Last Updated on September 19, 2022
Meb Faber, a famous money manager and writer, has several times stated he is a trend follower at heart. Back in 2015, he published an article about momentum and trend-following strategy called Meb Faber’s Three-Way Model in gold, stocks, and bonds– a model that Meb Faber (Mebane Faber) found in some research from Ned Davis Research. However, we would label the strategy more a momentum strategy than a trend-following strategy. How has this strategy performed and is it still working?
Yes, Meb Faber’s momentum/trend-following strategy in gold, stocks, and bonds is still working. The strategy is as simple as it gets: Initiate (or keep) a position if the monthly 3-bar moving average is above the 10-bar moving average. The strategy returned 13% annually from 1971 to 2015. We updated and backtested the strategy and show you the results after 2015.
Before we go on to explain Meb Faber’s momentum/trend-following strategy we start by explaining what momentum strategy investing is:
What is momentum in trading and investing?
Momentum investing is an approach that seeks to buy stocks with the best historical performance over a given period and then periodically rebalance the portfolio such that at any given time, it’s invested in the stocks with the highest momentum. Momentum is quite similar to trend-following strategies.
This approach is based on the momentum hypothesis, which believes that there is a strong correlation between 3-12 months historical return and 3-12 months future return. That is, stocks that performed the best over the medium term (3-12 months) are likely to continue performing well in the near future, say the next 3-12 months.
In backtests going back decades, even a century back, momentum has proven to be less significant for periods shorter than 3 months and longer than 12 months.
In one study, Alpha Architect analyzed stocks in the top 40% of the largest stocks on NYSE based on market value from 1974 to January 2016. They measured the return on each stock over the past 12 months and rebalanced the portfolio each month by investing in the top 10% stocks with the best price performance. They deducted 1% annually for transaction costs and another 1% for the annual management fee. Their findings showed an annual return of 14.4% against a 10.92% annual return for the S&P 500. This is a pretty significant improvement to buy and hold.
However, when you are testing the momentum anomaly you need to filter data for survivorship bias. Survivorship bias in trading and investing can have a major impact on the results! (Alpha Architect adjusted for that.)
How does momentum investing work – in practice?
Here’s an example of how you can utilize the momentum factor in practice:
- A momentum investor buys stocks that have performed best over the last 6 months (can be 3, 9, or 12 months)
- The investor rebalances his portfolio every month or 3 months by selling the least performing ones and buying new stocks that are currently performing best.
Pretty simple. Rinse and repeat 4 times per year.
What is Meb Faber’s Three-Way Momentum/Trend Following Model Strategy?
Mebane Faber published the strategy in a white paper. Although he is (perhaps) more famous for being crazy about podcasts, he is an avid reader and has published a lot of stuff. He’s a great resource.
Anyone that has read the writings of Meb Faber knows that he likes things simple and easy. That doesn’t mean mediocre strategies. Quite the opposite. When trading and investing, simplicity trumps complexity.
And Meb Fabers Three-Way momentum/trend-following strategy is really simple (it’s not a timing model). Here is all there is to it:
Three asset classes: Stocks, bonds, gold.
Invest equally in whatever is going up (defined as 3 month SMA > 10 month SMA).
The article didn’t have any reference to capital allocation, but we assume the whole equity is divided equally or 100% if it’s only one asset class that has positive momentum any month. If there are two positions, 50% each, etc.
How has Meb Faber’s Three-Way Momentum/Trend-Following Model Strategy performed?
Meb Faber tested the strategy from 1971, when the USD was released from the gold-peg, until 2015. The results looked like this:
|Momentum system||Equal weight||Stocks||Bonds||Gold|
As you can see, the simple momentum strategy performs the best – by far.
We updated and backtested Meb Faber’s Three-Way Momentum/Trend-Following Model Strategy
We backtested the strategy by using the most appropriate ETFs: GLD (gold), SPY (stocks), and TLT (bonds). We have data back to 2006 and start our backtest from then.
The most difficult part of this strategy is making the code to make the capital allocation dynamic:
- If any month has just one long position, this position gets 100% of the equity allocated.
- If we have two positions, we allocate 50% to each.
- If we have positions in all three asset classes, we allocate 33% to each.
The chart below shows (in the lower pane) how often we have 1,2, or 3 positions:
Since 2006, only twice have we been flat (no positions).
Let’s go on to backtest the performance:
We started with 100 000 in 2006 and ends up with 422 000. The annual return (CAGR) is 8.9%, the win rate is rather low at 57%, and the profit factor is 1.6. But the interesting part is the low max drawdown: only 18%.
Perhaps somewhat disappointing is the poor performance from 2015: only 5.3% annually, substantially lower than the stock market. However, the backtest reveals that the main strength of the strategy is the low drawdown. 2022, which has been a poor year for stocks, has shown just a small loss.
The monthly and annual performance looks like this:
Meb Faber’s Three-Way Model Momentum/Trend-Following Strategy with leverage
The strategy had a low drawdown, and it might be tempting to use a little leverage (we don’t recommend this in any way).
The chart below shows the performance with a leverage of 1.5:
Optimization of Meb Faber’s Three-Way Model Momentum/Trend-Following Strategy
Is the above test optimized? We have no idea why 3 and 10-month averages were chosen.
Let’s optimize and see what results from the nearby settings reveal:
All the other settings get pretty similar results and thus we can safely say this strategy is not curve fitted.
Do you want the Amibroker code for Meb Faber’s Trend-Following strategy?
The Amibroker code to backtest this strategy is not as simple as it sounds. You need to vary the position size and it could be quite cumbersome. For your convenience, we have put all of our code into one product which you can purchase, including the backtests in this article. The product has 80+ of our other free trading strategies. You can order or have a look at what you get on this link:
Meb Faber’s Three-Way Momentum/Trend-Following Model Strategy – ending remarks
The strategy is a perfect example of why simpler is better in trading. Performance might have been weak lately, still better than the stock market, but this is a strategy that needs to be evaluated over periods of ten years. Patience is important, and Meb Faber’s Three-Way Momentum/Trend-Following Model Strategy has proven to be defensive during major recessions, just like the 200-day moving average.