Members of our Quant Investing community are well aware that I am an advocate of combining Value and Momentum. By “Combining,” I mean, splitting one’s portfolio such that a portion of it (usually the larger portion) is allocated to Value stocks and the remaining portion – to Momentum stocks.
In this spotlight article, I would like to focus on the reasons why combining Value & Momentum is so important, especially to the retail investor. Whether you select your stocks quantitatively or manually, I will give you a few actionable tips to boost your current portfolio’s returns and reduce its volatility.
Recap on Combining Value and Momentum
Adding Momentum stocks to any Value Investing portfolio results in increasing (typically, but not always) the average annual returns and reducing the portfolio’s volatility and drawdowns. That’s because Momentum and Value both have positive and high return expectations, coupled with a low correlation between them.
In the February 2019 newsletter, I have written on the various methods to combine Value and Momentum. I have shown that the top researchers and practitioners are divided on this very topic. While Grey and Vogel or Alphaarchitect, as well as Corey Hoffstein of Newfound Research, believe that “Combining” (as described above) is the way to go, the respected folks from AQR believe that Value and Momentum criteria should be integrated into the stock selection process (i.e., select stocks which rank high both on Value and Momentum).
Why Does it work so well
Value and Momentum behave differently.
The common behavioral explanation of why Value works – is as follows. Stocks become cheap when companies experience problems, headwinds or risks. In such cases, investors tend to overreact and drive stocks down below their intrinsic economic value. Our human tendency to overreact upon bad news is a well-known cognitive bias, that was researched extensively by Nobel laureates Amos Tversky and Daniel Kahneman, as well as popular researchers such as Prof. Dan Ariely and James Montier. Moreover, institutional investors are often systematically biased to chase short-term gain, and would rather not hold stocks of failing companies due to career risk.
Moreover, by investing in cheap value stocks, investors front-run market expectations. They buy stocks when the near future looks dire. But more often than not – failing companies revert to the industry mean and perform on par with the industry averages. The scent of good news changes the general public’s expectations, and the stock goes up as a result. Investors who are invested early enough in that cycle, front-run the larger market’s expectations and could earn big returns.
Momentum stocks are the mirror image of Value stocks. Stocks go up and experience a good momentum typically after having good news. But it seems that most investors under-react to good news, and thus it takes time until the stock prices fully reflect the news. By buying momentum stocks, investors front-run expectations by entering positions before they have fully exhausted their appreciation potential.
Thus, Value and Momentum have very different mechanisms to realize Value. Both have been rigorously researched, and have shown to outperform beyond any reasonable doubt. I’ve been writing about the outperformance potential of Value and Momentum extensively in the blog, as well as referenced the major research on the topic (see also the Recommended Books section below for cornerstone resources).
But what if you combine those two assets together? Is the result better than the sum of its parts?
Well, it turns out it is.
Let’s examine why it works, at least in theory. It is a well-known mathematical fact, that when two random series are combined, the resulting series’ amplitude (i.e., returns) is the average of the two, but the volatility (as measured by the standard deviation of returns) is lower than each of the comprising series.
In the chart below, I show two random series, series1, and series2. Those are random walks generated by a computer and do not represent any real stocks. The grey line is the average of those two series. It is equivalent to investing half our money in the first asset (series1) and the other half in the second asset (series2). As expected, the combined series cumulative return is the midpoint of the two series.
But, notice the volatility. The combined series (in grey) is much more monotonous. Its peaks and throughs are dampened.
Let’s assume that the orange and blue lines represent the returns of the Value and Momentum portfolio. Let’s also assume that you do not know apriori which of the two, Value or Momentum, will deliver higher returns, but you know that both are similarly likely to outperform the benchmarks. Which of the three assets would you prefer to invest in? Value, Momentum, or the combined portfolio?
I know that I would always prefer the combined portfolio. Its relative stability will help me sleep better at night, and I will be much more likely to outperform the benchmarks even if either Value or Momentum do not perform well during the investment period.
And just for the fun of it, here are additional runs of 2 such random variables:
The examples above are for assets that are not correlated. The two are completely random, and they have nothing in common.
In reality, Value and Momentum ARE correlated. They do behave with some relation to each other. They both are correlated with the economy, investors sentiment, and market behavior in general. When the correlation of Value and Momentum is positive, they will behave as the theoretical charts above. When the correlation of Value and Momentum is negative (when one goes up, the other goes down), the combined return can even exceed both assets’ performance. For long-only portfolios, I’ve seen only low positive correlations. For long-short portfolios, researchers have reported negative correlations.
Here’s historical Value & Momentum charts, created for long-short portfolio, as taken from Cliff Assness’ 2013 paper, Value & Momentum Everywhere:
Due to the negative correlation of the long-short Value & Momentum portfolios, the Combo portfolios enjoyed returns, which are higher than both Value & Momentum.
Now that we have a better notion of the theory behind combining Value and Momentum let’s examine how the quantitative strategies on our Quant Investing community have performed.
