The math nobody explains about drawdowns — why a 20% loss needs 25% back, and why that gap is what actually kills accounts
Most people know "losses and gains aren't symmetric" as a vague warning. Here's the actual shape of it, and why it matters more than people think.
**The asymmetry**
Lose 10%, you need +11.1% to get back to even. Lose 20%, you need +25%. Lose 50%, you need to *double your money*. Lose 75%, you need +300%. Lose 90%, you need +900%.
The formula is simple: gain needed = loss / (1 − loss).
g = L / (1 - L)
It's not linear; it's not even close. Past about 50% down, the math turns from "annoying" into "structurally almost impossible" for most strategies.
**The part that actually gets people: time**
The percentage is only half the story. The other half is how many trades it takes to earn that percentage back, and how long that takes in calendar time.
Take a strategy with a 50% win rate and 1:2 risk-to-reward (genuinely good stats). One trader takes a 10% hit trading once a month — that's \~22 trades, or roughly *2 years*, to recover. Another trader with the same edge but trading 3x a week takes a 20% hit — 4 months to recover, versus 4 weeks if they'd kept the loss to 5%.
Same edge. Same win rate. The difference between a 4-week recovery and a 4-month one is entirely position sizing and discipline at the moment of the loss — not skill, not the strategy.
That gap is where most people actually blow up. Not from one bad trade, but from refusing to sit through the boring recovery and instead swinging bigger to "get it back faster" — which is how a manageable drawdown becomes an account-ending one.
**What actually prevents this (none of it requires being right more often):**
* **Risk per trade capped at 1-2%.** Ten losses in a row at 2% risk = 18% drawdown (recoverable). Same streak at 10% risk = 65% drawdown (probably not).
* **Set the stop-loss before entry; never widen it after.** Moving a stop mid-trade is how a 2% loss becomes a 10% one.
* **Diversification** across uncorrelated instruments or strategies, so one position can't do structural damage to the account.
* **A hard drawdown cutoff** — a predetermined point where you cut size or stop entirely, not a decision made in the moment while already down.
None of these makes you a better predictor of the market. They just cap how bad any single mistake is allowed to get—which lets the edge you already have keep compounding instead of getting wiped out.
Worth sitting with: research on investor behaviour shows people sell winners about 60% more often than losers — the exact opposite of what protects an account. The instinct to "wait until I'm even" before cutting a loser is normal, and it's also usually what turns a small loss into a large one.
I wrote this up in more depth (with the full break-even tables across different win rates and reward-to-risk ratios) [here](https://quantitativeteasing.substack.com/p/in-trading-a-thousand-cuts-keep-you)
