i ran 300 random three-ticker combinations looking for one that kept drawdown under 30% and a calmar above 2.
ten years, june 2016 through june 2026, forty-six candidates. not one combination cleared both bars.
so i ran the closest one instead. aapl at 2x, nvda, pypl, $100,000 each. ten years later that account holds $9,311,170, and it still lost.
what it lost to
it lost to doing nothing.
buy and hold on the same three names finished at $18,353,570. the rules gave up half the upside.
same stocks, same decade, same starting money. so on the return column alone this is not close.
the ride was the trade-off
the difference shows up in how you got there.
my worst drawdown was 29.8%. buy and hold sat 56.4% underwater at its low point.
half the hole, for half the upside. that's the actual exchange being made here, and it's worth knowing before you pick a side.
65 trades, 26 winners
there were 65 trades over the ten years. 26 of them made money.
that ratio is the engine, not a defect. a few long runs carry the account and every loser gets cut small.
the exits do one more thing. they turn paper into cash. that account closed with $1,947,537 already pulled out and parked in the vault, while the buy-and-hold total is still fully exposed to the market.
one name paid for the decade
nvda did the heavy lifting. its own $100,000, no rebalancing, no money moved between the three, and it ended at $7,217,108.
pypl ran identical rules and finished at $149,807. ten years for 50%.
i had no way to know in 2016 which of the three that would be. that is why the rules hold all three the same way instead of betting on the one that looks obvious in hindsight.
the red years are the entry fee
the decade was not a straight line.
2020 ran +112% through the pandemic. then 2022 took 25% back during the inflation squeeze, and 2026 is down 10% so far.
quit in 2022 and none of the years after it land in your account. that's the part a ten-year number never shows you.
the short version
three tickers, ten years, june 2016 to june 2026. a past simulation, not a forecast.
the rules: $9,311,170, worst drawdown 29.8%, $1,947,537 already banked in cash. buy and hold: $18,353,570, worst drawdown 56.4%, none of it realized.
my three tickers aren't the only three, and one filter run on 300 combinations found nothing that passed. that's the honest result, and it's the reason i'd read the drawdown column before the return column.