Moving-average trend following: how a trend filter works
A moving average turns a noisy price chart into one simple question: is the market trending up or not? The answer has historically helped investors avoid the worst of long declines.
In short
A moving-average trend filter holds a market when its price is above its average price over a set period, such as 200 days or 10 months, and steps aside when it falls below. Research going back to 1901 suggests rules like this have cut volatility and drawdowns while keeping much of the market’s return. The cost is whipsaws: in sideways markets, the price crosses the average repeatedly and each round trip loses a little.
What a moving average is
A moving average is the average closing price over a recent window, recalculated each day. A 200-day simple moving average adds up the last 200 closing prices and divides by 200. Tomorrow, the oldest price drops out and the newest is added.
Because it averages many days, the line is smooth and slow. It rises when the market has been rising for months and falls when it has been falling, which makes it a simple way to describe the direction of a trend.
The basic rule
The simplest trend filter has two states.
- When the price is above its moving average, hold the market.
- When the price is below its moving average, step aside, usually into cash or short-term Treasury bills.
Many investors check the rule monthly using a 10-month average, which is close to 200 trading days. Checking less often reduces trading but reacts more slowly.
What the research shows
Mebane Faber’s 2007 study, “A Quantitative Approach to Tactical Asset Allocation,” tested a 10-month moving-average rule on U.S. stocks back to 1901 and on several other asset classes. The rule produced returns similar to buying and holding, with lower volatility and much smaller drawdowns. Its main benefit was keeping investors out of most of the long bear markets.
Broader studies of trend following, including work by Hurst, Ooi and Pedersen covering more than a century of data across many markets, reach a similar conclusion: trend rules have tended to help most during extended declines.
Why it works when it works
Large market declines rarely happen in a single day. Most unfold over months as earnings, credit conditions and sentiment worsen. A trend filter will not catch the top, but it tends to exit before the deepest part of an extended decline and re-enter after a recovery is under way.
Historically, equity volatility has also tended to be higher when prices are below their long-term average. A trend filter therefore tends to keep investors out of the most turbulent periods.
The cost: whipsaws
A trend filter is wrong often. When a market moves sideways, the price can cross its average again and again. Each time, the rule sells after a small drop and buys back after a small rise, losing a little on each round trip.
This is the price of protection. Levment’s deepest drawdown came in 2011 and 2012, when the market moved sideways and trend signals reversed repeatedly. Trend following accepts those small, frequent losses in exchange for avoiding the rare, large ones.
Design choices
- Length. Shorter averages react faster and trade more often. Longer averages trade less but give back more before exiting.
- Type. A simple average weighs every day equally. An exponential average weighs recent days more heavily and turns faster.
- Buffers. Requiring the price to move a set distance past the average, or to stay there for several days, can reduce whipsaws at the cost of later signals.
- Frequency. Checking daily, weekly or monthly changes how quickly the rule acts and how often it trades.
Why trend filters pair with leverage
Leveraged ETFs do best in steady uptrends and lose value in choppy, volatile markets, a pattern called volatility decay. A trend filter tends to hold exposure in exactly the conditions leverage needs and step aside in the conditions that hurt it most.
Levment’s system starts from a moving-average framework and adds proprietary signals to refine entries and exits. The goal stays the same as the simple rule: amplify confirmed strength and protect capital when it fades. Read how the strategy works.
Sources
- Faber, M. (2007). “A Quantitative Approach to Tactical Asset Allocation.” Journal of Wealth Management 9(4).
- Hurst, B., Ooi, Y. H. and Pedersen, L. H. (2017). “A Century of Evidence on Trend-Following Investing.” Journal of Portfolio Management 44(1).
- Moskowitz, T., Ooi, Y. H. and Pedersen, L. H. (2012). “Time Series Momentum.” Journal of Financial Economics 104(2).
This article is educational and is not investment, tax or legal advice. Examples are illustrative and are not Levment results.