Faber’s 10-month moving average timing model

May 2007 – Sep 2026	Strategy	SPY
Annual return (CAGR)	9.2%	10.7%
Worst drawdown	-24.9%	-55.2%
Sharpe ratio	0.77	0.62
Volatility	12.5%	19.7%
Annual return since publication (Feb 2013)	9.5%	14.6%

Jan 1960 – Sep 2026	Strategy	SPY
Annual return (CAGR)	10.2%	10.6%
Worst drawdown	-32.9%	-55.2%
Sharpe ratio	0.89	0.70
Volatility	11.7%	16.4%
Annual return since publication (Feb 2013)	9.5%	14.6%

The 10-month moving average timing model comes from Mebane Faber's paper "A Quantitative Approach to Tactical Asset Allocation", first published in 2007 and updated in 2013. It is one of the most cited trend-following rules for individual investors. Hold the stock market while it trades above its average price of the last ten months; otherwise hold Treasury bills.

The idea#

Faber's observation was simple. Most of the stock market's worst losses happened after prices had already fallen below their long-term trend. A slow moving average filters out everyday noise. It catches the long, grinding bear markets that do the most damage, like 1973–74, 2000–02, and 2008.

The rule does not try to predict anything. It accepts that it will sell late and buy back late. The goal is to miss the middle of a long decline, not to call the top or the bottom.

How it works#

At the close of the last trading day of each month:

Compare the S&P 500's price with its 10-month moving average.
If the price is above the average, hold the S&P 500 (SPY).
If it is below, hold Treasury bills (BIL).

Faber applied the same rule to each of five asset classes in his GTAA portfolio. This page covers the single-asset version on US stocks. The five-asset GTAA did not meet our listing bar over the ETF era.

What the backtest shows#

Over the ETF era the timing model beat both the S&P 500 and a 60/40 portfolio on risk-adjusted return, and cut the worst drawdown by more than half. Its defining year was 2008, which it finished slightly positive after leaving stocks early in the year, while the S&P 500 lost more than a third.

The cost is visible too. In steady years it matched the market exactly, but it fell far behind in 2019 and 2023, when it sold into a decline and bought back after much of the rebound. It also lost a little more than the S&P 500 in 2022, a choppy decline in which it stepped aside and back in at the wrong times.

When it struggles#

Fast crashes and rebounds. A monthly check on a slow average reacts too late to a crash that reverses within weeks.
Sideways markets. Prices crossing back and forth around the average cause repeated small losses, known as whipsaws.

Using it on Tactfolio#

The live strategy above runs the rule on daily data with a 210-session average, the closest daily equivalent to ten month-ends. Copy it to test other lengths or other markets, or combine it with Faber's approach on bonds and gold.

Year by year

Year	Strategy	SPY
2026*	3.3%	14.0%
2025	12.3%	17.7%
2024	24.9%	24.9%
2023	9.3%	26.2%
2022	-20.5%	-18.2%
2021	28.7%	28.7%
2020	14.5%	18.3%
2019	10.3%	31.2%
2018	-6.4%	-4.6%
2017	21.7%	21.7%
2016	10.4%	12.0%
2015	-4.6%	1.2%
2014	13.5%	13.5%
2013	32.3%	32.3%
2012	16.0%	16.0%
2011	-2.2%	1.9%
2010	8.9%	15.1%
2009	22.6%	26.4%
2008	1.6%	-36.8%
2007*	-3.4%	-3.4%

Year	Strategy	SPY
2026*	3.3%	14.0%
2025	12.3%	17.7%
2024	24.9%	24.9%
2023	9.3%	26.2%
2022	-20.5%	-18.2%
2021	28.7%	28.7%
2020	14.5%	18.3%
2019	10.3%	31.2%
2018	-6.4%	-4.6%
2017	21.7%	21.7%
2016	10.4%	12.0%
2015	-4.6%	1.2%
2014	13.5%	13.5%
2013	32.3%	32.3%
2012	16.0%	16.0%
2011	-2.2%	1.9%
2010	8.9%	15.1%
2009	22.6%	26.4%
2008	1.6%	-36.8%
2007	5.0%	5.1%
2006	15.8%	15.8%
2005	4.8%	4.8%
2004	10.7%	10.7%
2003	22.8%	28.2%
2002	-4.8%	-21.6%
2001	3.3%	-11.8%
2000	-0.1%	-9.7%
1999	13.3%	20.4%
1998	12.4%	28.7%
1997	33.5%	33.5%
1996	22.5%	22.5%
1995	38.0%	38.0%
1994	-2.5%	0.4%
1993	9.7%	9.7%
1992	7.6%	7.6%
1991	17.1%	30.3%
1990	-14.9%	-3.2%
1989	31.5%	31.5%
1988	6.3%	16.4%
1987	7.4%	5.1%
1986	18.6%	18.6%
1985	31.6%	31.6%
1984	2.5%	6.2%
1983	22.4%	22.4%
1982	29.3%	21.6%
1981	-6.1%	-4.9%
1980	27.7%	32.4%
1979	12.3%	18.3%
1978	1.8%	6.5%
1977	-4.2%	-7.2%
1976	23.8%	23.8%
1975	22.1%	37.1%
1974	8.0%	-26.5%
1973	-15.2%	-14.7%
1972	18.9%	18.9%
1971	5.3%	14.1%
1970	15.7%	3.9%
1969	-4.3%	-8.4%
1968	5.2%	10.9%
1967	14.8%	23.8%
1966	-2.9%	-10.1%
1965	10.8%	12.4%
1964	16.4%	16.4%
1963	22.7%	22.7%
1962	-5.1%	-8.8%
1961	26.8%	26.8%
1960*	-2.1%	0.4%

* Partial year.

The rules as implemented

This is the exact tree Tactfolio runs, rebalanced monthly with signals and trades at the close. Open it to inspect or copy it.

Sources and caveats

Mebane Faber, A Quantitative Approach to Tactical Asset Allocation (2007, updated 2013)

The daily 210-session average stands in for Faber's average of ten month-end closes.
SPY stands in for the S&P 500 total return index and BIL for 90-day Treasury bills.
Signals and trades use the close of the last trading day of each month, as in the source.

Common questions#

What is Faber's 10-month moving average strategy?#

It is a trend-following rule from Mebane Faber's "A Quantitative Approach to Tactical Asset Allocation". You hold the S&P 500 while its month-end price is above its 10-month average, and Treasury bills otherwise.

Is the 10-month moving average the same as the 200-day?#

Nearly. Ten months is about 210 trading days, close to the widely watched 200-day average. Tactfolio uses a 210-session daily average.

How often does the timing model trade?#

It checks once a month and trades only when the price has crossed the average since the last check. In a steady trend it can hold the same position for years.

Does the 10-month moving average still work?#

Over the ETF era it cut the worst drawdown by more than half and beat a 60/40 portfolio on risk-adjusted return. Since 2013 it has trailed the S&P 500, as trend following usually does in a long bull market.