METHODOLOGY

How the simulation works, what it can tell you, and — more importantly — what it can't.

The rules (methodology v1)

Every monkey receives $10,000 and ten randomly selected US stocks, equally weighted at 10% each. Buy and hold — no trading, no rebalancing, no second thoughts. Performance is computed from monthly, split- and dividend-adjusted closing prices over a fixed ten-year window and compared against the S&P 500 including dividends (measured via SPY, the S&P 500 ETF). If a stock's price series ends during the window, the position is converted to cash at its last available price and sits there earning nothing, which is also what the monkey would have done.

The universe — and its bias

Monkeys currently pick from a universe of today's S&P 500 constituents. That matters: companies that went bankrupt or were delisted along the way are not in the draw. This is called survivorship bias, and it flatters the monkeys — a real random picker in the past could have drawn companies that no longer exist. We state this openly instead of pretending otherwise, because half the point of this site is showing how easily backtests flatter their subjects. A survivorship-bias-free universe including delisted stocks is on the roadmap.

The luck score

A monkey's luck score is its percentile among thousands of simulated monkeys run under identical rules — same universe, same window, same weighting. A luck score of 99.7 means: luckier than 99.7% of comparable monkeys. It is an empirical percentile, not a proprietary AI mystery metric.

What this is

An illustration of luck, survivorship bias, selection bias and the narrative fallacy in investing. When enough monkeys pick stocks at random, some of them will look like investment geniuses. The interesting question is how you would tell them apart from actual investment geniuses. (So far, nobody reliably can.)

What this is not

Wall Street Monkeys is an entertainment and educational simulation. Random portfolios are not investment recommendations. Historical performance is simulated and is not a prognosis. No monkeys were consulted on actual asset allocation.

Methodology version: v1 · monthly adjusted closes · benchmark: S&P 500 incl. dividends (via SPY) · universe: current S&P 500 constituents (survivorship bias applies)