Lesson 12 of 20 · 7 min

Delta as a rough probability, and the caveat

Delta is close to the chance of finishing in the money, but it is not that number and the gap is systematic.

There is a widespread shorthand that a 0.30 delta call has about a 30 percent chance of finishing in the money. The shorthand is useful and it is not correct, and knowing exactly how it fails is more valuable than either using it blindly or discarding it.

Recall from lesson 8 that the formula contains two probability-like quantities. Delta is N(d₁). The model's probability of finishing in the money is N(d₂). They are different numbers, and d₂ is always smaller than d₁ by exactly σ√T. So delta always overstates the model's own chance of finishing in the money, and it overstates it more when volatility is high or expiry is distant.

In the base case the gap is small: delta 0.533 against a 0.504 probability, under three points. Stretch the contract to two years at the same volatility and delta is 0.657 while the probability is 0.520 — a fourteen-point gap. The approximation is therefore decent for near-dated contracts and badly wrong for long-dated ones.

There is a second caveat that matters more. Both numbers are risk-neutral probabilities, which are the probabilities implied by no-arbitrage pricing, not forecasts of what will happen. And both are computed under the model's distribution, which lesson 10 already established the market does not believe. So the honest statement is: delta approximates a model quantity that itself approximates something the model gets wrong in the tails.

Delta is a hedge ratio that happens to look like a probability. Using it as a rough probability is fine; relying on it as one is not.

Key terms

N(d₂): The model's probability that the option finishes in the money. Always less than delta for a call.
Risk-neutral probability: A probability consistent with arbitrage-free pricing rather than with real-world expectation. Not a forecast.
Approximation error: Here, delta minus N(d₂), which equals the effect of σ√T. It grows with both volatility and time to expiry.

Worked example

How far apart delta and probability get

Underlying 100, strike 100, rate 4%, volatility 25%. Only the expiry changes.

30 days: delta: 0.533
30 days: N(d₂): 0.504
365 days: delta: 0.612
365 days: N(d₂): 0.514
2 years: delta: 0.657
2 years: N(d₂): 0.520

The shorthand loses about three points on a monthly contract and fourteen on a two-year one. The error is one-directional: delta always reads high.

Check yourself

A two-year at-the-money call has a delta of 0.657. Is its chance of finishing in the money about 66%?

Saved in this browser only. No account, nothing sent anywhere.

Educational only. Nothing in this lesson is investment, legal or tax advice, or a recommendation to enter into any transaction. Figures are illustrative and computed from the inputs stated in the lesson.