Expected Value in Trading: Why Bad Strategies Can Win Short-Term
A winning streak doesn't mean your strategy works. Expected value reveals whether you're actually profitable in the long run — and why the math can fool you over small sample sizes.
The Problem with Short-Term Wins
Seven wins in a row. You feel like you've cracked it. You increase your position size. Then the losses come — and they're bigger than all seven wins combined.
This is one of the most dangerous traps in trading: confusing a lucky streak with a working strategy. The fix isn't better intuition. It's math — specifically, expected value.
What Is Expected Value?
Expected value (EV) is the average outcome of a trade if you ran it an infinite number of times. It combines two things: how often you win, and how much you win or lose when you do.
The formula:
EV = (P_win × Avg_win) − (P_loss × Avg_loss)
Where:
P_win= probability of winning (e.g. 0.60 for 60%)Avg_win= average profit per winning tradeP_loss= probability of losing (1 − P_win)Avg_loss= average loss per losing trade
A positive EV means the strategy earns money over time. A negative EV means it loses — no matter how good it looks on any given week.
The Math Behind Good vs Bad Strategies
Here's where win rate alone breaks down:
| Strategy | Win Rate | Avg Win | Avg Loss | EV per Trade |
|---|---|---|---|---|
| A — "Feels great" | 70% | $50 | $200 | −$25 |
| B — Coin flip | 50% | $110 | $90 | +$10 |
| C — Trend following | 35% | $300 | $80 | +$53 |
Strategy A wins 7 out of 10 trades. It still loses money. Strategy C loses 65% of the time and still outperforms both others by a wide margin.
Key insight: A high win rate paired with large losses is one of the most common ways traders blow up their accounts. Chasing win rate without checking EV is a trap.
Why Bad Strategies Win in the Short Term
With 20 trades, variance rules. A strategy with −$25 EV can still show a profit over 20 trades simply by chance — the same way a biased coin can land heads 14 times out of 20.
The smaller your sample, the wider the range of outcomes. You need hundreds of trades before your results start to reflect the true expected value. This is why a single month of trading tells you almost nothing about whether a strategy actually works.
It's also why profit factor is more useful than win rate: it bakes both the size and frequency of wins and losses into a single number.
How to Calculate Your Strategy's EV
- Collect your trade history — at least 50 closed trades, ideally 200+
- Separate wins from losses
- Calculate average win: total profit from winning trades ÷ number of winning trades
- Calculate average loss: total loss from losing trades ÷ number of losing trades (use absolute value)
- Calculate win rate: winning trades ÷ total trades
- Plug into the formula:
For example, 100 trades: 45 wins averaging $120, 55 losses averaging $80.
EV = (0.45 × $120) − (0.55 × $80)
EV = $54 − $44
EV = +$10 per trade
That's a positive-EV strategy. Over 1,000 trades, you'd expect roughly $10,000 in profit — before fees.
What This Means for Automated Trading
A bot removes the psychological pressure to override a strategy mid-run. But it doesn't protect you from pulling the plug after a bad week.
If your strategy has a verified positive EV over a large backtest, a 5-trade losing streak is noise — not signal. Stopping the bot after losses means locking in the downside without ever capturing the long-run edge.
The correct response to losses is to check whether the strategy parameters still make sense, not to abandon a mathematically sound approach because recent results felt bad.
Beet Robot runs DCA Grid and Trend Following strategies that are designed around positive expected value: small, frequent take-profits and hard position sizing rules that keep losses bounded. The bot never overrides its own logic mid-trade based on emotion. Start automating with a consistent edge →