If smart beta is smart, is my beta dumb?
First, a definition: smart beta strategies are (allegedly) better ways of getting exposure to equities, bonds and other asset classes than via traditional indices. What precisely does this mean?
Let’s assume my core equity allocation is an S&P 500 index fund. Smart beta guys think I have a problem: the index is weighted by market cap. Whenever Standard & Poor’s rebalances it, the weights of winners go up and losers down. In other words, after Google rose 45% in 2015, I buy more Google (ok, “Alphabet”); as Chesapeake Energy dropped 75% plus, I sell some. In essence, I’m a closet momentum investor.
This wasn’t done by accident. The Dow Jones Industrial Average used to be everyone’s proxy for the “market.” But few consider it representative today: only 30 stocks, weighted by the dollar price of one share (just plain weird), and a bias toward old economy companies (hence, “industrial”).
Compared to the DJIA, the S&P 500 looks pretty smart: 500 stocks, weighted by market cap and representative of a much broader swath of publicly traded equities. At the end of the day, my S&P 500 fund gets me broad exposure to equities with a touch of momentum. But is momentum bad? Probably not: ironically, one popular smart beta approach is a more concentrated form of momentum than I get from the S&P 500. Yes, the smart beta guys like it when they design it, but not when S&P does.
I am now cutting my flowers and watering my weeds.
So why use another weighting scheme? Many investors have a value bias and cringe at the idea of chasing a rising stock. Instead, we could weight each stock equally. With 500 stocks, I now own 0.2% of Google and 0.2% of Chesapeake Energy. Google rises, and I cut it back; Chesapeake Energy falls and I buy more. As Warren Buffett warns, though, I am now cutting my flowers and watering my weeds.
Given my value bias, why don’t I overweight “cheap” stocks? Great in theory, but tough in practice. What metrics should I use? The original work on value stocks was very straightforward: simply buy the stocks each year with the lowest price to book ratio. This would have outperformed during the 1960s through 1980s, but underperformed badly in recent years (please see our Value Factor Reconsidered paper). When I look at the actual shares of companies with low PB ratios, I find a lot of struggling companies like Chesapeake Energy. Plenty of stocks are cheap for a good reason (“cheap and built to stay that way”). Moreover, the price-to-book value screen results in weird outcomes. It loved Chesapeake when it was trading at $20 per share – a steep discount to book value at the time – but hated it at $3 per share after a massive write down had decimated its reported book value. Not sure I want to tie my fate to such a simplistic approach.
So what about some multiple of earnings or cash flow? Now we’re subtly moving away from an index to an investment strategy. The beauty of the S&P 500 is that we all agree what it is — warts and all. The issue with these smart beta products is that each practitioner puts a different twist on how to build his or her respective “index.” Use historical or expected earnings? Reported, adjusted or cash earnings? Rebalance annually or more frequently? Cap a particular sector if it gets too large? The list goes on. The more parameters, the more it looks and feels like active management than index investing. I don’t have to think very hard about whether to buy a Vanguard or iShares S&P 500 fund, but now I find myself making a call on who’s got the better quants.
There’s another big problem with this approach: data mining. The beauty of the S&P 500 is that we have live data going back decades. Indices or products created today tend to overstate how they would have done: trust me, you won’t see a smart beta product with back-filled underperformance. The cold reality is that it’s hard for researchers to resist turning the dials until the results looks good. Unfortunately, this isn’t likely to translate into better results going forward, which is all I care about.
So why all the brouhaha? For one thing, firms don’t make enough money selling S&P 500 funds – it’s too commoditized. Smart beta practitioners want investors to switch from the lowest cost, commoditized funds to “value added” indices where they can charge more money. While it’s a certainty that this is good for asset managers, the jury is still out on whether it’s good for investors.