How many losing trades in a row is normal?

Ten losing trades in a row feels like proof that something is broken. For a strategy that loses 70% of its trades, it is a normal event. This page uses 33,251 closed trades from one systematic crypto futures strategy to show how long losing streaks get in practice, and why simple probability predicts them.

What the record shows

Across 134 markets the strategy went through 6,812 separate losing streaks. Most were short: 76% lasted between 1 and 4 trades. 1,664 reached 5 losses or more, 307 reached 10 or more, and the longest lasted 29 trades on 1INCH.

Per market, the longest streak was typically 13 trades, and 110 of the 134 markets had at least one run of 10 losses or more.

Losing streaks by length1 in a row1,9792 in a row1,4683 in a row9874 in a row7145 in a row5046 in a row3247 in a row2348 in a row1819 in a row11410 in a row7911 in a row6212 in a row5013 in a row3714 in a row2115 or more58

Probability predicts it

If each trade loses with probability q, the chance that the next k trades all lose is q to the power of k. At this strategy's loss rate of 0.70:

  • 3 losses in a row: 33.9%
  • 5 losses in a row: 16.4%
  • 8 losses in a row: 5.6%
  • 10 losses in a row: 2.7%

Each of those probabilities applies at every trade, and a market produces hundreds of trades, so long runs become likely somewhere along the way. A standard approximation for the longest losing run in n trades is log(n × (1 - q)) / log(1 / q). A typical market here had about 248 trades, which gives an expected longest streak of about 12. The real median was 13.

What a streak costs

Ten losses at the average loss of 3.0% compound to a 27% drawdown on the balance trading that market. That number explains why the stop matters as much as the entry. With every loss capped near 3%, a long streak is painful and survivable, and the occasional large winner, +19.2% on average, pays for it.

Takeaways for testing any system

  • Work out the streak your loss rate predicts before you start, and check that your position size survives it.
  • A streak inside the predicted range is ordinary variance. A streak far beyond it deserves a closer look at the strategy.
  • Judge a strategy over hundreds of trades. Ten or twenty trades mostly measure luck.

About the data

Every figure on this page is computed from AlphaFeed's closed trades: 33,251 trades across 134 crypto perpetual futures markets on 1-hour bars, from 2020 to October 2026. Returns are the percentage gained or lost on each trade, losing trades included. The numbers update as new trades close.

This is research on one systematic strategy and is not financial advice. Past results do not predict future ones, and leveraged futures can lose more than the money put in.

The full results are on the record, with each market on the markets page. New trades, winners and losers, are posted in the free AlphaFeed Telegram channel.