcome:
Consecutive losses | Next stake | Total committed |
0 | $10 | $10 |
3 | $80 | $150 |
6 | $640 | $1,270 |
9 | $5,120 | $10,230 |
The system looks safe because nine losses in a row is rare: about 0.2% (1 in 512) per sequence. But it is repeated hundreds of times. Over 500 sequences, the chance of at least one nine-loss streak is roughly 60%. The stake needed grows exponentially while the account does not, so sooner or later the account, the margin, or the broker’s maximum lot size runs out. The one big loss typically wipes out all the accumulated small profits.
Grid trading
The system places buy or sell orders at fixed price intervals (for example every 20 pips) and closes each position for a small profit. In a ranging market it keeps collecting profits. In a strong trend against the grid, positions pile up in the wrong direction.
Illustration: a grid buying 0.01 lots on EURUSD every 20 pips lower. After a 500-pip fall there are roughly 26 open positions and an unrealised loss of about $650. On a $1,000 account that is most of the balance, and the market has not stopped falling. Many grids also increase lot size at each level, which makes this much faster.
Averaging down and “recovery” systems
Instead of closing a losing position, the trader adds to it at a lower price to reduce the average entry. Hedging or lock-and-recover variants open an opposite position and reopen later. All of these avoid realising a loss, so the account appears healthy until margin runs out. The loss is deferred, not avoided.
Trading without a stop-loss
Holding every position until it turns profitable gives a very high win rate, because almost every trade eventually closes in profit, until the one that never recovers within the account’s capacity. It is the same pattern in simpler form.
Why they usually fail
• Negative skew. Many small gains, rare huge losses. A 90% win rate says nothing if the average loss is 30 times the average win.
• Leverage and margin. The floating loss grows faster than the balance. A margin call can close the whole position at the worst moment.
• Exponential position size. Doubling is mathematically unsustainable; no account is large enough for a long losing streak.
• Trends and gaps. Markets can move in one direction far longer than expected. News events and weekend gaps can skip any planned exit.
• Hidden costs. Spreads, swaps (overnight fees) on large open positions, and slippage grow with exposure.
• Short track records. A few good months is not a sample of the market’s full range. A system can run through a calm period without ever meeting the conditions that break it.
• Survivorship bias. The failed accounts disappear from leaderboards, so the visible survivors look better than the average.
How to recognise them on a signal provider’s profile
You can often spot these systems before you copy them:
• A very high win rate (often above 85%) combined with an average losing trade much larger than the average winning trade.
• A big gap between realised results and the maximum drawdown or floating loss.
• Lot sizes that increase after losses, or many trades open at the same time in the same direction.
• Trades held for unusually long periods, or no stop-loss at all.
• A history shorter than a year, or one that never covered a strong trend.
• Equity much lower than balance during open drawdown periods.
A hint is not a verdict. A provider using a grid with strict limits and a real stop can be a legitimate strategy. The questions to ask are what the maximum loss is, and what happens when the market goes the wrong way for a long time.
If you still want to copy one
These systems can make money for a while, and some investors choose to accept that risk knowingly. If you do, reduce the damage of the likely failure:
• Allocate a small amount. Use only money you can afford to lose entirely, a small share of your total capital.
• Set your own protection. Use the copier-side stop-loss or maximum drawdown limit, so your copy stops even if the provider’s account does not.
• Limit the copy ratio. Do not copy at a higher multiplier than needed.
• Withdraw profits regularly. Treat the strategy as having a limited lifespan and take money out while it works.
• Diversify. Do not copy only martingale or grid providers; combine them with other styles and several providers.
• Check the open exposure, not just the closed trades. Look at equity, floating loss, and the number of open positions at least weekly.
• Decide your exit in advance. Define the drawdown at which you stop copying and stick to it.
Conclusion
Martingale, grid, and recovery systems are attractive because they win often and look safe, and they can run well for a long time. They are also designed so that the failure, when it comes, is large and fast, and most of them eventually fail. They are not a reason to avoid copy trading, but they are a reason to read the full risk profile, size your allocation accordingly, and put your own limits in place.
This article is for educational purposes only and does not constitute investment advice. Trading leveraged products such as forex and CFDs carries a high risk of loss, and past performance does not guarantee future results. Copying a signal provider does not remove that risk.
En fallende aksjeindeks kan skape inntrykk av at salg sprer seg over hele markedet. Handelen 22. oktober viste hvorfor den tolkningen kan være misvisende. S&P 500 falt 1,2 %, og Nasdaq Composite falt 2,6 %, men omtrent to tredjedeler av aksjene i S&P 500 endte høyere.
Les mer →Microns rekordoverskudd har gjenopplivet en velkjent bekymring for hvor mye av en oppgang i minnemarkedet som kan overleve den neste tilbudsutvidelsen. KI endrer den beregningen gjennom kapasitetskrevende produkter, langsomme fabrikkutbygginger og bindende kundeforpliktelser. Disse kontraktene tilbyr en måte å dele investeringsrisiko på, selv om utgiftene de støtter til slutt kan skape forutsetningene for en ny nedtur.
Les mer →