Are Multiple Factors Better?
Source: Ken French Data, JQR Capital calculations
Are Multiple Factors Better?
Two weeks ago we looked at the idea of combining more than one factor into a multi factor model portfolio. We calculated and presented a correlation matrix showing the average annual correlation between the slopes of the factor payoffs. In this post, we look not at the slopes of the factor payoffs, but of the return streams from investing in just one of the fractiles. We will utilize the best quintile for each factor. This was the smallest quintile (Q1) for the market equity (ME) and investment divided by total assets (Inv2TA) factors. For the other five factors, the best quintile was the highest (Q5). As a reminder, here is the growth of $1 from 1964 to 2022 for ME Q1.
Source: Ken French Data, JQR Capital calculations
This chart shows the ME factor grew from $1 invested at the beginning of 1964 to $1,420 at the end of 2022. This growth pattern was not all peaches and cream with steep losses in 1968 to 1974 and drawdowns in 2006 to 2008.
Which Factors To Choose?
When looking at the results of investing in just the most attractive quintile each calendar year between 1964 and 2022, the individual factor strategies rank as follows.
$1 invested using the BE2ME factor grew to $25,428
$1 invested using the CF2ME factor grew to $15,405
$1 invested using the NI2ME factor grew to $11,821
$1 invested using the Inv2TA factor grew to $11,650
$1 invested using the DP2ME factor grew to $2,383
$1 invested using the OP2BE factor grew to $1,462
$1 invested using the ME factor grew to $1,420
Source: Ken French Data, JQR Capital calculations
At first glance, it is tempting to use just the first four ranked factors that produced the highest annualized returns over the 59 year period. They included three income statement ratios (DP2ME, NI2ME, and CF2ME) and one balance sheet ratio (BE2ME). The visual history charted above shows returns that may have significant correlations - making a combination of the factors into a multi factor model less powerful than a single factor.
How To Combine Them?
Correlation is a measure for how two things (in our case, two factor return histories) move in the same - or the opposite - direction over time. The table shown below confirms the relatively high annual return correlation between the seven factors in the Ken French data that was shown in the previous chart.
Source: Ken French Data, JQR Capital calculations
The conditional formatting for this table clearly identifies the red diagonal and the muted reds of most factor combinations. The only “bright” spots appear to be the ME + DP2ME and the DP2ME + Inv2TA combinations. Unfortunately, there do not appear to be any lower correlations. You may recall that lower correlations effectively smooth the “ride” over time. Based on this table, it may be advantageous to use the ME + DP2ME + Inv2TA factors in a combined multi factor model.
What Is The Effect?
Here is the effect of investing $1 into the best quintile by selecting based on an equal combination of our three multiple factors from the universe in each of the years starting in 1964 and ending in 2022. We can see that the combination of the three factors mutes some of the drawdowns in each of the three biggest loss periods experienced by most of the individual factors. This muting of the “pain” also limits the overall upside.
Source: Ken French Data, JQR Capital calculations
Perhaps there is a way to be somewhat smarter about combining these factors. There might even be a larger universe (or zoo - as some describe) of factors from which to select. We will dive deeper into those topics in early 2024. Our next stop is the asset allocation pie chart. Please stay tuned!
“Without the method of learning, you're like a one-legged man in an ass-kicking contest. It's just not going to work very well." - Charlie Munger
Disclosure
Past performance is no guarantee of future results. Any investment involves some amount of risk and may not be suitable for all — or any — individuals. You should consult with your investment advisor before acting on this information.
References
https://mba.tuck.dartmouth.edu/pages/faculty/ken.french/data_library.html
https://www.investopedia.com/terms/c/correlation.asp
Comments
Post a Comment