Where’s the Beef?! Is Your Low-Fee Smart Beta Product Ripping You Off?
Though many of us love having a hamburger out at a restaurant, we hate getting ripped off. But how do we assess whether we are getting our money’s worth?
Most people would focus on the menu prices and compare how much a hamburger costs across restaurants. The decision then becomes easy: Go to the restaurant with the lowest price. Unfortunately, price is an imperfect measure of value (or lack thereof).
Why? Though a hamburger is technically a combination of a beef patty and bun, the key value-add component is the beef patty. Yet, beef patty size can vary a lot from one restaurant to another. Some restaurants offer a lot of “bun,” but not a lot of “beef.”
Back in the 1980s, the Wendy’s fast food chain featured a television commercial in which a little old lady receives a tiny patty accompanied by a giant bun from a fictional competitor. “Where’s the beef?!” she angrily exclaims.
To properly address the fundamental question of overall hamburger value, we need to focus on the beef patty, determine its weight, and calculate a price per pound. Assuming buns cost very little, and assuming similar beef quality, the price per pound for beef is the key metric of value for a buyer to consider. Restaurants selling patties at a low price per pound are a good value irrespective of the headline menu price, while restaurants selling hamburgers at a high price per pound are expensive even if the headline menu price is low. For example, if Restaurant A offers a ½-pound beef patty for $10 (≈ $20 per pound) and Restaurant B offers a ¼-pound beef patty for $7.50 (≈ $30 per pound), Restaurant A is clearly the better value, even though it has the higher headline menu price.
Hamburger Implications for Smart Beta
So what do hamburgers and price per pound have to do with equity-oriented, long-only smart beta products? A lot more than you think.
Long-only strategies are a combination of two return sources: general passive market risk exposure and a “differentiated smart” risk exposure. The general market risk exposure is akin to the hamburger bun, while the differentiated smart risk exposure is the beef.
Differentiated smart risk exposure varies from one smart beta provider to another. The headline basis point (bps) fee charged by a smart beta product is an imperfect measure of value. Why? Because it doesn’t tell us anything about the beef! In order to properly evaluate this headline fee, we need to know the “size” of the smart beta patty. Some smart beta products deliver a lot of differentiated smart risk, while others deliver a lot of general market risk.
Is a smart beta product with a larger amount of differentiated smart risk exposure better than one with less? Not necessarily. We need more information to make a judgment. Though it’s critical to know the amount of differentiated smart risk exposure we’re getting, in order to assess value, we also need to know what we are paying for that exposure.
In order to properly address this fundamental question of value, we need to focus on the differentiated smart risk exposure, figure out the amount of that risk exposure, and calculate a fee per unit of differentiated smart risk exposure (i.e., fee per unit volatility). Assuming similar quality (i.e., similar return per unit of risk), the fee per unit of differentiated smart risk exposure is the key metric of value.
Asset managers selling smart beta products at a low fee per unit of differentiated smart risk exposure are a good value irrespective of the headline bps fee. Consider the following situation:
1. Asset Manager A offers a 15% volatility smart beta strategy for 50 bps. The volatility of only the smart differentiated component (i.e., risk exposure) is 5%.
2. Asset Manager B offers a 15% volatility smart beta strategy for 35 bps. The volatility of only the smart differentiated component is 2.5%, which is equivalent to 70 bps per 5% risk exposure (which is higher than the 50 bps per 5% risk exposure we get with A).
So if we want 5% smart differentiated risk exposure, we can either make an allocation to Asset Manager A or double the allocation and use Asset Manager B. Both options correspond to a total differentiated smart risk exposure of 5%, allowing for an apples-to-apples fee comparison.
Clearly, Asset Manager A is the better value despite the higher headline bps fee. This simple analysis implicitly assumes that generic market risk is essentially free.
Measuring the Smart Beta Hamburger: How Much Beef vs. Bun?
Sounds good in theory. But in practice, in order to make our key metric of value operational, we need to isolate the differentiated smart component of the asset manager’s return stream. Fortunately, this task is a simple application of standard linear regression techniques and is easy to do in Microsoft Excel.
Here’s the recipe:
1. Obtain the smart beta manager’s historical returns, the S&P 500’s (or another relevant market index’s) historical returns, and historical T-bill returns (risk-free rate proxy).
2. Calculate the manager’s historical returns in excess of T-bills by subtracting off the T-bill return. Do the same for the S&P 500.
3. Calculate the manager’s exposure to the S&P 500 by using Excel’s SLOPE function. In Excel, type “= SLOPE (Manager’s Excess Returns, S&P 500’s Excess Returns).” Let’s call this number “BETA.”
4. The manager’s historical differentiated smart return component is obtained by netting out — subtracting — the S&P 500 return component, which is BETA*S&P 500 Excess Return. In other words, DIFFERENTIATED SMART RETURN = MANAGER’S EXCESS RETURN – BETA*S&P 500 EXCESS RETURN. Do this calculation for each holding period.
