The second entry is the first one with a rule in it. The rule is mostly about when not to be in the market.
The idea
Own the Dow through a single fund, DIA. Watch one number: how violently it’s moving. When that volatility spikes far above normal, sell everything and wait in cash. When it calms down, buy back in.
The reasoning behind it: most of the money in stocks is made in slow, steady climbs, and very little is made in the panics. So the job isn’t to predict the market. It’s to recognize when the market has stopped behaving normally and step aside until it starts again.
What it did
| Vol Filtering | Holding the Dow (DIA) | |
|---|---|---|
| CAGR | 13.4% | 13.1% |
| Total return | +289% | +275% |
| Worst drop | 20.8% | 35.2% |
| Sharpe ratio | 0.94 | 0.82 |
The return is slightly better than just holding the Dow. The bigger difference is the ride. The volatility exit triggered only three times in almost eleven years: March 2020, June 2020, and April 2025. The first one did most of the work. It was in cash from early March to early May 2020, so it missed most of the COVID crash, the moment when holding the index meant losing more than a third of its value. The worst stretch it did live through was the slow 2022 decline, about 21% from peak to trough. That’s the kind of grinding drop a volatility rule isn’t built to catch.
What didn’t work, and why it’s one fund
Before landing here, I tested a lot of ways to be cleverer than owning the whole Dow: picking only the top 10 Dow stocks by recent performance, by how steadily they’d been trending, by both, with short and long lookbacks. None of them beat owning all 30 with the same risk rule. Most did no better than picking at random. The stock picking added effort and trading costs and nothing else.
Owning all 30 stocks in equal amounts, rebalanced regularly, with the same volatility exit, came out almost identical: 13.2% a year with a 21.7% worst drop. Since the results are the same, the simpler instrument wins. One fund means one trade each time the rule fires, instead of thirty stocks to buy, sell and keep balanced. An equal-weight Dow fund was also tested; it did slightly worse and trades very thinly.
The test: passed for its purpose
Vol Filtering isn’t trying to beat the market. Its job is narrower: keep roughly the market’s return while cutting the worst drop. It trades only a handful of times a decade, so it can never reach the 100 trades the full bar on the bench requires. It’s judged on that goal instead, over the most recent 10 years of data, with cash earning the actual 3-month Treasury bill rate.
| Oct 2016 – Oct 2026 | Vol Filtering | Holding the Dow (DIA) |
|---|---|---|
| CAGR | 13.7% | 13.2% |
| Worst drop | 20.8% | 35.2% |
| Return for risk (Sharpe over T-bills) | 0.78 | 0.68 |
It met the goal: a worst drop about 40% smaller, with a slightly higher return. Its timing also beat 96.5% of random versions that sat in cash just as often, and it held up at much higher trading costs. On that basis it’s marked Passed.