An In-Depth Look at Two Dodge & Cox Funds

The table demonstrates that for the 15-year period, DODGX clearly outperformed DFLVX. It provided higher returns with less volatility. The Sharpe ratio for DODGX during this period was 0.51, versus 0.46 for DFLVX. It also shows that the difference in alphas (1.5 percentage points) accounted for more than the difference in returns. While the size loadings in both cases were virtually identical, DFLVX’s returns benefitted from its higher loadings on the market, value and quality factors. It did, however, have a slightly greater negative loading on momentum, but note that value funds tend to load negatively on momentum. Thus, given DFLVX’s higher loading on value, you would expect it to have the greater negative loading on momentum that it possesses.

It’s also interesting to note that if we don’t include the quality factor, the alphas change quite a bit. DODGX’s alpha increases from 0.4% all the way to 2.7%, and DFLVX’s rises from a negative 1.1% to a positive 1.7%. DODGX goes from a small annual alpha to a huge one. In other words, the fund benefitted more from the type of stocks it bought than the specific stocks it bought. However you cut it, the fund delivered a great experience for investors over the 15-year period.

Before turning to a review of the more recent 10-year period, it’s important to observe that successful active management sows the seeds of its own demise. In his paper, “The Five Myths of Active Management,” University of California at Berkeley professor Jonathan Berk suggested the following thought process: “Who gets money to manage? Well, as investors know who the skilled managers are, money will flow to the best manager first. Eventually, this manager will receive so much money that it will impact the manager’s ability to generate superior returns, and expected return will be driven down to the second-best manager’s expected return. At that point, investors will be indifferent between investing with either manager, so funds will flow to both managers until their expected returns are driven down to the third-best manager. This process will continue until the expected return of investing with any manager is driven down to the expected return investors can expect to receive by investing in a passive strategy of similar riskiness (the benchmark expected return). At this point, investors are indifferent between investing with active managers or just indexing, and an equilibrium is achieved.”

In other words, investors observe benchmark-beating performance and funds flow into the top performers. The investment inflow eliminates return persistence because fund managers then face diminishing returns to scale. Thus, even if past results were the result of skill and not a lucky random outcome, the evidence clearly demonstrates that persistence in outperformance is difficult to maintain.

Empirical evidence supports this theory. The academic research¹ on the performance of pension plans demonstrates that plan sponsors hire managers with records of large alpha in the three years prior to their hiring. Unfortunately, that alpha has a nasty tendency to disappear after these managers come on board.

Returning to our comparison of DODGX and DFLVX, the table below shows the results of the regression analysis for the more recent 10-year period ending September 2014.

Over this period, DODGX produced a lower return than DFLVX – though it did have somewhat lower volatility. In terms of the Sharpe ratio, DODGX’s was lower at 0.44 versus 0.49 for DFLVX. We also see that for this period, DODGX’s alpha completely disappeared and was, in fact, more negative than that of DFLVX. The remainder of DFLVX’s outperformance was due to its higher loadings on the market, SMB and HML factors.

These results indicate that despite outperforming over the full 15-year period, the outperformance was concentrated in the October 1999-September 2004 timeframe.

We now turn to examining the regression analysis results for the most recent 5-year period.

               
The relative results were very similar to those of the 10-year period, with DFLVX producing higher returns with somewhat higher volatility. In terms of the Sharpe ratio, DODGX’s was lower at 1.10 versus 1.18 for DFLVX. In this case, both funds produced alphas relative to our five-factor model. During this period, there was virtually no size premium and the value premium was slightly negative. Thus, the higher HML loading negatively impacted DFLVX’s performance relative to DODGX. DFLVX, however, did benefit from a higher loading on the market factor.

Before expanding our evaluation to another Dodge & Cox fund, there are three points worth mentioning. First, the Dimensional Fund Advisors (DFA) fund benefited from lower expenses. DODGX’s expense ratio is 0.52% while DFLVX’s is just 0.27%. Second, DODGX investors accept significantly more idiosyncratic risk than DFLVX investors. As of September 2014, DODGX held just 69 stocks while DFLVX held 254. Third, DFLVX, as can be seen in the HML loadings, maintained a much more consistent exposure to value stocks than DODGX.

Let’s turn our attention to another of Dodge & Cox’s highly regarded funds, the Dodge & Cox Balanced (DODBX).

Dodge & Cox Balanced Fund

As of December 19, 2014, Morningstar gives the fund a 3rd percentile ranking for its 15-year performance. The fund’s current holdings are approximately 70% stocks and 30% bonds. However, it does have the ability to shift allocations between stocks and bonds, and also between long-term and short-term bonds, making a simple analysis a bit more difficult. For example, more than 40% of the fund’s bond holdings currently have maturities of greater than 15 years. Keep in mind we’ve basically been in a bond rally for most of the time period under consideration. We’ll compare DODBX’s annualized returns and annual standard deviation to those of a portfolio allocated 70% to DFA’s U.S. Large Cap Value Fund (DFLVX) and 30% to DFA’s 5-Year Global Fixed Income Fund (DFGBX) with annual rebalancing.  

The picture is pretty similar to what we saw with Dodge & Cox’s Stock Fund (DODGX). For the full 15-year period, DODBX provided returns 1 percentage point higher than the DFA stock/bond fund portfolio. But in this case, it did so with higher volatility. However, the Sharpe ratio was higher at 0.60 versus 0.53. In the more recent 10- and 5-year periods, DODBX provided slightly lower returns and also somewhat higher volatility, resulting in lower Sharpe ratios in both cases.

Summary

Despite the excellent long-term results of the two Dodge & Cox funds we examined, Jonathan Berk’s warnings still apply. While the funds delivered outperformance over the full 15-year period, all of that outperformance occurred in the first one-third of that timeframe. For the last 5- and 10-year periods, the two funds have failed to outperform well-designed passive, lower cost alternatives.

I would add this warning to Berk’s: We only know today that these two Dodge & Cox funds have performed quite well. And – with the benefit of hindsight – we have been able to use two of the market’s best performing funds (according to Morningstar’s rankings) in our comparison with similar passively managed funds. Clearly, the average actively managed fund was far worse in comparison.

Larry Swedroe is director of research for the BAM Alliance, a community of more than 150 independent registered investment advisors throughout the country.

Read more articles by Larry Swedroe