Trading

Martingale Forex: How It Works, Why It Fails, and What to Use Instead

Trader reviewing a forex position at her workstation

Martingale in forex means doubling your position size after every losing trade, so that one eventual winner recovers all earlier losses plus a small profit. It looks airtight on paper. In practice it ends the same way whenever the losing streak outlasts your account. The mirror image, anti-martingale, adds size only after wins, and pyramiding adds to a trade that is already working. Those two keep your risk under control.

We have watched plenty of traders reach for martingale after a bad week. It feels like a plan, but it is really a revenge trade with a spreadsheet attached. Below we cover the mechanics, the numbers, the UK rules that apply to you, and the sizing methods that hold up better.

Tradergibkey
tradergibkey.eu
Build a More Structured Trading Approach
Trader Gibkey teaches practical price action strategies shaped by more than 18 years of live market experience.
Visit

TL;DR

  • Martingale doubles your position after each loss. It wins small, often, until one streak wipes out the account.
  • Tests and backtests show why: Edge or Myth’s EUR/USD test since 1999 found single losing trades that took accounts below zero.
  • UK retail leverage is capped at 30:1 on major pairs, so the huge positions martingale needs are out of reach for most accounts.
  • Anti-martingale (size up after wins) and pyramiding (add to winners) point risk in the opposite direction and keep losses small.
  • Fixed risk per trade beats any doubling system. Pick a hard rule and keep it.

Table of Contents

What martingale means in forex

The martingale idea started as a gambling system in 18th-century France: double your stake after each loss, and the first win recovers everything. Investopedia’s overview of martingale in forex describes the same logic applied to currency trading, where traders double their position after a loss to recover earlier losses and make a profit.

In forex it usually looks like this:

  1. You buy EUR/USD with 1 lot.
  2. Price drops 20 pips. You close or hold, then open 2 lots in the same direction.
  3. Price drops again. You add 4 lots, then 8, then 16.
  4. When price finally bounces, your average entry sits much lower, so a small rally gets you back to break-even and a little profit.

The appeal is real. Currencies rarely go to zero, and the retracement you are waiting for often does arrive. That is exactly why the system feels like it works for weeks or months.

The maths: what doubling does to your account

Take Investopedia’s EUR/USD example. As the lot size doubles from 1 up to 32, the accumulated loss climbs from $0 to $6,200, while the move needed to break even is only +19 pips. A trader with a $5,000 account would be bankrupt before EUR/USD reached 1.2550, even though the pair only had to reach 1.2569 for the 32-lot position to break even.

That is the shape of the problem in one line: the break-even distance shrinks while the capital you need explodes.

Here is a simple version you can run yourself. Start with 0.10 lot, double after each loss, use a 25-pip stop and a £10,000 account:

Step Lot size Cumulative stop-outs
1 0.10 1
2 0.20 2
3 0.40 3
4 0.80 4
5 1.60 5
6 3.20 6

In a martingale calculator audit using these inputs (with 10,000 in equity), six steps bring the cumulative loss to 1,575, which is 15.75% of the starting equity. Six losses in a row sounds rare. Now picture the seventh step at 6.40 lots. Then the eighth. The loss is not linear, so the pain arrives very late and very fast.

Two more things quietly make it worse:

  • Costs scale with size. Capital.com’s guide to martingale risks notes that spreads and slippage erode profits as position size increases.
  • The system assumes unlimited capital. The same guide points out that martingale needs unlimited capital and a guaranteed eventual win, and neither exists in real markets.

Sounds strange, but it’s true: a martingale system can post a 100% win rate right up to the trade that ends the account. Investopedia’s Kathy Lien makes the same point, noting that strict martingale produces a perfect success rate until it ends in the complete loss of capital.

What the tests actually found

Backtests are where martingale looks best and where it hides most. Two sources are worth knowing.

