What is a losing streak in trading?

A losing streak is a run of trades that each end in a loss, one after another. Every method has them, because a positive trading expectancy is an average over many trades, not a promise about the next few. A method that wins 40% of the time loses 60% of the time, and losses that common will bunch together by chance.

The drawing at the top shows one simulated run of 100 trades from such a method, with winners of +2R and losers of −1R, where R is the amount risked per trade. Its expectancy is 0.4 × 2 − 0.6 × 1 = +0.2R, and this run ends at +20R, exactly on target. Along the way it loses 10 trades in a row, trades 51 to 60, and falls 18R from its peak. Nothing was broken; that is what a 40% hit rate looks like.

Losing streak
A run of consecutive losing trades. Its length depends mostly on the loss rate and the number of trades, and long streaks are normal even for a method with a positive expectancy.

What is the probability of N losses in a row?

If each trade loses with probability q (one minus your win rate, or hit rate), and trades are independent, the chance that the next N trades all lose is q to the power N. With a 50% hit rate, q = 0.5; with a 40% hit rate, q = 0.6:

  • Hit rate 50%: 3 losses in a row 12.5%, 5 in a row 3.1%, 8 in a row 0.39%, 10 in a row 0.10%.
  • Hit rate 40%: 3 in a row 21.6%, 5 in a row 7.8%, 8 in a row 1.7%, 10 in a row 0.60%.
  • Somewhere in 100 trades at a 40% hit rate: 5 or more in a row 97.6%, 8 or more 49.0%, 10 or more 20.5%.
  • Somewhere in 250 trades at a 40% hit rate: 10 or more in a row 45.2%, 15 or more 4.4%.

How long a losing streak should you expect?

The first two lines answer the question “will the next N trades all lose?”, which is rare. The question that matters is the last two: “will a streak of N happen somewhere in my next 100 or 250 trades?”, which is common. The table below gives the expected longest losing streak, computed exactly for independent trades, and the length that the longest streak stays at or below in 95% of runs.

Two things stand out. The win rate matters far more than the number of trades: going from 100 to 250 trades adds only 1.0 to 2.5 losses to the expected longest streak, while going from a 50% to a 40% hit rate adds about two. And a trend follower or breakout trader with a 35% hit rate should expect a longest streak of about 11 losses over 250 trades, and up to 17 in 95% of runs. Real trades are not perfectly independent: losses cluster when the market changes character, so treat these numbers as a floor, not a ceiling.

Turn your phone sideways to read the image. Tap to close.

A hand-drawn table of the longest losing streak to expect, computed exactly for independent trades. Rows are win rates of 30% to 60%. Expected longest streak over 100 trades: 10.7, 9.1, 7.9, 6.9, 6.0, 5.3 and 4.6. Over 250 trades: 13.2, 11.2, 9.7, 8.4, 7.3, 6.4 and 5.6. In 95% of 250-trade runs, the longest streak stays at or below 20, 17, 14, 12, 11, 9 and 8. Red squares draw the 100-trade streak, about one square per loss, and the 40% row is highlighted.

Why a long streak is normal even with a positive expectancy

Expectancy describes where a large sample ends up, not the path. The simulated run at the top earns exactly its +0.2R per trade, yet a trader living through trades 46 to 73 would see 18R disappear: 10 straight losses, one win, four more losses, one win, then seven more losses. Judged on those weeks alone, the method looks broken. Judged on 100 trades, it did its job. The maximum drawdown lesson shows how to measure that fall from a peak.

The streak itself is rarely what ruins a trader; the reaction is. In a study of local traders in Treasury bond futures at the Chicago Board of Trade, Joshua Coval and Tyler Shumway found that traders regularly took above-average risk in the afternoon to recover morning losses (Journal of Finance, 2005). For a retail trader it is a bigger size to “get it back”, or a stop moved once too often, and it turns a normal streak into a real hole. The pull to believe a win is now due is the gambler’s fallacy, covered in cognitive biases in trading.

What does a 10-trade losing streak do to your account?

With a fixed share r of the current balance at risk, N straight losses leave (1 − r) to the power N of the account. Hypothetical: a $10,000 account takes 10 losses in a row, as the 40% method did above.

  • At 1% risk per trade: 0.99 to the power 10 leaves $9,044, a drawdown of 9.6%, and the account needs a gain of 10.6% to get back.
  • At 3% risk per trade: 0.97 to the power 10 leaves $7,374, a drawdown of 26.3%, and the account needs a gain of 35.6% to get back.
  • At the 95% streak for 250 trades at a 40% hit rate, 14 losses: −13.1% at 1% risk and −34.7% at 3% risk.
  • Over the whole simulated run at the top, the deepest drawdown is 16.6% at 1% risk and 42.7% at 3% risk.

Turn your phone sideways to read the image. Tap to close.

Two staircases of a hypothetical $10,000 account through 10 losses in a row. The 1% staircase steps down gently to $9,044, a 9.6% drawdown that needs +10.6% to recover. The 3% staircase steps down steeply to $7,374, a 26.3% drawdown that needs +35.6% to recover. Labels mark both end values; a note says three times the risk makes the hole almost three times deeper and the climb back more than three times higher.

How to size your risk from the streak math

Bigger risk also grows faster when a run ends well, which is exactly why it tempts: the 3% account in the simulation finished higher. The real question is whether you would still follow the plan at −43%. Most traders would not, so size for the streak first and growth second:

  1. Measure your win rate on at least 100 logged trades of one setup, or from a backtest that you discount for costs.
  2. Read the 95% streak for your next 250 trades from the table: 14 at a 40% hit rate.
  3. Choose the drawdown you would accept from a streak alone, for example 15%.
  4. Risk per trade = 1 − (1 − 0.15) to the power 1/14, about 1.15%; round down to 1%.
  5. Check it against the Kelly criterion: for a 40% hit rate and 2R winners, full Kelly is 10% per trade, so 1% is a tenth of it, which leaves room for an edge you have overestimated.
  6. Run your numbers through the risk of ruin calculator on this page to see how deep a bad run could go.
Risk of ruin
The probability that a run of losses takes an account down to a level you define as ruin, such as a 50% drawdown from its peak, given a win rate, a payoff and a risk per trade.

What to do during a losing streak: pause, then review

Decide the pause rule before the streak, because during one every rule feels negotiable. A common set, with hypothetical numbers: stop for the day after three losses or −3R; cut the risk per trade in half after a 10% drawdown from the peak; return to full size only after recovering half of it. Risk management explains daily and weekly loss limits in more detail, and trading psychology the revenge-trading loop a streak can start.

Then review without rewriting the system after every loss. For each losing trade, ask one question in your trading journal: did I follow the plan? If yes, the loss is variance and the plan stays. If no, the fix is the rule you broke, not the method. Change the method itself only at a scheduled review after a full sample, or if a streak runs longer than the 99% value for your win rate, 17 at 40% over 250 trades, which is a reason to retest the edge on paper before trading it again. If a streak starts to cost you sleep, see trading anxiety and burnout.