Probability ITM: A Model Output, Not a Win Rate
For educational purposes only; not investment advice.
Direct answer
Section titled “Direct answer”Probability ITM is a model estimate that an option will finish with positive intrinsic value at expiration: Sᵀ > K for a Call and Sᵀ < K for a Put. The number depends on the selected distribution, option price or implied volatility, rates, dividends, time, exercise assumptions, and data convention.
It is not a win rate, a probability of profit, a probability of touching the strike, or a forecast calibrated to the investor’s beliefs. Many platforms display a risk-neutral probability inferred from option prices; some use historical or proprietary methods. The label alone does not identify which one.
Why d₂, Delta, and profit probability differ
Section titled “Why d₂, Delta, and profit probability differ”Under a simplified Black-Scholes-Merton model for a European option with continuous dividend yield q:
d₂ = [ln(S/K) + (r − q − 0.5σ²)T] ÷ (σ√T)
risk-neutral Call probability ITM = N(d₂)
risk-neutral Put probability ITM = N(−d₂)
Here S is spot, K strike, r continuously compounded risk-free rate, σ volatility, T years to expiration, and N() the standard normal cumulative distribution. This is a pricing-measure probability, not a claim that investors expect the stock to earn r or that actual returns are lognormal.
For the same European Call, model Delta is e^(−qT)N(d₁), where d₁ = d₂ + σ√T. Delta and N(d₂) can be numerically close for short maturities but answer different questions: Delta is a local price sensitivity; N(d₂) is a terminal event probability under the model. American exercise, discrete dividends, skew, jumps, and vendor methods increase the difference.
ITM also differs from profitable. A long Call bought for premium C has an expiration breakeven near K + C; finishing one cent above K is ITM but still loses most of the premium. A short option can retain premium and show profit without expiring worthless if it is closed earlier or expires only modestly ITM. Transaction costs and assignment further change realized results.
Probability of touching a level before expiration is a path event, while Probability ITM is a terminal event. A stock can cross the strike and finish below it, or never cross until the final observation. One cannot be substituted for the other.
A 25.1% model probability is not a 25.1% chance of profit
Section titled “A 25.1% model probability is not a 25.1% chance of profit”Assume a European Call with:
- spot
S = $100; - strike
K = $105; - annual volatility
σ = 25%; - time
T = 30/365years; - rate
r = 4%and dividend yieldq = 0.
The simplified calculation gives:
d₂ ≈ −0.671
N(d₂) ≈ 25.1%
d₁ = d₂ + σ√T ≈ −0.599, so model Call Delta is approximately N(d₁) ≈ 27.5%. Using 27.5% Delta as exactly 27.5% Probability ITM would already mix two different quantities.
Suppose the Call costs $2.40. Its expiration breakeven is $107.40, not $105. The event Sᵀ > $105 has the modeled 25.1% probability above, while profitability before costs requires Sᵀ > $107.40. That second probability is lower under the same assumptions. A platform’s probability of profit may also model early closing, Bid/Ask, or a proprietary target, so its definition must be read.
Change volatility, rate, dividend, timestamp, or quote input and the output changes. After an event, the entire implied-volatility surface can move; a probability displayed at entry is not fixed until expiration.
Interpretation checklist
Section titled “Interpretation checklist”- Identify whether the platform uses risk-neutral, historical, subjective, or proprietary probability.
- Record option style, settlement, spot or forward, strike, exact expiration, timestamp, rate, dividend, and volatility input.
- Verify whether volatility comes from Bid, Ask, midpoint, last trade, or a fitted surface.
- Distinguish
Probability ITM,Probability OTM,Probability of Profit, andProbability of Touch. - Calculate expiration breakevens from actual net premium and fees; do not use the strike alone.
- Do not treat Delta as an exact probability, especially for long-dated, dividend-paying, skewed, or American-style options.
- Recalculate after price, IV, skew, time, rates, dividends, or event expectations change.
- Compare model probabilities with historical frequencies only on matched definitions and samples.
- Inspect payoff size as well as probability; many small wins can be offset by rare large losses.
- Size from stress loss and liquidity rather than a high displayed probability.
- Model exercise, assignment, settlement, and resulting shares or cash separately.
- Preserve the vendor methodology and timestamp when reviewing a past decision.
Common misconceptions
Section titled “Common misconceptions”“Probability ITM is the trade’s win rate.” ITM is defined at the strike; profit depends on premium, exit timing, costs, and the complete strategy.
“A 30 Delta option has exactly a 30% chance of expiring ITM.” Delta is sensitivity and only approximates a particular model probability under limited conditions.
“Risk-neutral probability is the market’s literal forecast.” It is a pricing-measure output embedded with risk premia and model assumptions, not necessarily a real-world frequency forecast.
“High probability means favorable expected value.” Expected value also depends on the size and timing of gains and losses.
“The displayed percentage stays valid until expiration.” Every input and the fitted volatility surface can change continuously.