Fama-French 5-Factor Asset Pricing Model

Motivated by a question related to the Morningstar style box, what factors motivate the ITA investment philosophy.  Is the style box or SAA model valid even when the securities used to populate the various boxes are highly correlated?  The SAA model came on hard times during the Great Recession when seemingly low correlated investments all moved together – DOWN.  The Great Recession caused investors to rethink their models and it appears as if “factor investing” gain significant strength.

In every portfolio review of the 14 portfolios tracked and reported on monthly, readers see what we call the Dashboard.  The Dashboard table comes right out of the TLH Spreadsheet and lays out the Strategic Asset Allocation (SAA) plan for each portfolio.  Regardless whether or not the SAA is followed and rebalanced according to the target percentages, we maintain an asset allocation for each portfolio so as to calculate the ITA Index.  The ITA Index is a reference point.  If the managerial model of choice does not outperform the ITA Index, then it behooves one to examine or shed the model and use the SAA model.

While the SAA model is used with the Schrodinger, Copernicus, and Pasteur portfolios, eleven portfolios deviate from the SAA model in an effort to reduce portfolio risk. And now we come to the Fama-French 5-Factor Asset Pricing Model.  Which of the five factors can we make use of with ETFs and what are the drawbacks to using these factors.

Here are the five factors.  Let me state right off that the factor count is either murky, or what counts as a specific factor has some overlap with other factors.  Beta, the original factor, has been swallowed up by the size and value factors.

1. Small-Cap stocks tend to outperform the broad market.
2. Value stocks (low price/book/share) tend to outperform the broad market.
3. Momentum or stocks moving up tend to continue to move up.  Stock trending down continue to move down. [Editor Note:  Momentum is not one of the FF factors.  Beta is a factor, which I do not include in this list.]
4. Profitability – This factor was uncovered by a former student of mine, Robert Novy-Marx.
   • Here is the abstract of Novy-Marx paper.
   • “Profitability, measured by gross profits-to-assets, has roughly the same power as book-to-market predicting the cross section of average returns. Profitable firms generate significantly higher returns than unprofitable firms, despite having significantly higher valuation ratios. Controlling for profitability also dramatically increases the performance of value strategies, especially among the largest, most liquid stocks. These results are difficult to reconcile with popular explanations of the value premium, as profitable firms are less prone to distress, have longer cash flow durations, and have lower levels of operating leverage. Controlling for gross profitability explains most earnings related anomalies and a wide range of seemingly unrelated profitable trading strategies.”
5. Investment – Here is a quote from Phil DeMuth that sheds a little light on the meaning of the Investment Factor.

“Imagine that a company announces they are going to invest a lot of money in some new project.  Is this good news or bad news?  Should you buy or sell?  Recall that firms tend to invest a lot when their profitability is high and their cost of capital is low.  It sounds promising, right?

Before you run out beating bongos in the woods, consider this:  In 20o4, researchers Titman, Wie and Xie controlled for the relevant variables and found that firms that significantly increase capital investment tend to achieve sub-par subsequent returns.

“The new ‘investment’ factor has a high correlation to the value and profitability factors.  The investment effect is perhaps half as strong, but it is still reliable and significant.  Surprisingly, when you statistically analyze the performance of stocks, the value factor completely drops out of the equation and can be replaced by the beta, size, profitability and investment factors.  That’s interesting.  At a distance, value gives you a quick and dirty approximation for beta, size, profitability and investment.  But up close and personal, once you take beta, size, profitability and investment into account, value isn’t bringing anything new to the hootenanny.

By the new model, the highest expected returns can be expected from companies that are small, value (high book-to-market, for example), and profitable that are not embarking on major growth initiatives.”  Value is many times defined as low Price/Book/Share.

Larry Swedroe writes a good article on the Five-Factor Model.

Of the above five factors, we are able to find ETFs that satisfy small size, value, and momentum using our own rating system.  That leaves out the two most recent factors, profitability and investment.  Never fear as the first three also cover some of the ground occupied by the last two on the list.

And now for a little honesty about these factors.  There is nothing fundamental driving these factors.  As frequently stated elsewhere on this blog, we don’t have a clue why momentum works.  If you run a regression analysis, value, size, and momentum add alpha to a portfolio.  At least they have in the past.  If we don’t know what is driving these factors there is a non-trivial probability they will not work as expected in the future.  Fama and French do not attempt to interpret or explain their results.  Running regressions may squeeze out results, but it is not a theory unless there is strong evidence why the results turn out as they do.  Thus far those explanations go wanting.  Most of the ITA portfolios have been testing the merits of momentum and it appears to be working based on performance trends.  Will this continue and how much do we want to back on this single factor?

Take an extreme example that might eventually show up in the Rutherford Portfolio.  If one ended up with only VTI outperforming SHY, “heart burn” would likely be at a minimum if the model directed one to invest a million dollars in a single ETF, VTI.  However, what if the model directed us to invest the million dollars in gold (GLD) and only GLD?  Would that give you pause?  It would me and that is why I still keep one foot in the SAA model door.  One modification, and it plays a role in the SAA model, is to place a cap on the percentage one can invest in any single ETF.  Regardless of the portfolio model used, judgment is involved.