Same cadence, same 10 bps/side cost, execution at next open. × multiple = value ÷ dollars contributed. $ value = absolute dollars at $1,000 per buy, log scale. Note how nearly all of ROTATOR's separation from QQQ opens up in the final stretch (2023+).
If you had started this DCA N years ago and run it to today. Caveat: every "to-present" horizon includes the 2023-2026 boom, which flatters all of these — see the rolling-window table below for the unbiased view.
| started | ROTATOR × | ROTATOR IRR | QQQ × | QQQ IRR | SPY × | SPY IRR | ROTATOR vs QQQ |
|---|
Within-year growth of the running portfolio (net of that year's new contributions), vs the same DCA into QQQ and SPY. This is where ROTATOR's variance shows: huge up years, but also flat/down years the cash switch and concentration produce.
| year | ROTATOR | QQQ DCA | SPY DCA |
|---|
Across all rolling start dates (244 windows), how often each holding length actually beat QQQ-DCA. SUMMIT shown for contrast.
| holding length | windows | ROTATOR beat QQQ | ROTATOR worst | SUMMIT beat QQQ | SUMMIT worst |
|---|---|---|---|---|---|
| 3 years | 70 | 51% | −45.3% | 84% | −10.6% |
| 5 years | 62 | 50% | −55.6% | 92% | −7.4% |
| 10 years | 42 | 52% | −57.5% | 100% | +8.4% |
| to present | 70 | 100% | +40.3% | 100% | +45.9% |
ROTATOR's "to-present" row is 100% only because every such window rides 2023-2026. On independent 3/5/10-year holds it is roughly a coin-flip vs QQQ with deep worst cases — the price of three-name concentration plus a cash switch that whipsaws.
| regime | multiple | ROTATOR vs QQQ | SUMMIT vs QQQ |
|---|---|---|---|
| GFC 2007–2009 | 1.32× | +13% | +9% |
| Recovery 2009–2012 | 1.08× | −20% | +7% |
| Bull 2013–2017 | 2.23× | +40% | +27% |
| Sideways 2015–2016 | 1.25× | +14% | +4% |
| Vol shock 2018–19 | 0.95× | −13% | +1% |
| COVID 2020–2021 | 1.33× | −4% | +14% |
| Bear 2022–mid-2023 | 1.19× | −0% | +18% |
| AI bull 2023–2026 | 4.94× | +206% | +48% |
ROTATOR's cash switch whipsaws in choppy recoveries (2009-12, 2018, COVID rebound) — it sells into the bottom and re-enters higher. SUMMIT stays positive in every regime by switching what it buys rather than going to cash.
Both run on the identical engine, S&P 500 point-in-time universe, next-open fills, and 244-window grid. ROTATOR at its native k=3 / 10 bps; SUMMIT at k=2 / 5 bps.
| metric | SUMMIT | ROTATOR | edge |
|---|---|---|---|
| Beat QQQ (244 windows) | 93% | 65% | SUMMIT |
| Beat SPY | 98% | 77% | SUMMIT |
| Median excess vs QQQ | +28.8% | +21.3% | SUMMIT |
| 10th-pct window vs QQQ | +3.0% | −23.4% | SUMMIT |
| Worst window vs QQQ | −10.6% | −57.5% | SUMMIT |
| Full-period multiple (2006→) | 20.0× | 30.0× | ROTATOR |
| Best regime (AI bull) | +48% | +206% | ROTATOR |
| Trades / year | ~0 (never sells) | ~26 | SUMMIT |
| Cost drag 5→10 bps (full mult) | none | 32.8×→30.0× | SUMMIT |
| Tax profile | deferred / LT | mostly short-term | SUMMIT |
In one line: ROTATOR has the higher ceiling, SUMMIT the higher floor. ROTATOR wins terminal wealth and explosive regimes; SUMMIT wins consistency, drawdown, turnover and taxes. For buy-and-hold money you can't perfectly time the entry of, SUMMIT's 93%/−10.6% profile is the safer expectation; ROTATOR is a higher-variance satellite whose recent dominance leans heavily on one concentrated theme.
Rank by average(3-month, 6-month return) × fraction of rising weeks over 6 months. "Strong AND steady beats strong alone."
Exclude any name whose biggest 1-day jump in the past month is in the universe's top 5% — "names that just went vertical tend to give it back."
Hold the top 3, equal weight. Keep a position while it stays in the top 8; sell once it drops below rank 8. Buy at next open.
If SPY closes below its 210-day average, sell everything to cash; re-enter at the first rebalance after SPY reclaims it.