Monte Carlo Simulation

Many financial models depend on uncertain assumptions, such as sales growth, prices, costs, interest rates or investment returns. Instead of using a single best estimate for each, Monte Carlo simulation assigns each uncertain input a range or probability distribution and repeatedly draws random values from it.

Each run produces one possible outcome. After thousands of runs, the results form a distribution showing how likely different outcomes are.

For example, a retirement planner might simulate 10,000 possible paths of investment returns and inflation and find that a client's savings last at least 30 years in 85% of the simulations. A company might use simulation to estimate the probability that a project's net present value will be negative.

The technique is named after the Monte Carlo casino, because it relies on randomness. Its results are only as good as its inputs: if the assumed distributions or the relationships between variables are wrong, the results can give false confidence.