Edge or Myth, EUR/USD since 1999. In their published test of martingale on EUR/USD, 45.0% of buying martingale accounts and 51.9% of selling ones ended with a profit. That sounds fine until you read the tail: in 41 buying accounts and 61 selling accounts, a single losing trade took the balance below zero, because the rate ran far past 50 pips before the trader could act.

FTMO’s 2016 backtest. A martingale system tested on a funded forex account made 25% to 35% profit in the first three tests. Yet 7 out of 20 test cycles lost the entire capital, which implies roughly a 35% chance of bankruptcy in that year. The article’s warning is that the risk is hidden from historical results.

Martingale backtest profit and bankruptcy results

That last point deserves a pause. A smooth equity curve is not proof of low risk. It can simply mean the bad streak has not shown up in your sample yet.

The one fair argument for martingale

In fairness, Lien notes that FX offers some advantage to well-capitalised traders, since a currency rarely falls to zero and interest income on positive-carry pairs can offset part of the losses. That helps a deep-pocketed institution with long horizons. It does little for a retail account holding leveraged positions with a margin requirement.

UK rules that collide with martingale

If you trade with an FCA-regulated provider, the rules cut against martingale in three ways.

Leverage caps. Under FCA Policy Statement PS19/18, in force since 1 August 2019, retail leverage is capped at 30:1 on major currency pairs and 20:1 on non-major pairs, gold and major equity indices. Summaries of the rules also note the 50% margin close-out and negative balance protection. Doubling a position five or six times on a small account hits that ceiling quickly. Edge or Myth makes the same point: in the EU and UK a trade of 1,280,000 euros on what is left of a $10,000 account is far beyond the cap.

Forced close-out. Under COBS 22.5.13R, a firm must make sure a retail client’s net equity in an account used for restricted speculative investments does not fall below 50% of the margin requirement, and must close positions as soon as market conditions allow when that level is breached. The Financial Ombudsman Service decision DRN-5208030 sets this out. It also confirms that closing the largest losing position first can be consistent with acting in the client’s best interests. In a martingale ladder, the largest losing position is the one you just doubled into.

Protections around the edges. The FCA’s original proposal for permanent retail CFD measures included leverage limits, close-out at 50% of required margin, protection against losses beyond the account funds, and a standardised risk warning showing the percentage of retail accounts that lose money. That framework is why a margin call in the UK is a rule-driven event, not a negotiation. CMC Markets’ explainer on UK spread betting protections lists the same family of safeguards, and notes providers must be FCA authorised.

Negative balance protection means a UK retail client should not owe more than the account holds. It does not mean your account survives. The account is gone.

UK rules that collide with martingale

Anti-martingale: size up when you are right

Anti-martingale (also called reverse martingale) flips the rule. You increase size after a win and go back to your base size after a loss.

A basic version:

  1. Choose a base risk per trade, for example 1% of the account.
  2. After a winning trade, raise the next position to 1.5% or 2%.
  3. After a second win, you may go one step further, then reset to base.
  4. After any loss, return to base risk immediately.

Why traders like it:

  • Losses stay small. A losing streak never gets a larger position, so the worst sequence costs you base-size trades.
  • Winners compound. When a trend is paying you, your size grows with it.
  • It matches how edges behave. Strategies often have runs of good and bad conditions. Sizing up in good conditions and down in bad ones fits that.

The trade-off is that a string of small losses will cost you steadily, and one big loss right after a size increase can eat several wins. That is why we cap the step-up. A hard rule such as “never more than 2% on any single trade, even on a hot streak” does that job.

Pyramiding: adding to a trade that is already working

Pyramiding is the trade-level cousin of anti-martingale. You open a starter position, and if price moves in your favour and confirms the idea, you add smaller positions at higher levels (for a long trade). The key difference from martingale: you add to a winner, never to a loser.

A simple pyramid for a long trade:

  1. Starter: 1% risk at your first setup, with a stop below structure.
  2. First add: after price breaks and retests a level in your favour, add a smaller position, for example half the starter.
  3. Move the stop: bring the stop on the whole position to at least break-even, so the added risk is paid for by open profit.
  4. Second add: only if the trend still shows clean structure, and smaller again.

