Monte Carlo estimator
V_0_hat = DF(0,T) (1/N) sum_i Pi_iValue is estimated as the discounted sample mean of risk-neutral payoffs.
Foundations
Monte Carlo Pricing
Price derivatives by simulating many risk-neutral paths and averaging discounted payoffs.
Mental model
Monte Carlo estimates value by averaging discounted payoffs across simulated risk-neutral paths.
Monte Carlo pricing simulates many paths of the underlying under the risk-neutral measure.
For each path it computes the payoff at maturity, discounts it back, and averages across paths.
Explore the mechanism
How does sampling more risk-neutral paths change estimated value and numerical uncertainty?
This visual shows Monte Carlo pricing of a European option: many simulated risk-neutral paths for S_t, with the option value given by the discounted average payoff at maturity.
MC price: 10.7802BS price: 9.4134Paths: 300 · Std error ≈ 0.8313
Numbers
Monte Carlo price ≈ 10.7802
Black–Scholes price ≈ 9.4134
Absolute error ≈ 1.3668 · relative ≈ 14.52%
Sampling standard error ≈ 0.8313 (shrinks as 1/√N)
Interpretation
Each blue line is one simulated risk-neutral path for the underlying. The option payoff depends only on S_T and whether it finishes above or below the strike K; the Monte Carlo price is the discounted average payoff across all paths.
As you increase the number of paths, the Monte Carlo estimate fluctuates less and moves closer to the analytic Black–Scholes price, but convergence is slow: the noise falls only like 1/√N. Monte Carlo is flexible and model-agnostic but computationally expensive, which is why variance reduction and good random generators matter in production systems.
UNDERSTAND
What to notice
The law of large numbers ensures convergence to the arbitrage-free price, but sampling error shrinks only as 1/√N.
The quantitative structure
Notation
N- number of simulated paths
Pi_i- payoff on path i
DF(0,T)- discount factor to maturity
Standard error
SE(V_0_hat) = DF(0,T) s_Pi / sqrt(N)Sampling uncertainty falls only with the square root of the path count.
This model assumes
- Paths are generated under dynamics and a measure consistent with the pricing problem.
- The simulated payoff has finite variance for the usual standard-error estimate.
A small example
A four-path payoff estimate
Four simulated payoffs are 0, 10, 20, and 30. The discount factor is 0.95.
- Average the payoffs: (0 + 10 + 20 + 30) / 4 = 15.
- Discount the mean: 0.95 * 15 = 14.25.
ResultThe four-path price estimate is 14.25; a production estimate also reports sampling error.
Where it appears
- pricing path-dependent derivatives
- valuing portfolios with many stochastic risk factors
- estimating exposure profiles
Where the intuition breaks
What this view leaves out
- Monte Carlo converges slowly at the usual square-root rate and can give precise-looking answers from a misspecified path model.
Common mistakes
“More simulation paths remove model risk.”
Not quite.More paths reduce sampling error; they do not correct wrong dynamics, calibration, payoff logic, or discounting.
Where this concept sits
Follow the strongest prerequisites, applications, extensions, and contrasts.
Connections for Monte Carlo Pricing
Monte Carlo PricingYOU ARE HERE
Built onRisk-Neutral Measure (Intuition)Risk-neutral Monte Carlo estimates discounted expected payoffs under a pricing measure.
UsesOption Payoff Types and Path DependenceMonte Carlo is especially natural when a derivative payoff depends on an entire simulated path.
Supported byGeometric Brownian Motion (GBM)GBM is a canonical path model used to illustrate risk-neutral Monte Carlo pricing.
Validated byMonte Carlo Model ValidationSimulation-based pricers need convergence, confidence-interval and benchmark checks to establish numerical reliability.
Used bySwing Options and Volume FlexibilitySimulation is a common building block for valuing flexible contracts when future decisions depend on uncertain price paths.
Used byXVA (CVA/FVA) MechanicsExposure profiles for XVA are often generated by simulating future market states and portfolio values.
Built on · Risk-Neutral Measure (Intuition)
Risk-neutral Monte Carlo estimates discounted expected payoffs under a pricing measure.
Explore concept →