Does anything beat buy & hold on Russell 2000 (IWM)?
Every setup we tested on Russell 2000 (IWM) — ranked out-of-sample, corrected for multiple testing, and forward-tracked in public from the day this page published. The honest answer is the headline.
No setup beat simply holding once tested honestly. We say so plainly.
Its best setup only beat buy-and-hold in one window — a regime artifact, not a strategy. Buy-and-hold benchmark: +14.8% CAGR over 2.9 years (+29.6% CAGR in the out-of-sample window).
Educational research from historical backtests — not investment advice. Past performance does not predict future results.
Russell 2000: Nothing Beat Buy-and-Hold, and That Is the Honest Answer
Broad, diversified instruments like Russell 2000 are where indicator strategies go to disappoint. We ran 1,511 setups against IWM, and none cleared the bar once scored honestly — on data the strategy never saw. The best of the batch, Order-Flow Reversion on the 1-hour timeframe, posted an out-of-sample Sharpe of 2.99, short of the 4.03 hurdle we require before calling anything real. For an index fund this is the expected result: whatever inefficiency exists in single names tends to average away in the basket, leaving buy-and-hold's +14.8% annualized return as the number nothing here managed to beat.
Read these figures with the selection problem in mind. Test 1,511 indicators, keep the best, and the winner looks impressive by construction — which is exactly why the hurdle exists instead of applause for a lucky draw. Here, only 0.7% of setups outperformed buy-and-hold even in-sample, and the top candidate produced +5.8% annual alpha over 0.9 unseen years, across 81 trades with a 69.1% win rate and a -15.9% drawdown. That pattern reads as noise, not signal. Markets also change, so even a genuine past edge can fade. This page documents what failed — useful to know before assuming something works.
Every figure above is computed from our own backtests — nothing is estimated or invented. Hypothetical results; not investment advice.
The least-bad setups — shown with their failure numbers
Nothing here earned a verdict — these are the best of a losing field, published so you can see exactly how "best" still failed.
Order-Flow Reversion
Mechanical rule (exactly as backtested): Net-liquidity reversion — fades a 2-sigma price stretch only when signed-volume order-flow shows sellers are exhausted (imbalance z-score), exits on the mean. Needs real volume, so it abstains on feeds that don't report it. Signals are evaluated at 1-hour-bar close, the position changes on the NEXT bar, 0.08% cost per side, long/flat only — no leverage, no shorting.
Out-of-sample (last ~30% of the window, never used to pick this setup): Sharpe 2.99 · alpha +5.8% · 29 trades over 0.9 yrs.
Bollinger Mean-Reversion
Mechanical rule (exactly as backtested): Buy the lower band, sell the middle — fade stretches below the 20/2 band. Signals are evaluated at 1-hour-bar close, the position changes on the NEXT bar, 0.08% cost per side, long/flat only — no leverage, no shorting.
Out-of-sample (last ~30% of the window, never used to pick this setup): Sharpe 2.26 · alpha -2.7% · 32 trades over 0.9 yrs.
Fibonacci Bands
Mechanical rule (exactly as backtested): Fibonacci-scaled volatility bands — buy below the lower band, exit at the basis. Signals are evaluated at 1-hour-bar close, the position changes on the NEXT bar, 0.08% cost per side, long/flat only — no leverage, no shorting.
Out-of-sample (last ~30% of the window, never used to pick this setup): Sharpe 2.26 · alpha -2.7% · 32 trades over 0.9 yrs.
Since publication — including if it loses
The forward record is just getting started — the gap between the two is the honest score. Marked to market nightly from real prices, rules frozen at publication, as of 2026-07-02. Currently FLAT.
We tested 1,511 setups (indicator × parameters × timeframe) on Russell 2000 (IWM). Only setups with ≥30 trades qualify (1,341 did). Setups are ranked by out-of-sample Sharpe — the last ~30% of history, which standard-parameter rules never saw during selection. Because picking the best of 1,511 tries mines even the holdout, the VALIDATED verdict additionally requires the top setup’s OOS Sharpe to clear a selection hurdle of 4.03 (√(2 ln N)/√T) AND positive alpha in both windows. Of the eligible setups, 0.7% had positive out-of-sample alpha (median OOS Sharpe 0.12) — the table below is truncated, but this summary covers all of them. Full recipe: methodology · the engine’s contract lives in the repo as STRATEGY_METHODOLOGY.md.
