Does trading IWM on its 200-day moving average beat buy and hold?
No: from Nov 13, 2006 to Nov 13, 2026, the rule “buy when the close crosses above the 200-day average, sell when it crosses below” returned +151.8% on IWM after costs, while simply holding IWM returned +372.8%. It made 81 trades and was invested 66% of the time, with a worst drop of -30.7% against -58.6% for holding. Costs were modelled at 2 bp of slippage per side and no commission, and cash earned the 3-month Treasury bill rate while the rule was out of the market.
| The rule | Holding IWM | |
|---|---|---|
| Total return after costs | +151.8% | +372.8% |
| Return per year | +4.7% | +8.1% |
| Worst drop from a high | -30.7% | -58.6% |
| Time invested | 66% | 100% |
| Trades | 81 | 1 (buy once) |
| Sharpe ratio over T-bills | 0.28 | 0.37 |
Before you rely on it
- Data: 5,032 daily sessions of IWM from Twelve Data, Nov 13, 2006 to Nov 13, 2026, adjusted for splits and dividends. Computed Nov 14, 2026 with the estockes Reality Check engine.
- The rule is read on each day's close and trades at the next day's open, long only, with the whole account. Holding is a buy on the first day with no costs.
- At 10 bp of slippage per side, the rule returned +121.2%; before any costs it returned +160.1%.
- This is a historical simulation. It describes the past, does not predict the future, and is not advice.
Related questions
Why compare a rule with simply holding?
Holding IWM needs no decisions and no trades. A rule is only worth its effort and risk if it ends with more than that, after costs, over the same days.
Did the rule at least lose less in bad markets?
Yes. Its worst drop from a high was -30.7%, against -58.6% for holding IWM, mostly because it was in cash 34% of the time.
Did it work in the last two years?
From Nov 13, 2024 to Nov 13, 2026 the rule returned +19.0% after costs on 6 trades. Over the whole test, 4 of 5 equal four-year periods made money.
Can I run this test myself?
Yes. The link on this page opens the same rule in the free Reality Check, which tests the last 5 years without an account. Testing 20 years needs Pro.