Buy Last Year's Losers
Mean reversionOn the first trading day of each January, buy the 10 S&P 500 stocks with the LOWEST total return during the prior calendar year (that are still in the index), equal weighted. Hold for the year, then rebalance.
See this in the terminal — live earnings bands, fair value and rates.
Start free trialRead the story
Buying the beaten-down losers sounds brave. It was mostly just painful.
The rule. On the first trading day of each January, buy the 10 S&P 500 stocks with the LOWEST total return during the prior calendar year (that are still in the index), equal weighted. Hold for the year, then rebalance.
What actually happened.
- $10,000 became $33,091. The same $10,000 in S&P 500 (SPY) became $86,775. This strategy trailed the benchmark by $53,684.
- Compounded at 8.3% a year vs 15.5% for the benchmark.
- Best year: 2013 (+54%). Worst year: 2024 (-33%). This is not a smooth ride.
The catch. At its worst, this strategy fell -58% from a prior peak — deeper than the benchmark's -34%. Would you have held through that? Most people don't. That's the whole game.
The takeaway. The lesson isn't "this idea is stupid." It's that the simple, boring benchmark is harder to beat than the pitch decks admit. Before you try to be clever, respect the index.
Holding right now (2026-01-02): ARE, CHTR, CMG, DECK, FDS, FISV, IT, LULU, MOH, TTD.
Drawdown from prior peak
Calendar-year returns
| Line | 11* | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26* |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Last year's 10 losers | +11% | +35% | +54% | +25% | -29% | +21% | +15% | +2% | +26% | -4% | +27% | -25% | +53% | -33% | +27% | -23% |
| S&P 500 (SPY) | +10% | +16% | +32% | +13% | +1% | +12% | +22% | -5% | +31% | +18% | +29% | -18% | +26% | +25% | +18% | +14% |
* Partial year — the backtest does not cover this whole calendar year, so it is excluded from the best and worst year figures.
Holding now · 2026-01-02
Includes stocks later dropped from the S&P 500 where price history exists (138 of 345). The rest are missing, which can flatter momentum-style ideas. How we test
Rebalance: annual (first trading day of January), equal weight · Window: 2011-09-28 → 2026-09-28 · Last run Sep 28, 2026 · JSON