The method was born at Los Alamos in 1946 when Stanislaw Ulam, recovering from illness, wondered whether solitaire winning odds could be estimated by dealing hands randomly rather than enumerating combinatorics. John von Neumann recognized its value for neutron diffusion and the classified project was code-named Monte Carlo after the Monaco casino.

Key milestones:

1949: Metropolis & Ulam publish the first formal paper on statistical sampling.
1953: Metropolis algorithm enables sampling from arbitrary distributions — foundation of modern MCMC.
1970s: Boyle, Phelan and others bring MC to option pricing; Black-Scholes gives analytic benchmarks.
1990s–2000s: Cheap compute makes 10,000-path retirement simulations routine for advisors.

Era	Typical path count	Application
1950s	100s	Neutron shielding
1980s	1,000s	Exotic options
2020s	1,000,000s	Retirement, climate, AI