Does the 200-day rule work?

The rule

Own stocks when the market is above its average price of the last 200 trading days. Sell to cash when it drops below. Check once a month. That is the whole thing, and its appeal is obvious: it would have had you out of stocks for most of 1929 to 1932, 2000 to 2002 and 2008.

What it buys

A smaller worst fall. Run on US stocks between July 1926 and June 2026 with a realistic trading cost, the rule's deepest loss was −43.1% against 84% for holding. That is the one result in this whole study that is not in doubt, and it holds because the arithmetic is mechanical: you cannot lose 84% if you are in cash for most of the fall.

Meb Faber, who popularised the rule, tested it across several asset classes on data through 2005. Concretum ran the same rule forward on the years after his paper, 2006 to March 2025, and found the worst fall stayed small at 12% while the return dropped from 11.7% a year in his sample to 6.1% after it. The protection survives. The return edge does not.

What it costs

A lower return than holding. Over the full century the rule earned 0.73 points a year less than buy and hold. Since 2017, run on SPY, it earned 8.5% a year against 15.0% for holding, and in nine of ten round trips it sold and then bought back at a higher price. Across the century, 58 of 73 of its exits lost money. The worst run was twelve losing exits in a row between 1948 and 1960, and the rule has not made a new high against buy and hold since June 1932.

Then tax. Every exit in a taxable account is a sale. In a top bracket that cost 1.92 points a year on top of the return gap, which puts the rule about 3 points a year behind holding. Sheltered, the gap is about 1 point.

One more comparison. The rule is out of the market about a quarter of the time, so on average it holds 73% stocks. Against a portfolio that simply holds 73% stocks and 27% cash all the time, the rule came out +0.74 points a year ahead with a range so wide the test could not tell whether the timing added anything. Holding fewer stocks buys a worst fall of 39% for free, with no trading and no tax.

The fair case for it

The strongest case I can make for the rule is this. I will probably earn less than holding. I will trade about once a year and be wrong most of those times. I should only do it in an account where tax cannot reach me. What I am buying is a smoother ride and a rule that stops me from panicking at the bottom, which is worth something to a person who knows they would otherwise sell everything in March 2009. If that is you, the rule is a fair purchase, and you should know you are buying a smaller fall, not a bigger return.

2x and 3x funds

A 3x fund borrows to triple the market's daily move. Held over the century since 1926, it would have compounded at 12.5% a year against 10.2% for the plain index, lost 99.9% at its worst, and spent 28 years under water.

The popular fix is to hold the 3x fund only when the market is above its 200-day average, the idea in Michael Gayed's paper on holding borrowed stock exposure only above the 200-day average. Over the same century that turned 19.9% a year with a worst fall of 85%, in a run of whipsaws between 1933 and 1935 where the rule kept getting in and out at 3x. October 1987 cost it 67% because a three-week crash cannot be exited by a 200-day average. In 2022 it lost 35% while the index lost 25%. Once you compare the timed 3x fund with the same 3x fund held at the 73% the rule averages, all the time, the timing itself is worth somewhere between minus 1 and plus 11 points a year, and the test cannot say which. The return is the borrowing. The rule just decides when you are exposed to it.

The other popular version is HFEA, 55% of a 3x stock fund and 45% of a 3x long Treasury fund, rebalanced. Since 1926 it returned 16% a year with a worst fall of 93%. Between December 1972 and September 1981 it lost 52% while the plain index gained 60%, because bonds and stocks fell together for nine years. It lost 47% in 2022 for the same reason. It is a bet that bonds will rise when stocks fall, and that bet has not paid since 2020.

In a taxable account the 3x rule costs about 3.3 points a year in tax. The only version I could defend is a small sheltered 2x holding under the rule, at most 10 to 15% of a portfolio, for someone who has decided they want borrowed exposure to stocks and wants a brake for slow crashes. That is an argument about what you can hold through, not what you will earn. The default is none.