The pyramid gets narrower toward the top for a reason. Each add has a worse entry price, so a pullback hits the later layers hardest. Keeping adds smaller and stops tightened means the total risk on the position does not grow even as exposure does.

The common failure is pyramiding without a stop change. Then you have simply made a bigger position at a worse average price, with the same stop distance. That is martingale logic wearing a trend-following coat.

Martingale, anti-martingale and pyramiding side by side

Martingale Anti-martingale Pyramiding
When you add size After a loss After a win While a trade is in profit
Average entry Improves Not applicable Worsens
Worst case Account wiped by one streak Several small losses, one large loss after step-up Pullback hits the later layers
Typical win rate profile High, with rare catastrophic losses Normal, with variable size Normal, varies by trend
Fits Almost no retail account Trend-following, momentum systems Clean trending markets
Main control None that scales Capped step-up and reset Smaller adds and a moved stop

No sizing method makes a bad entry good. Sizing only decides how much a good or bad series costs you. If you want the maths behind that, our guide to win rate versus risk/reward shows how break-even changes with each combination.

What to do instead of doubling down

Most people reach for martingale after a losing run. The brain stops trading the chart and starts trading the pain. These steps swap that impulse for a rule.

  1. Fix your risk per trade. Choose a percentage and keep it. Our guide to setting risk per trade walks through the sizing formula.
  2. Write a daily loss limit. When you hit it, you stop. No exceptions. See our notes on position sizing in practice.
  3. Calculate risk of ruin. Know how likely a streak is to end your account, using our article on risk of ruin in trading.
  4. Test before you trust. Run any sizing idea on historical data, then on demo, as laid out in our backtesting guide.
  5. Journal every size change. If you ever increase size after a loss, write down why. Usually the honest reason is “I wanted it back.”

Perspective

We built Trader Gibkey on more than 18 years of live market experience, and one pattern shows up again and again: traders do not blow up because they lack a clever sizing formula. They blow up because they are trying to fix a feeling.

Martingale is attractive because it promises that a loss is only temporary. A good mentor will tell you the opposite. A loss is information, and the right response is to protect your decision quality, not to bet bigger on a hunch.

That is why our teaching starts with price action and risk before it gets to any sizing trick. When you read structure well, you take fewer, better setups, and a fixed percentage risk is enough. Anti-martingale and pyramiding make sense inside that framework, because you add only when the market has already shown you it agrees.

If you have been stuck in a loop of revenge trades, a structured plan and someone to review your journal with you changes the outcome more than any new system does.

Gabriel

Work on your sizing with us

If you want to replace guesswork with a structured approach, we teach price action, risk management and trading psychology step by step. The Price Action (Core) plan is £149 per month, and our New Generation Liquidity plan is €349 per year. If you want the full programme for good, Diamond (Lifetime Access) is a €3,499 one-off payment.

Tradergibkey

Prefer direct feedback on your own trades and sizing? Mentorship is available as a single session at €149, three sessions at €399, or five sessions at €559. You can see every option on the Trader Gibkey site.

FAQ

Does martingale work in forex?

It often wins small for a long time, but one long losing streak can wipe out the account. In Edge or Myth’s EUR/USD test, a single losing trade took some accounts below zero. Fixed-percentage risk is the safer base.

What is the difference between martingale and anti-martingale?

Martingale increases size after a loss. Anti-martingale increases size after a win and returns to base size after a loss. The second keeps losing streaks small.

Is pyramiding the same as averaging down?

No. Pyramiding adds to a trade that is already in profit, ideally with the stop moved up. Averaging down adds to a losing trade, which is martingale logic.

Sources

Want to learn the full system?

Join the mentorship and work directly with Gibkey for 60 days. Personal trade reviews, live sessions, and a complete trading plan tailored to you.

Explore Mentorship