What it means
Option pricing uses a beautiful trick: pretend investors are indifferent to risk, compute the expected payoff in that pretend world, and discount at the risk-free rate. The probabilities in the pretend world are risk-neutral.
They are not forecasts: the risk-neutral probability of a stock rising is usually different from the real-world chance, because it bakes in the risk premium investors actually demand. The justification is arbitrage, not psychology: if a derivative's payoff can be replicated by trading the underlying, its price must equal the replication cost, and the risk-neutral measure is the bookkeeping device that computes it.
The Bank of England's working paper on implied risk-neutral probability densities shows the reverse journey: from a strip of option prices, one can extract the whole probability distribution the market is pricing. That extraction is a policy instrument: central banks read risk-neutral densities from index options to see the tail risks markets are paying to insure, a sentiment gauge no survey can match.
The maths has a strict pedigree: under no-arbitrage, a risk-neutral measure exists, and under market completeness it is unique, which is why the technique is exact in textbook models and approximate in life. Practitioners hold the distinction like a talisman: pricing and hedging live in the risk-neutral world, while risk management and capital live in the real-world one, and confusing the two misprices both.
For a non-finance reader, risk-neutral probabilities are the exchange rate between risk and price: not the odds of what will happen, but the odds consistent with what protection costs. The binomial intuition generalises: every pricing model, from closed forms to Monte Carlo engines, simulates the risk-neutral world rather than the one investors actually inhabit.
Calibration connects theory to the desk: model parameters are fitted to observed option prices, so the risk-neutral measure in use is the one the market is currently trading, not the one a textbook assumes. Change of measure is the deep idea: adjusting probabilities is mathematically equivalent to adjusting payoffs for risk, and the measure chosen is the one where the adjustment disappears.
In practice
Real-world examples.
Example
A binomial model computes the risk-neutral up probability from the risk-free rate and the move sizes, then prices the option by replication. A trader checks the answer by building the replicating portfolio and finding that it costs the same.
Example
A central bank extracts the market's implied crash probability from an index options chain before an election. The surveys and the prices disagreed, and the markets team explains that the options figure includes the premium investors pay for protection.
Example
A risk manager keeps real-world probabilities for loss estimation and risk-neutral ones strictly for pricing. Her value-at-risk report uses historical scenarios, while the pricing library for the bank's derivatives uses the risk-neutral measure.
Formula
Calculation
Derivative price equals the discounted expected payoff under risk-neutral probabilities. For a one-step binomial move with up factor u and down factor d, the risk-neutral up probability is q = (e^{rT} minus d) / (u minus d) with continuous compounding, or q = (1 + r minus d) / (u minus d) with simple annual compounding, computed from the risk-free rate and the move sizes.
Worked example, using simple annual compounding for easy arithmetic. A fictional stock trades at $100 and in one year will be either $120 (u = 1.2) or $80 (d = 0.8). The risk-free rate is 5%, so q = (1 + 0.05 - 0.8) / (1.2 - 0.8) = 0.25 / 0.40 = 0.625, and the down probability is 1 - 0.625 = 0.375.
Now price a one-year call option with a $100 strike. Its payoff is $20 if the stock reaches $120 and $0 if it falls to $80. The price is the discounted risk-neutral expectation: (0.625 x $20 + 0.375 x $0) / 1.05 = $12.50 / 1.05 = $11.90.
Check by replication. Buying 0.5 shares costs $50 and is worth $60 or $40 next year. Borrowing $38.10 today means repaying $40 at 5%, so the position pays $20 or $0 exactly like the option. Its cost is $50 - $38.10 = $11.90, which matches. The real-world chance of the stock rising never entered the calculation.Case study
Seen in the real world.
This case study is fictional and illustrative. A made-up central bank's markets team watches index options ahead of an election. The risk-neutral density extracted from the options chain shows the market pricing a 9 percent chance of a 20 percent one-month drop, triple the historical frequency, while surveys of economists put the probability near zero. The governor's briefing frames the two numbers correctly: neither is a forecast, but the risk-neutral figure reveals what institutions are paying to insure, and that insurance demand is itself information about positioning and fear.
After the election passes quietly, the tail price collapses, and the team writes its internal postscript: the options were expensive insurance that went unused, exactly as most insurance should. The desk's junior quant learns the accompanying discipline the same week: her risk report uses real-world probabilities for expected losses and the risk-neutral ones only inside the pricing library, because a predecessor once mixed the worlds and produced a value-at-risk figure that was neither one thing nor the other. The two probability measures, she is told, are colleagues who must never be allowed to do each other's job.
Watch out
Common mistakes.
- Reading risk-neutral probabilities as forecasts; they encode prices and risk premia, and usually overstate the real-world odds of the feared outcomes.
- Using them for risk management; expected losses belong to real-world probabilities, and mixing the two corrupts both reports.
- Assuming uniqueness outside textbooks; incomplete markets admit many risk-neutral measures, and the model's choice is part of the answer.
Questions
People also ask.
What are risk-neutral probabilities?
Adjusted probabilities under which assets earn the risk-free rate, used to price derivatives by discounting expected payoffs, justified by arbitrage rather than belief.
How do they differ from real-world odds?
They embed risk premia: feared outcomes carry higher risk-neutral than real-world probability, which is why implied crash odds look pessimistic.
Why extract them from options?
A chain of option prices reveals the whole implied distribution, giving policymakers and traders a market-priced map of tail risks.
From the founder's library
Take it further with the book.
Build your financial confidence beyond this definition. Shihan's full-length guide, Accounting Fundamentals, takes the same plain-English approach and turns it into a complete, practical playbook for non-finance managers, business owners and students - with chapter-end quiz answers and presentation slides included.
25% off with code MMHQ25, applied at checkout. Priced in USD - checkout may show the equivalent in your local currency.
View the book and save 25%