How Long Is a Normal Losing Streak?

The longest losing run across N signals is approximately ln(N × p) ÷ ln(1 ÷ (1 − p)), where p is the win rate. At a 40% win rate that is about 7 consecutive losses in 100 signals and about 12 in 1,000 — for a strategy that is profitable, since break-even at 2:1 reward-to-risk is 33%. A losing streak is a property of the win rate, not evidence the edge has gone; distinguishing a normal run from a broken strategy takes hundreds of resolved outcomes. What determines survival is position size: a run of twelve costs about 20% of the account at 2% risk per signal, but about 70% at 10%.

Almost everyone who abandons a systematic strategy does it during a losing streak, and almost everyone who does assumes the streak means the strategy stopped working. Usually it does not. Long losing runs are not evidence of failure — they are an arithmetic certainty, and you can work out in advance roughly how long yours will be.

How long is normal

For a strategy with win rate p, the longest losing run you should expect across N signals is approximately ln(N × p) ÷ ln(1 ÷ (1 − p)). That gives:

Read the 40% row carefully, because at a fixed 2:1 reward-to-risk a 40% win rate is a genuinely good result — break-even is 33%. A profitable strategy at that hit rate will still hand you a run of roughly a dozen consecutive losses inside a thousand signals. Not as a bad year. As the expected case.

These are approximations, and the run you actually get can be longer. The point is the order of magnitude: if you were braced for four losses in a row and you get eleven, nothing has gone wrong that the arithmetic did not already predict.

What a streak does and does not tell you

• It does not tell you the edge is gone. Distinguishing a normal run from a broken strategy needs far more data than the run itself contains — typically hundreds of resolved outcomes, not the ten that just went against you.
• It does tell you whether your position size was honest. A streak that threatens the account was always going to happen; it was the sizing that made it fatal. The position size calculator works backwards from the run you can survive.
• It tells you nothing you could not have computed beforehand. Which is the entire argument for computing it beforehand.

The number that actually matters

Streak length is only half the question. The half that decides whether you are still trading afterwards is what each loss costs. Risking 2% per signal, a run of twelve takes roughly a fifth of the account and is survivable. Risking 10%, the same run — the ordinary, expected run — takes about seventy percent, and recovering from that needs a gain of more than 200%.

That asymmetry is the whole of risk management. Work through your own numbers with the risk-of-ruin calculator and the losing-streak calculator, and see why a 40% win rate can beat a 90% one for how win rate and reward-to-risk combine.

Our own streaks

Every signal we publish is timestamped at publication and scored later against the real price series, losses included. That means our losing runs are visible rather than described — you can read them in sequence, per asset, on the track record, and see exactly how each outcome was decided in the methodology.

A provider who publishes only their wins can tell you their longest losing streak is short. They can tell you that because nothing in what they publish would contradict it. The checklist for spotting that is in how to verify a signal provider's win rate.

Educational content, not financial advice. Trading carries risk, including loss of capital.