Risk warning: binary options are high-risk. Most retail accounts lose money. Not offered to residents of the EU, UK, US or Australia.We earn a commission from some links.

Strategies

Why martingale fails in binary options (with the numbers)

By Updated 2 min read

Short answer

Martingale means increasing the stake after each loss so one win recovers everything. With payouts below 100% you must more than double each time, so stakes grow very fast: starting at $1 with an 85% payout, the 7th trade in a losing streak needs over $100, and after 8 losses the total lost passes $400. Losing streaks of that length are common, so martingale eventually empties the account.

Practise without risk $50,000 in virtual funds. No deposit. Try a $50,000 demo account Affiliate link. High risk of losing money.

How the stake grows

With an 85% payout, to recover all previous losses plus the original $0.85 profit, each stake must be about 2.18× the previous one.

Loss #StakeTotal lost so far
1$1.00$1.00
2$2.18$3.18
3$4.74$7.91
4$10.31$18.22
5$22.44$40.66
6$48.84$89.50
7$106.30$195.80
8$231.35$427.15

Stake n = (total lost + $0.85) ÷ 0.85.

How often streaks happen

At a 55% win rate, the chance of at least one run of 6 losses within 100 trades is about 36%. A run of 8 is about 8%. Over a few hundred trades, one of these is close to certain. See the losing streak calculator.

Why it feels like it works

Most of the time the streak ends after 1–3 losses and you collect a small profit. The rare long streak then takes back weeks of gains in minutes. The results look smooth — until they are not.

What to do instead

Use a fixed percentage stake (1–2% of balance) and a daily stop. Details in risk management.

FAQ

Is a soft martingale (×1.5) safer?

It slows the growth but has the same problem: a long enough streak still wipes the balance, and it no longer recovers losses in one win.

Research desk We open broker accounts from our target countries, test deposits and withdrawals, and log what we measure.

How this page was made: drafted with AI assistance from public sources and our own test logs, then checked and edited by the ExpiryLab editorial team. The date above shows the last check. How we test · Editorial policy