Swedroe: Do Funds Earn Their Fees?

January 10, 2018

In a recent article, The New York Times highlighted the high cost of funds from the Baron Funds group, which manages about $26 billion in assets. For example, the flagship fund, Baron Growth (BGRFX), which manages more than $6 billion, has an expense ratio of 1.3%. Baron Funds founder Ronald Baron argues in the article his skills and experience justify the fund family’s high fees.

To see if this is actually the case, we can put BGRFX’s performance through a regression analysis, which shows the source of its returns. This allows us to determine whether the fund’s returns were a result of loading on common factors, or if they derive from stock-picking and market-timing skill (alpha).

A Look At The Data

Before we do that, however, we can look at the raw performance figures. Morningstar classifies the fund as a midcap growth fund. Thus, we’ll compare its performance to that of the SPDR S&P 400 Mid Cap Growth ETF (MDYG), which has an expense ratio of 0.15%.

For the 10-year period ending December 2017, BGRFX returned 8.3% and underperformed the benchmark MDYG by a full 2 percentage points per year. Given that the difference in expense ratios was just 1.15 percentage points, the stock-selection and/or market-timing efforts of BGRFX negatively impacted returns by a further 0.85 percentage points.

We’ll now move on to the regression analysis. The regression tool available at Portfolio Visualizer permits us to review results back to January 1995 through October 2017. Through the lens of a six-factor (market beta, size, value, momentum, quality and low beta) regression, we find that the fund had generated virtually no alpha (0.1%) over the full period.

The way to think about this is that the fund demonstrated just enough stock-picking/market-timing skill to offset its 1.3% expense ratio. However, after January 2000, the fund’s alpha has turned negative (-1.1% a year). Since then, the fund’s 1.3% expense drag has become too much of a burden to overcome.

Additional Factors

There are a few additional points we need to cover. Several of the factors used in the regression analysis were not commonly known back in 1995, when Baron started his fund. The momentum factor was first introduced into asset pricing models by Mark Carhart in 1997.

And the quality and low-beta factors are much newer still. Over its full life, we see that BGRFX did have statistically significant loadings on three factors: market beta (1.04), size (0.52) and quality (0.21). It had very low and statistically insignificant loadings on the other three factors (value, momentum and low beta).

While today we would consider loading on quality to be beta (that is, exposure to a common factor), when Baron started his fund, it was alpha. As Andrew Berkin and I explain in our book, “The Incredible Shrinking Alpha,” academic research has been converting what once was alpha into beta.

That said, because some of the fund’s performance was due to exposure to quality prior to quality becoming a factor in asset pricing models, its loading on quality at that time should be considered alpha.

However, investors today don’t have to pay high fees to gain exposure to what we now know to be common factors. For example, low-cost ETFs that provide exposure to the quality factor include the PowerShares S&P 500 Quality Portfolio (SPHQ) and the iShares Edge MSCI USA Quality Factor ETF (QUAL).

A Fama-French Lens

We’ll take one more interesting look at BGRFX’s performance through the lens of the Fama-French three-factor (market beta, size and value) model, which was the workhorse asset pricing model until fairly recently. From inception, the three-factor regression shows fund alpha of 2.6%. This contrasts pretty dramatically with the alpha of virtually zero that we obtained from the six-factor analysis.

We can thus conclude the alpha found in the three-factor analysis was due to exposure to other factors (specifically, quality). However, one of the trends we identified in “The Incredible Shrinking Alpha” is that markets are becoming ever more efficient, not only because academics have been converting alpha into beta, but because the skill level of the competition has been persistently rising.

With that in mind, I split the fund’s performance into two roughly equal periods, from 1995 through 2007, and the succeeding period from the beginning of 2008 through October 2017. In the first period, the three-factor regression shows a fund alpha of 3.4%. However, in the succeeding period, the alpha became -0.8%. It seems that market efficiency had caught up with Baron.

No matter how we look at BGRFX’s performance, I don’t believe the fund’s relatively high expense ratio can be justified any longer—the academic research and the efficiency of the market have changed the picture entirely. Viewed properly, there’s no longer any evidence that would justify paying 1.3% a year for performance below that of appropriate risk-adjusted benchmarks.

Larry Swedroe is the director of research for The BAM Alliance, a community of more than 140 independent registered investment advisors throughout the country.

Find your next ETF

CLEAR FILTER