Short answer: the martingale strategy is a betting system where you double your stake after every loss. The first win then recovers every earlier loss and adds one unit of profit. It works on paper because a losing streak cannot last forever. It fails in practice because your bankroll, the betting limits, and your nerve all run out before "forever" arrives. In trading, the same idea shows up as doubling position size after a losing trade or averaging down into a falling market. Treat it as a lesson in risk management, not as a plan.
This page is for traders who have seen the martingale system praised in a forum or an expert advisor listing and want a plain verdict.
How the martingale strategy works
The mechanics fit on a napkin. Pick an even-money bet, one where the odds are close to 50/50 and a win pays the same as the stake. Wager one unit. If you win, pocket the unit and start again. If you lose, double the stake and place it again. Keep doubling until a win arrives. Then reset to one unit.
Here is the ladder on a 10,000 bankroll with a 100 base stake.
| Step | Stake | Total lost if this step loses | Profit if this step wins |
|---|---|---|---|
| 1 | 100 | 100 | 100 |
| 2 | 200 | 300 | 100 |
| 3 | 400 | 700 | 100 |
| 4 | 800 | 1,500 | 100 |
| 5 | 1,600 | 3,100 | 100 |
| 6 | 3,200 | 6,300 | 100 |
| 7 | 6,400 | 12,700 | 100 |
Two things jump out. First, the reward never grows. Whether you win on step one or step six, you make exactly 100. Second, the exposure grows violently. By step six you are risking 3,200 to win 100, and you have already lost 6,300. Step seven needs 6,400, and after six losses only 3,700 remains. The ladder breaks there. The whole sequence ends with a 6,300 loss that would take sixty-three earlier wins to cover.
That asymmetry is the entire story. A martingale converts many small, tidy wins into one rare, catastrophic loss. On a fair game the average outcome is unchanged; only the shape of the results changes, and it changes for the worse.
Where the martingale came from and which games it was built for
The name is usually traced to eighteenth-century France, where the doubling method was popular among gamblers at even-money games. The classic home for the martingale strategy is roulette: red or black, odd or even, high or low. Baccarat, craps and sports bets priced close to evens get the same treatment.
None of these games is truly 50/50. A roulette wheel with a zero pays even money on a bet that wins less than half the time. That house edge is small on any single spin, but the martingale asks you to place many spins, and the edge is charged on each one. A bookmaker's margin plays the same role. Every doubling step gets slightly worse odds than the ladder assumes.
Then there are betting limits. Casinos set a maximum stake on every table, and sportsbooks cap what they will accept on a single line. On a 10 minimum table with a 1,000 maximum, the seventh step of the ladder (640) still fits but the eighth (1,280) does not. Limits vary widely between venues, so a ladder that survives at one table can be cut short at the next.
Risks and limitations of the martingale strategy
The advantages are honest but thin. A martingale gives frequent small wins, a simple rule that removes hesitancy about stake size, and a reset that feels like a fresh start. Its attractiveness is emotional: it turns an ordinary losing session into a story about the comeback.
The limitations are consequential.
- The payoff is stingy. The best case on every ladder is one base unit.
- The worst case is unforgiving. It equals the sum of every stake on the ladder, and it arrives all at once.
- Long losing streaks are common. On a fair coin, six straight losses happen about once in every 64 sequences, and ten straight losses about once in 1,024.
- The bankroll caps the ladder. With 10,000 and a 100 base stake you can absorb six losses, and no more.
- Betting limits cap it again from the other side.
- Costs compound against you. Whether it is a house edge, a bookmaker margin, or spread and commission cost on a trade, the charge grows with the stake.
The long-term maths is blunt. Run the six-step ladder above and stop at the cap. In most sequences you collect 100. In roughly one sequence out of 64 you lose 6,300, which is exactly sixty-three wins of 100. So over many cycles the expected result is close to zero on a fair game and negative once any edge or cost is added. The system does not create money; it only rearranges when you lose it, and it does so needlessly.
What martingale looks like in forex, indices, metals and crypto
Traders rarely call it a martingale. They call it adding to the position, averaging down, scaling in, or running a grid. The mechanics are identical. A short on gold moves against you by 50 pips, so you sell again with double the lots. It moves another 50, so you double again, hoping one small bounce brings everything back to breakeven.
