Play · timing vs. time in the market
The Long Game
Updated 2026-07-30 · Sources: 2 · Real S&P 500 total returns, 1928 to 2025
Twenty years, one click each. You start with $10,000 in a broad index fund, and before every year you decide one thing: stay in, or jump to cash. Beside you a tortoise does nothing at all; it just holds. The years are not invented: they are a real, randomly chosen run of S&P 500 history, revealed one at a time, exactly as you would have met them. At the end the tool adds up what each of your exits actually cost or saved.
Predict first
Playing timing against buy-and-hold for twenty years, most players end up:
Pick one, or skip the prediction.
Choose a position, then press “Next year” to reveal the first year's return.
n/a
What each exit did
| Exit | Years out | Index moved | Had you stayed in | Verdict |
|---|
Year by year
| Year | You were | Index return | Your pile | Tortoise |
|---|---|---|---|---|
| No years played yet. | ||||
Real data, teaching model. Each year's return is the S&P 500 annual total return (dividends reinvested), nominal, from Aswath Damodaran's “Historical Returns on Stocks, Bonds and Bills” dataset (NYU Stern), covering 1928 to 2025 and fetched 2026-07-30. Cross-checked against the S&P 500 total-return column on Wikipedia's S&P 500 article (sourced from S&P Dow Jones Indices): 56 overlapping years, mean absolute difference 0.23 percentage points, largest single-year gap 1.13 points; the two series tell the same story at the scale this game teaches at. Cash is modelled at a flat 0%, which is generous to the timer in high-rate decades and harsh in low-rate ones; there are no taxes, no trading costs, no bid-ask spread and no capital-gains bill on the way out, all of which make real-world timing worse than it looks here. Nominal returns, so inflation is not removed. Teaching only, not advice, and not a forecast.
He that observeth the wind shall not sow; and he that regardeth the clouds shall not reap.Ecclesiastes 11:4
KJV. The farmer scanning the sky for perfect conditions looks careful and ends the season with nothing in the ground. Waiting is not a neutral act; it is a decision with a price, and this game charges it. The next verses answer the objection: you cannot know which venture will prosper, so spread your seed and act (Ecclesiastes 11:2, 6).
What the whole dataset says
There are 79 different twenty-year windows between 1928 and 2025. Run the simplest timing rule a frightened person actually follows (after a losing year, sit out the next one; come back after a winning year) through every one of them, and it beats buy-and-hold in 1 of 79 (the 1983-2002 window). The median result is about 60% of the tortoise's pile, a shortfall of roughly 2.7 percentage points a year, from a median of just 8 trades in twenty years.
Meanwhile the tortoise had a dull time of it and did fine: across those 79 windows the worst twenty-year stretch still compounded at 2.4% a year, the median at 10.9%, the best at 17.7%, and the number of those windows that ended below where they started is 0, even though 26 of the 98 individual years were losses.
Here is the honest complication, because the argument is weaker without it: timing works beautifully backwards. Dodging a window's single worst year saves more than missing its single best year costs in 54 of 79 windows. The problem was never that avoiding crashes wouldn't help. The problem is that the choice has to be made in advance, in the dark, and the rule above is what that choice actually looks like when fear is the one making it. That is the gap Morningstar measures in real accounts; see behavior and speculation.