We start by recapping the Lion Portfolio historical performance but now extended from 19 years, from the June 30th, 1999, to June 30th, 2018. I’d like to remind you that we’re always simulated June 30th to June 30th, to remain consistent across all our publications, and to be consistent with the academic research convention.
Here’s how the Quantitative Value portfolio of the Lion portfolio performed during this time period. The portfolio consists of 30 stocks with a market cap in the upper 60% (approx. $200M and up).
And here’s how the Quantitative Momentum portion performed. The portfolio consists of 20 stocks with similar market cap restrictions.
And now to the combined Value (60%) and Momentum (40%) portfolio:
As you can see, the Sharpe ratio for the combined portfolio (1.18x) is higher than that of the QV portfolio (1.04x) and the QM portfolio (1.13x). The return of the combined portfolio (17.11% per year) is almost as high as QM (17.13%) and higher than QV (16.8%).
Let’s look at additional statistics:
(minute differences between the figures above result from simulation constraints, but are immaterial)
The volatility of the combined portfolio (12.66%) is lower than that of QV (14.22%) and QM (13.37%). The Beta and max drawdown is between those of QM and QV.
Overall, an investor is better off investing in the combined portfolio, getting returns that are roughly the same, but which much reduced volatility.
Nevertheless, combining portfolios doesn’t mean that it beats the market every single year, or that even it gains positive returns every single year.
Here’s how year over year returns look like in our simulation:
Our combined model portfolio made negative returns during 1999, 2007, 2008 and 2015, and lagged the S&P 500 during 1999, 2013-2015, and in 2017.
Winning in the long-term requires surviving short-term underperformance. As we always say, “It works because it doesn’t always work.” As individual investors, our biggest edge is not being subjected to career risk. No one will fire us if we don’t beat the market every quarter or every year (except maybe our spouse). We CAN adopt a long term view, and we should (if we fancy super high returns, of course).
But does combining Value & Momentum result favorably with portfolios other than Quantitative Value?
Here is the performance for a portfolio of 30 Deep Value stocks (selected with VC2 and a low beta rule), market cap in the upper 70% (approx. $30M and up):
The momentum portfolio portion will be similar to the Lion portfolio case above.
The combined DV and QM portfolio’s performance is shown below:
In this case, as well, the total returns are not impaired, while the volatility, as measured by both std. Deviation and beta are lower than the Deep Value portfolio standalone.
However, in this case, the Sharpe ratio of the combined portfolio did not exceed that of the QM portion.
As we cannot know apriori which of the portfolios, Value or Momentum, will perform better, it would have been wise to combine them, even in this case.
Tips for Discretionary Investors
What should you do if you already have a Value portfolio and wish not to alter it in any way? Let’s say that you are a Buffett-like investor, picking stocks after careful manual analysis, and your portfolio is concentrated with a few big bets.
How can you benefit from the merits of combining Value and Momentum portfolios?
For this case, let’s simulate a concentrated Value portfolio by using Graham’s defensive strategy. We will construct a portfolio of just 15 stocks, all S&P 500 constituents. Those will serve as our proxy to the big value names that many value investors hold.
The performance of such a portfolio over the same time period of mid-1999 to mid-2018 is as follows:
Suppose our portfolio value is $100K.
Now let’s assume that we take $20K and investing them in a Quantitative Momentum portfolio of 20 stocks. Each position will be a mere $1,000. If we use a discount broker such as Interactive Brokers, we will pay for a transaction a mere $1, which is 0.1%, a very reasonable rate that will not impact our overall performance in any material way.
The Combined Value portfolio (80%) and QM portfolio perform as follows:
As shown above, adding a little bit of Momentum boosted the annual portfolio returns from 9.4% to 11.06%. Not only that, but it also reduced its volatility from 11.45% to 10.77%. As a result, it increases the Sharpe ratio from 0.69x to 0.87x.
Spicing up a value portfolio with even just a little bit of MomentumMomentum contributes to higher returns and lower volatility.
Even conservative investors are better off adding some momentum.
If the portion of the QM portfolio rose to 40%, the results would be even better:
Aggressive investors may ask if Momentum is so great, why not go all-in?
Well, you should make your own decision, but I personally would not go that path. Momentum has historically been a volatile strategy, with prolonged periods of underperformance. As a diversifier, it works great, but standalone Momentum may be too risky, even to my taste.
Many value investors, especially conservative ones, avoid investing in MomentumMomentum. Momentum is perceived as speculative. Standalone Momentum is indeed volatile, risky and requires a different mindset than Value.
But by avoiding Combining a Value portfolio with Momentum, investors are making a compromise.
Allocating some of the portfolio funds to Momentum contributes to reducing volatility and risk, due to the simple mathematical fact that the correlation between Value and Momentum is low, and sometimes even negative. The conservative choice would be to combine Value and Momentum, rather than stick only with Value.