Top 20 of 1,341 eligible setups
Ranked by out-of-sample Sharpe. Full + out-of-sample columns, costs included. Hypothetical.
| # | Setup | TF | Total ret | Sharpe | Max DD | Win | Trades | α vs B&H | OOS Sharpe | OOS α | OOS trades |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Order-Flow Reversion | 1-Hour | +25.2% | 0.76 | -15.9% | 69.1% | 81 | -6.7% | 2.99 | +5.8% | 29 |
| 2 | Bollinger Mean-Reversion | 1-Hour | +20.6% | 0.62 | -13.7% | 67.0% | 94 | -8.1% | 2.26 | -2.7% | 32 |
| 3 | Fibonacci Bands | 1-Hour | +20.6% | 0.62 | -13.7% | 67.0% | 94 | -8.1% | 2.26 | -2.7% | 32 |
| 4 | VWAP Bands | 1-Hour | +21.4% | 0.63 | -14.2% | 61.2% | 85 | -7.8% | 2.25 | -5.7% | 26 |
| 5 | Keltner Mean-Reversion | 1-Hour | +5.1% | 0.21 | -18.4% | 65.3% | 75 | -13.1% | 1.93 | -7.8% | 26 |
| 6 | Holy Grail Confluence | 1-Hour | +23.0% | 0.56 | -22.4% | 71.1% | 45 | -7.3% | 1.89 | -2.0% | 15 |
| 7 | Williams %R | 4-Hour | +15.2% | 0.35 | -23.6% | 74.4% | 43 | -7.2% | 1.81 | -3.4% | 14 |
| 8 | RSI Mean-Reversion | 1-Hour | -2.4% | -0.01 | -22.7% | 61.5% | 39 | -15.6% | 1.68 | -15.4% | 12 |
| 9 | Projection Bands | 4-Hour | +16.9% | 0.4 | -26.3% | 63.3% | 60 | -6.8% | 1.67 | -5.2% | 20 |
| 10 | Murrey Math Lines | 1-Hour | +10.4% | 0.31 | -20.9% | 69.8% | 53 | -11.3% | 1.42 | -11.8% | 18 |
| 11 | CCI | 1-Hour | +24.9% | 0.6 | -18.7% | 66.7% | 96 | -6.8% | 1.36 | -9.0% | 32 |
| 12 | Connors RSI | 4-Hour | +18.7% | 0.44 | -18.1% | 59.5% | 84 | -6.4% | 1.3 | -8.4% | 26 |
| 13 | Projection Bands | 1-Hour | +2.5% | 0.13 | -23.7% | 59.7% | 176 | -13.9% | 1.21 | -14.3% | 55 |
| 14 | SMC: Liquidity Sweep | 1-Hour | +20.1% | 0.45 | -29.5% | 64.5% | 62 | -8.2% | 1.19 | -11.0% | 20 |
| 15 | Connors RSI-2 | 4-Hour | +17.5% | 0.44 | -14.5% | 60.7% | 89 | -6.7% | 1.17 | -11.1% | 28 |
| 16 | McGinley 200 Trend | 1-Hour | +23.8% | 0.53 | -18.8% | 29.6% | 54 | -7.1% | 1.16 | -8.5% | 11 |
| 17 | Williams %R | 1-Hour | +5.9% | 0.21 | -24.3% | 65.1% | 129 | -12.8% | 1.15 | -14.0% | 40 |
| 18 | Stochastic | 1-Hour | +9.2% | 0.28 | -25.5% | 67.0% | 100 | -11.7% | 1.1 | -13.6% | 32 |
| 19 | Twiggs Money Flow | 4-Hour | +8.5% | 0.22 | -24.5% | 38.5% | 109 | -8.8% | 1.05 | -7.1% | 25 |
| 20 | Demand Index | 1-Hour | -10.7% | -0.17 | -32.5% | 52.7% | 184 | -18.6% | 1.04 | -13.8% | 61 |
Hypothetical backtests with 0.08%/side costs. Not investment advice — see the full disclaimer.
These are historical backtests of mechanical rules. They are educational research, not investment advice, not a recommendation, and not tailored to you. Educational information only — not investment advice. Hypothetical backtested results; past performance does not guarantee future results. Trading involves risk of loss.