A Primer on Smart Beta

With $42.8B of inflows in 2013—up from $20.5 in 2012[i]—“smart beta” products have gone mainstream. The debate between traditional market-capitalization weighted indexes and “smart” strategies or indexes is an important one because the choice between the two can lead to significant differences in returns over long holding periods. Smart beta simply means a strategy that chooses stocks based on some proven criteria other than size or market capitalization (e.g. value, momentum, etc.). In this brief primer, I’ll summarize the problem with market cap weighted benchmarks, review what to look for in a smart beta strategy, and address a common criticism that smart beta strategies are becoming over-popular.

Stupid Beta

Smart beta strategies have emerged because buying and weighted stocks according to their size is stupid. Instead of a cap-weighted benchmark like the S&P 500, I think the fairest benchmark for investing success is an equal weighted group of investable stocks (at least $200MM market cap, adjusted for inflation). This is the fairest benchmark because this is your opportunity set—if you can’t beat this benchmark, you aren’t adding any real value.  The historical return of this equal-weighted benchmark is 11.7% (1962-2013). Today, there are about 3,500 investable stocks here in the U.S.  If you split this investable universe into ten equal buckets based on their size (so roughly 350 stocks in each), one major effect sticks out: the biggest stocks (the 10% of stocks with the largest market caps) grew by 9.6% historically (1962-2013), underperforming our equal-weighted benchmark by 2.1% per year. That is a huge gap. Over a 20-year investment period, that 2.1% per year would balloon to almost 300% because of the power of compounding. So here is the problem: 78% of the S&P500 currently falls in this largest-10% bucket.

For any basketball fans, I find the following analogy useful.  If you “salary” weighted an NBA all-star team (which would be like market-cap weighted), this would be your starting five: Kobe Bryant’s upper body, Dirk Nowitzki, Amar’e Stoudemire, Joe Johnson and Carmelo Anthony.  If you “value” weighted an NBA all start team (I use PER as the best measure of player ability), then your starting five would be Kevin Durant, LeBron James, Kevin Love, Anthony Davis and Chris Paul.  A game between these two teams would be a bloodbath. Bigger is not always better (well, at least when it comes to salaries and market caps).

Value is “smart,” what else is?

One way of buying stocks that makes much more sense than size is value. No matter how you define value (price-to-book, price-to-earnings, price-to-sales, EBITDA/enterprise value), when you pay less for stocks you earn more from them.  Value investing is easier said than done—because all the fun stuff tends to be expensive (Facebook, Tesla, etc.)—but it works.  Some argue that the value effect is mostly a small-cap phenomenon, but it works in large stocks too. Between 1963 and 2013, the cheapest 10% of large stocks outperformed an equal-weighted benchmark by 4.1% per year[ii]. The figure below gives a few examples of strategies that have worked well in the large cap space, both before and after the average industry fee.

Compustat raw data used.

Compustat raw data used. “500” means the top 500 stocks by market cap only.

Other strategies work as well.  High momentum stocks outperform, low volatility stocks outperform (to a lesser degree, but they do very well on the downside), highly profitable stocks outperform (return on capital or lower accruals, often referred to as quality), and stocks that send more cash back to their shareholders outperform. CXO Advisory is a great resource if you want to explore factors further. Of course, this is all in the past, and there is no guarantee that they will always work.  But arguably these factors are rooted in human nature—mispricings exist because we are boundedly rational and commit systematic trading mistakes (like over or underreacting to new information, or extrapolating recent trends too far into the future).  The good news is human nature isn’t changing any time soon, so these strategies should continue to be effective in the future.  And, of course, the alternative is to believe that more expensive, poorer quality stocks that send less cash (or no cash) back to shareholders and have worse recent market momentum will do well.