5. With the manager’s historical differentiated smart return component in hand, it’s easy to calculate the average return and volatility using the Excel functions AVERAGE and STDEV, respectively.
Let’s quickly test this on a simple book-to-market equity (B/M) value smart beta strategy made famous by Eugene Fama and Kenneth French. We’ll study the time period from July 1963 — the start date of the original Fama–French paper — through October 2015. Our low-risk-exposure smart beta strategy will be proxied by a value weight portfolio of the top 30th percentile B/M stocks, while our high-risk-exposure smart beta strategy will be proxied by an equal weight portfolio of the top 10th percentile B/M stocks. The CRSP value weight index will be our market proxy. (Results are the gross of the management fee and transaction costs for illustrative purposes only. All returns are total returns and include the reinvestment of distributions.)
The table below reports the main results. First, both smart beta strategies have a BETA close to 1. Second, the equal weight top 10th percentile B/M strategy has more than double the differentiated smart risk exposure (volatility) when compared to the value weight top 30th percentile B/M strategy.
It also has more than double the average return, making both strategies roughly similar on a return-per-unit risk basis. Thus, if the equal weight top 10th percentile B/M strategy charged about double the bps fee, it would still be comparable to the value weight top 30th percentile B/M strategy on a fee-per-unit basis.
The equal weight top 10th percentile B/M strategy is a better value — even if its headline bps fee is higher — as long as it isn’t more than double the fee charged by the value weight top 30th percentile B/M strategy.
Beef vs. Bun: A Real-World Application
In the real world, just how misleading is the headline bps fee as a measure of value in a fund context? To what extent are there low-headline-fee smart beta managers with high fees per unit of differentiated smart risk exposure?
We now have the tools to answer this question. To provide some insight into a specific situation, let’s analyze the returns and fees of two real funds, the first of which, we’ll call it Fund A, sports a management fee of 20 bps. The construction of this portfolio is analogous to the VW Top 30th B/M portfolio outlined above, with 691 holdings and a value weight construction.
Next, we’ll look at the more expensive Fund B, which has a management fee of 35 bps management fee, holds 116 stocks, and has a “style-attractiveness-weighting” construction (i.e., the stocks with stronger value characteristics get larger weights). Since Fund B’s portfolio construction focuses on the stocks with the strongest value characteristics and avoids a size bias, it is similar in spirit to the EW Top 10th B/M portfolio.
Both of these strategies aim to capture the so-called value premium, which is the historically documented spread in returns between cheap stocks (e.g., high B/M) and expensive stocks (e.g., low B/M). In the table below, we outline the characteristics of the two funds under consideration.
We analyze the live, net-of-fee performance of these two funds over the last three years — 1 January 2013 through 31 December 2015. All returns are total returns and include the reinvestment of distributions. Data are from Bloomberg and publicly available sources. The figure below tabulates the results.
While Fund A is the “cheapest” exposure with a 20 bps fee, one is actually overpaying for the differentiated “value” piece the strategy is trying to deliver, since the “patty” is so small (the monthly risk exposure is 0.63%). In contrast, Fund B is more “expensive” on a headline basis at 35 bps, but there is a lot more beef associated with this hamburger — the monthly risk exposure is 1.48%.
On a fee-per-unit basis, Fund B is cheaper than Fund A — 0.24 is less than 0.32. Fund B offers more than double the risk exposure but charges less than double the fee. Investment professionals who focus on the absolute price of investment without considering the product will surely be surprised when they see this.
An important caveat is that we implicitly assumed both products were of similar quality, which is hard to verify without large amounts of historical data. In contrast to volatility estimates, expected returns are notoriously difficult to assess. During the time period studied here, Fund B realized a slightly better return-per-unit risk than Fund A, but both numbers were negative — value strategies underperformed during the past three years.
Remember, we had over 50 years of data for our Fama–French example, making it easier to evaluate the “similar quality” assumption. Furthermore, the academic research also makes clear that fees are always an important consideration for long-term performance. Nonetheless, the example highlights that the headline price for a fund is not the only thing we should consider when purchasing a fund. Portfolio construction elements, such as value weight versus equal weight versus style attractiveness, are also important since they impact the size of the “smart” part of a fund.
So the next time you consider whether to spend on a smart beta fund that charges 0.25% versus one that charges 0.50%, don’t immediately buy the cheapest option. Instead, ask, “Where’s the beef?!” Otherwise, you might end up overpaying for a lot of bun.
This article was co-written with Peter Hecht at Evanston Capital Management.
For readers looking for a more in-depth analysis on the same subject, see the Alpha Architect post on “How to Pick Smart Beta ETFs.“
If you liked this post, don’t forget to subscribe to the Enterprising Investor.
All posts are the opinion of the author. As such, they should not be construed as investment advice, nor do the opinions expressed necessarily reflect the views of CFA Institute or the author’s employer.
Image credit: ©iStockphoto.com/ericb007