Tilting the three factor model

By John De Goey | August 19, 2011 | Last updated on August 19, 2011
4 min read

Many advisors who use Dimensional Fund Advisors (DFA) often struggle in explaining how DFA products differ from various other products that investors might choose. One might say that DFA’s approach is essentially passive, but does not use conventional cap-weighted indexes (like Barclay’s or Vanguard). Right away, that confuses people.

Many people who only think of comparing and contrasting traditional active products and strategies with non-intuitive passive strategies don’t understand how different DFA’s are. There are other differences in DFA’s approach too, but the commitment to an academic approach that eschews market capitalization-weighting in favour of ‘enhancements’ that aim to capture the dimensions of risk and return is easily the most notable.

I’ve recently come across a metaphor that I hope will make it easier to explain this. Let’s say you have an ice-cube tray shaped like a tic tac toe board – a 3 x 3 grid. This grid represents the ‘style boxes’ that are now quite well-known and widely accepted since they are used extensively by Morningstar (and others). In actual fact, the style box depictions are simply a stylized version of the Fama/French Three Factor Model, so we’re really just getting back to basics here.

To review, the factors are:

• stock market exposure as compared to bonds; • small company exposure as compared to large companies; and • value exposure as compared to growth.

The Fama/French Three Factor Model builds on earlier work done by William F. Sharpe called the Capital Asset Pricing Model (CAPM). It turns out that the Sharpe Model explains about 2/3 of risk and return. Adding the two additional risk/return factors of size and value allows us to explain about 90% of risk and return.

In short, the model suggests that risk and return are related – all you have to do in order to get higher returns (on average) is to take on more risk. Since small companies are riskier than big ones and value companies are riskier than growth companies, one can reasonably expect to get above-market rates of return over time by tilting a portfolio toward them. Ice cube trays, as anyone with a fridge that is less than entirely level can tell you, allow water to flow to the lowest point before it freezes. The same kind of tilting can be done with style boxes.

We would expect this because no one would rationally invest in something that was riskier than a viable alternative if there wasn’t a payoff for doing so. Stated a little differently, no one would rationally invest in a risky alternative if something with an identical expected rate of return (but with lower variability) existed.

That’s where the ice cube tray comes in. A simple 3 x 3 grid can illustrate how one might ‘tilt’ exposure to certain market elements that allow for greater than market returns because they involve greater than market levels of risk. If one were to think of the three sets of compartments going from left to right as being value, blend and growth and the three compartments going from top to bottom as being small cap, blend and large cap, then it becomes easy to see how one might get better performance – simply tilt the ice cube tray toward the top left so that there’s relatively more invested in the small cap and value compartments and less in the large cap and growth compartments.

When looked at in this manner, a traditional market cap portfolio (what most people think of when discussing ‘indexing’) is simply an ice cube tray on a flat surface. There’s no tilting going on and, as a result, there’s no extra emphasis on any part of the overall market. As such, products like those offered by Barclay’s and Vanguard are ‘flat ice cube tray’ offerings.

In contrast, the DFA products might be thought of as ‘tilted ice cube tray offerings’ that allow investors to more discreetly calibrate their own personal risk preference (risk/ reward trade-off). Flat ice cube tray offerings are legitimate products that aim to provide market level rates of return (minus costs) in exchange for market weighted levels of risk.

Perhaps the most important element of this trade-off is that historical data shows how people can improve upon their expected risk-adjusted return. In short, the relationship is not linear. A medium-sized increase in expected return with only a small increase in expected risk is something many people would be very interested in.

Therefore, one can reasonably expect to add more return than risk when doing the tilting. If you could get an additional 2 or 3 units of expected return in exchange for taking on only 1 or 2 units of expected risk, would you do it? Most people would, but the degree to which they would do so would be a highly personal decision based on ‘risk ‘preference’.

Once again, I must stress that market levels of return with market levels of risk is a totally honourable decision and that the decision is a highly personal one.

There is no ‘right’ or ‘wrong’ decision; there are only points along the continuum that are consistent with each investor’s own views of how risk and return are perceived.

John De Goey, CFP, is the vice president of Burgeonvest Bick Securities Limited (BBSL) and author of The Professional Financial Advisor II. The views expressed are not necessarily shared by BBSL. You can learn more about John at his Web site: www.johndegoey.com.

John De Goey