Three things make trading worse than the casino for this idea. First, there is no fixed payout. A trade can lose far more than the planned stop if price gaps or the market turns disorderly. Second, margin replaces the table limit. Each doubling step demands more margin, and a margin close-out can shut the whole stack at the worst possible moment. Third, trends exist. A coin has no memory, but a market can fall stubbornly for weeks. A doubled short into a strong crypto rally is a precarious place to be.
A 10,000 CFD account running a martingale from 0.1 lots on an index reaches 3.2 lots by step six. The stop that once cost 100 now costs 3,200 per 50-point move. The trader who was calm at 0.1 lots is jittery and skittish at 3.2. This is the point where a bad sequence slides into revenge trading, because the ladder has quietly become an argument with the market.
Martingale versus other staking systems
The clearest way to judge a martingale is next to the alternatives.
| System | After a loss | After a win | Worst case | Who it suits |
|---|---|---|---|---|
| Martingale | Double the stake | Reset to base | Sum of the whole ladder, all at once | Almost nobody with a finite bankroll |
| Anti-martingale (Paroli) | Reset to base | Double for a set number of steps | Give back one streak's profit | Traders who want size to follow winning streaks |
| Fixed fractional | Risk shrinks with the balance | Risk grows with the balance | A slow, lumpy drawdown | Most retail traders |
| D'Alembert | Add one unit | Remove one unit | Long grind, sluggish recovery | Gamblers who want a gentler ladder |
| Kelly fraction | Stake follows edge and odds | Same rule | Very wobbly equity if the edge is overestimated | Traders with a measured, stable edge |
Fixed fractional sizing, risking a steady one percent of the current balance, is the unremarkable default for good reason. It decelerates your losses automatically when things go wrong, the exact opposite of a martingale, and that makes a bad run far more forgiving. A consistent edge sized this way beats any staking trick, and our guide on how to grow a small trading account covers that path.
A decision framework for anyone still tempted
Before you gamble on a martingale, or its trading cousin, run through this checklist and be honest.
- Bankroll size. Count the steps you can afford. With 10,000 and a 100 base, the answer is six. If that feels small, the base stake is too big.
- Betting limits. Check the table maximum, or the margin your broker demands at the top of the ladder. A ladder that outruns the limit is certain to break.
- Odds and cost. If the game or the trade is not close to even money after the house edge, the spread, and the commission, every step is priced against you.
- Risk tolerance. Ask whether you could watch 6,300 disappear in one sequence and still follow your rules the next day. Most people cannot.
- Time horizon. Over a handful of sessions a martingale usually looks brilliant. Over hundreds of sessions the rare streak arrives, and that is the horizon you actually live in.
- Recovery maths. Once an account is cut in half, it must double just to return to where it started. A doubling system makes that cushion threadbare very fast.
The risk of ruin framing is the cleanest verdict. For a fixed-fraction trader, ruin is a remote tail. For a martingale, ruin is the scheduled end of the sequence, and the only unknown is the date. Never let one decision put more than a fixed, small slice of the account on the line, and never make the next stake depend on the last result. This is what bankroll management means in practice.
Martingale inside prop-firm evaluation limits
Evaluation accounts add hard lines that a martingale crashes into very early. Suppose a 100,000 evaluation account carries a 5,000 daily loss limit and a 10,000 total drawdown. A trader starts a ladder at 1,000 of risk. Step one loses 1,000. Step two loses 2,000, total 3,000. Step three loses 4,000, total 7,000. The daily line was breached on step three, before the ladder reached the win it was waiting for. Three ordinary losses in a row, which a fair coin delivers one time in eight, ended the attempt.
Consistency rules make it worse, because a lopsided equity curve of tiny wins and one enormous loss is exactly what they are built to catch. Static total drawdown does not save a doubling ladder either; it only means the floor stays put while the stakes climb toward it.
Pivex states both lines before you start an attempt. The total drawdown is static rather than trailing, and the daily loss limit resets each day. Its rule set also bars automated grid systems, which are the most common way a martingale gets packaged into an expert advisor. The built-in assistant watches your limits and flags a stake that is about to cross one. That is useful, but the better fix is to never build the ladder in the first place.
If a martingale ladder is already part of how you trade, write down a fixed risk per trade and stop letting the last result choose the next stake. Test that rule on a sample of trades before scaling anything. Pivex's one-step evaluation has no time limit and no minimum trading days, which removes the usual excuse for forcing size after a bad week.
Related guides: risk of ruin in trading and revenge trading.
Before participating, review Pivex's current program rules and decide whether a simulated proprietary trading evaluation fits your experience and risk limits. Participation does not guarantee a reward, income, or any particular trading result.