Does a moving-average crossover beat buying and holding SPY? 26 years of daily data, with trading costs
Site edge: one trading claim tested on data with costs, shown whether it held. No advice, no signals. What we tested: instrument SPY (dividend-adjusted daily closes, Yahoo Finance chart data, 3 January 2000 to 20 November 2026, 6,763 days); rule: hold SPY when the fast simple moving average of closes is above the slow one, otherwise hold cash earning nothing; positions take the next day's return; costs of 0, 5 and 10 basis points per switch; two rule pairs tried (50/200 and 10/50), both reported. Who this applies to: this one instrument and window. Not financial advice.
What we found
- Buying and holding SPY grew 1 dollar to about 8.13 dollars over the 50/200 test window, about 8.38% a year.
- The 50/200 crossover switched 27 times and was invested 73.2% of the days; with no costs it grew 1 dollar to 7.90 dollars, about 8.26% a year.
- With 10 basis points per switch, the 50/200 crossover reached 7.69 dollars, about 8.15% a year: below buy and hold, though with less time in the market.
- The faster 10/50 crossover switched 174 times; with no costs it reached 3.45 dollars, about 4.76% a year, against 8.07% a year for buy and hold over its window.
- With 10 basis points per switch, the 10/50 crossover reached 2.90 dollars, about 4.08% a year.
- The claim that a crossover beats buy and hold on SPY did not hold in this window for either pair, before or after costs.
How the numbers work
| Figure | Value | Where it comes from |
|---|---|---|
| Buy and hold, 50/200 window | 8.13x, 8.38% a year | calculated |
| 50/200, no cost | 7.90x, 8.26% a year, 27 switches | calculated |
| 50/200, 10 bp | 7.69x, 8.15% a year | calculated |
| 10/50, no cost | 3.45x, 4.76% a year, 174 switches | calculated |
| 10/50, 10 bp | 2.90x, 4.08% a year | calculated |
| Buy and hold, 10/50 window | 8.07% a year | calculated |
What we are unsure of
- Cash earns nothing in this test; with Treasury bill interest on the days out of the market, the crossover would look better.
- The crossover is often argued for its smaller drawdowns, which this test did not measure.
- Two rule pairs were tried; trying many and reporting the best would overstate any edge.
Data
- SPY daily chart data, Yahoo Finance, downloaded 2026-11-21.