The 200-day moving average strategy: rules, evidence, and a backtest

The 200-day moving average strategy is the simplest trend-following rule in common use. When an asset's price is above the average of its last 200 daily closes, hold it. When it is below, hold cash. Checked once a month, the rule has historically kept an investor out of the deepest parts of major bear markets at the cost of some false exits during bull markets. The rule is one sentence long, and that sentence is the prompt at the bottom of this page.

What are the exact rules?

The version described here uses two funds: SPY for US large-cap equities and BIL for one-to-three-month Treasury bills.

1. On the last trading day of each month, compute the simple average of SPY's closing prices over the trailing 200 trading days.
2. If SPY's closing price is above that average, hold SPY for the next month.
3. If it is below, hold BIL for the next month.
4. Trade only when the signal changes from the previous month.

The rule is binary. The portfolio is either 100 percent in equities or 100 percent in cash. There is no partial position, no volatility scaling, and no judgment.

Mebane Faber's 2007 paper A Quantitative Approach to Tactical Asset Allocation is the most cited source for this rule. Faber used a 10-month simple moving average, checked at month end, applied separately to five asset classes. Ten months of trading days is close to 200, and the two versions behave almost identically, which is why the terms are used interchangeably.

Why might it work?

The rule works if trends persist, and there is substantial evidence that they do at horizons of several months to a year. Moskowitz, Ooi, and Pedersen's Time Series Momentum documents the pattern across dozens of futures markets; Hurst, Ooi, and Pedersen's A Century of Evidence on Trend-Following Investing extends it back to 1880. The usual explanations involve slow diffusion of information, investors anchoring on old prices, and forced selling that pushes prices past fundamental value and keeps pushing.

A moving average is a crude but robust way to measure whether a trend is up or down. It does not forecast; it describes where the price is relative to where it has recently been. The rule's value comes from a specific feature of equity markets: the worst returns and the highest volatility have historically clustered in periods when prices were already below their long-term average. By holding cash in those periods, the strategy avoids the parts of the distribution that do the most damage to compounding.

The mechanism cuts both ways. Bull markets have corrections, and a correction that pushes the price below the average and then reverses produces a sell at the low and a buy back higher. Over a long bull market those false signals accumulate into a meaningful gap versus buy and hold.

When does it fail?

The 200-day rule's failure mode is whipsaw: a signal that flips, costs a round trip, and flips back.

Fast V-shaped corrections. The most damaging environment is a sharp decline that reverses within a month or two. The rule sells after the decline has already happened and repurchases after much of the recovery. 2020 is the recent example: the month-end signal turned negative at the end of March, after a 34 percent decline, and turned positive again at the end of May, with the market already well off the lows. The strategy avoided none of the crash and missed part of the rebound.

Sideways markets. When the price oscillates around its average for months, as it did through much of 2011, 2015, and 2016, the rule generates several round trips that each cost a little and gain nothing. In a taxable account, the costs include short-term capital gains.

Long bull markets with few sustained declines. From 2009 to 2019, the S&P 500 spent most months above its 200-day average, and each brief excursion below it was a false exit. The rule kept pace with buy and hold in that decade only because the drawdown savings from 2008 carried forward; measured from 2009 alone it lagged.

The rule's strongest historical periods are the extended bear markets of 2000 to 2002 and 2007 to 2009, when it moved to cash early and stayed there for a year or more. Its record therefore depends heavily on whether the sample includes such a period, which is one of the common backtesting mistakes to watch for.

What does a backtest show?

An ENSEMBLE model for this exact rule is being published and its live tear sheet card will appear here. In the meantime, the prompt below builds it in under a minute; the resulting tear sheet shows the equity curve against SPY, the drawdown chart, the holdings over time, and the number of signal changes.

Things to look for on that tear sheet:

• Time in market. The rule has historically been in equities roughly 70 to 75 percent of months. The other quarter is where the drawdown protection and the opportunity cost both live.
• The drawdown chart against SPY. In 2008 the difference is large. In 2020 it is small or negative. In 2022 the rule exited in the first quarter and re-entered in early 2023, which produced a shallower drawdown but not a painless one.
• Signal changes per year. Expect one to three in most years and more in choppy ones. Every change is a round trip; the cost of the strategy is in this number.

For the definitions behind the metrics, see backtest metrics explained.

How to run it and change it

Variations worth testing, each a small edit:

• Lookback. Replace 200 days with 100 or 250. Shorter averages react faster and whipsaw more; the historical results for anything between roughly 6 and 12 months have been similar, which is a good sign that the rule is not overfit to one number.
• Signal frequency. Ask for the signal to be evaluated weekly or daily instead of monthly and compare the trade count. This is the most instructive variation on the page.
• A buffer. Require the price to be at least 2 percent above the average to buy and 2 percent below to sell. Hysteresis reduces whipsaws at the cost of later entries and exits.
• Multiple assets. Faber's original version applied the rule independently to US stocks, foreign stocks, bonds, commodities, and real estate, 20 percent each. Ask for that and you have his Global Tactical Asset Allocation model.
• A filter on a static allocation. "60% SPY and 40% TLT, but move the SPY sleeve to BIL when SPY closes below its 200-day average" combines this rule with the 60/40 portfolio and is one of the most common ways it is used in practice.

Practical notes

The 200-day rule is a risk management tool more than a return tool. Its historical appeal is a smaller maximum drawdown and a smoother path, not a higher compound return. Investors who adopt it hoping to beat the index in bull markets will be disappointed; investors who adopt it to stay invested through bear markets they would otherwise have abandoned may not be.

It is sensitive to execution details that a backtest can hide. Checking the signal at month end versus daily, trading at the close versus the next open, and the exact number of days in the average all change the results at the margin. The prompt above states each of those choices explicitly so the backtest tests the rule you mean.

Finally, the rule is binary, and a binary rule concentrates all the decision risk into a handful of dates per year. Many practitioners soften it by scaling the position rather than switching it, or by combining it with a relative-strength rule like dual momentum so that the portfolio rotates rather than exits. Both are a sentence away.

Backtests are hypothetical, past performance does not guarantee future results, and ENSEMBLE is research software rather than an investment adviser.