How to Evaluate

When choosing between the countless smart beta ETF options now available, there are just a few things that matter:

1)      Fees – It goes without saying that you want to spend as little as possible. The market-cap weighted SPY charges 0.09%…that’s $90 for every $100,000 invested—an amazing deal. Smart Beta products charge more, so the question is how much more should you pay for a superior strategy.  Which brings us to…

2)      Strategy—There are many smart beta ETFs that use one factor (e.g. value, momentum, low volatility). You should look instead for ETFs that use a variety of different factors.  This is because while these factors work well on their own, they work even better together.  Take value for example. If you were a value investor, meaning you only bought stocks in the cheapest 20% of the market, you’d have done well in the past.  But if you had further differentiated those value stocks based on their momentum, you’d have done even better.  Here is a universe of value stocks split up by their momentum. 

Annualized returns. Compustat raw data used.

Annualized returns. Compustat raw data used.

You can see, it’s much better to be “smart” in two ways than in one…and three ways are better than two, and so on.

3)      Concentration—You should favor ETFs (with a good strategy) that have fewer holdings and take bigger bets. For ample evidence that you must be very different from the cap-weighted benchmark to beat the benchmark, read this paper by Cremers and Petajisto. If a smart beta ETF is just tilting an existing cap-weighted benchmark, it may not be worth the higher fees. Look for ones that have no more than a few hundred holdings.

If an ETF strategy uses several factors and is concentrated, then higher fees can be justified.  It’ll be hard to find a perfect mix of these three, but look for ~50-200 holdings, a combination of factors, and fees between 0.15%-0.75%.

Criticisms

Given the asset flows into smart beta strategies, it’s no surprise that many are criticizing them as the same old stock market returns in shining new (and more expensive) packaging.  Others are concerned that their popularity will reduce their effectiveness, because as more people crowd into value stocks, they’ll be bid up these cheap stocks, neglect the expensive stocks, and the edge will disappear. One way to evaluate the “opportunity” in value stocks is to look at the spread or gap between valuations of cheap stocks vs. expensive stocks, seen in the figure below.  If the value trade is crowded, you’d expect to see the spread compress.  Indeed using price-to-sales, we are near the all-time low spread, so this is a potentially legitimate concern. No surprise, the “peaks” of this chart were some of the best valuation opportunities ever (1973, 1981, 1991, 2009).

Price-to-Sales ratio of the stock in the 10th percentile vs. stock in the 90th percentile across all investable stocks. Compustat raw data used.

Price-to-Sales ratio of the stock in the 10th percentile vs. stock in the 90th percentile across all investable stocks. Compustat raw data used.

But this concern highlights why it is important to use several different factors rather than piling into just one like value alone. It also highlights that If you elect to move your money into smart beta strategies, you must understand that it is not a trade, it is a long-term strategy.  The reason that these factors have continued to work—despite widespread knowledge of their efficacy since the 90’s—is that it is very hard to stick with them when they are sucking. Every smart beta factor can and will have long (5-year+) periods of underperformance. In a world of quarterly performance evaluations, it’s hard to stick with something after 5 years of bad returns. After these periods of bad returns for certain ETFs, people will probably abandon them—at exactly the wrong time.  Peter Lynch said the key to making money in stocks is not getting scared out of them.  The same is true for smart beta.  It may a bit of a fad, but the philosophy behind the best smart beta ETFS—that value, momentum, quality, and yield lead to superior returns—is sound. You can do well in them so long as have patience and discipline, two rare commodities in today’s market.

Ultimately smart beta offers access strategies proven to be more effective than market-cap weighting. Setting aside what has worked in the past for a moment, imagine you had two investment options to choose between with the caveat being that after making your choice you weren’t allowed do anything for 20-30 years. One is the SPY—a cheap, but cap-weighted index fund.  The other is a group of stocks that are cheaper than the market, are more profitable, send more cash back to their stakeholders, and have strong recent price trends.  Which group would you prefer to own for twenty years? I know which strategy I would pick.

 

[i] http://online.wsj.com/news/articles/SB10001424052702304680904579369112512612596

[ii] Large stocks defined as anything larger than average (current floor would be $7B).  Value defined using a variety of value factors like price-to-earnings, price-to-sales, etc.  Using data from Compustat.