Burton Malkiel has a new piece up on Wealthfront’s website calling smart beta strategies into question. It’s a very interesting and hot debate at the moment, and Malkiel’s perspective is worth a read—many of his points are important to consider.

The problem with this debate is that it is impossible to evaluate “smart beta” as an entire category relative to cap-weighted indexes, because smart beta strategies can be very different from one another. For example, a low volatility smart beta strategy will provide very different returns than a high momentum strategy. Any comparison between cap-weighted indexes and smart beta should be made on a strategy by strategy basis.

I think smart beta strategies should be assessed just like any strategy, and here I’ll lay out the criteria by which I think investment strategies should be evaluated. Hopefully you can use these criteria when making a decision between a smart beta strategy—or a traditional active strategy—and a traditional cap-weighted index.

Passive Investing—Misnamed, Misunderstood, and Flawed

First off, the debate between “passive” and “active” is misleading. As Cullen Roche recently pointed out in a post, everyone is active to some extent. Every strategy is an active bet that certain stock selection and weighting criteria are better than others.

Cap-weighting is itself a strategy that says “buy big stocks, the bigger the better.” Cap-weighted indexes are hard to beat because they are so cheap (and the fee differential between index funds and other strategies compounds over time) and the fact that rules-based cap-weighted indexes are inherently disciplined…they never deviate from their strategy. The goal of Smart Beta indexes/ETFs is to use the same discipline, but build indexes with better rules.

Malkiel’s best point is that costs (higher fees, taxes, and turnover) make some smart beta strategies inferior. But the piece, which denounces smart beta as a marketing tactic, is pretty light on empirical evidence. In a bit of cherry picking, Malkiel invokes the example of RAFI indexes taking on more risk in 2009 by overweighting bank stocks (which worked for RAFI), but leaves out the similar risky bets taken by cap-weighted indexes, like the massive weight to tech stocks in 2000 (which didn’t work for the S&P 500). Oh, and RAFI would have worked in the tech wreck, too, because fundamental indexes would have been very underweight the tech stocks that were producing very little sales/earnings/cash flows/dividends.

Evaluating Any Investment Strategy

Given that any investment strategy is active to some extent, I think there is a small list of important criteria when choosing where to put your money—whether it is indexing, smart beta/indexing, or traditional active. There are four main criteria worth considering.

#1 – Strategy consistency

Is the strategy rules based (and therefore consistent) or not? Consistency is very important for long-term success. Indexes—and Smart Beta—have this in their favor. The S&P 500 “strategy” has never changed, so it has consistency baked in. Non-index managers are less consistent. According to this recent SPIVA report, over the past five years, just 46% of funds stayed consistent to their style. Switching strategies/styles at the right time is very hard to do.

Bottom line: consistent approaches are better.

#2 – Selection Criteria

Rules-based is good, but the rules themselves are very important too.

I spend a few chapters on this topic in my book, but the bottom line is that market cap is a convenient but sub-par weighting/selection criteria. Indexing is the right strategy for a lot of people and is a great basic way to own equities, but I also believe that the right kind of “smart” strategy will deliver superior results. Here’s one example why.

Imagine breaking up all of the investable stocks[i] in the U.S. market up into 10 equal groups (deciles) based on their market capitalizations. Decile 1 would have the 10% of companies with the biggest market caps, and decile 10 those with the smallest. Today, there are around 300 companies in each decile, and 3,000 investable stocks total. Then imagine doing the same thing every year back in time for the past 50 years.

The figure below illustrates how each size decile has performed relative to the equal weighted opportunity set. Not much information in the 8 smallest deciles, but the largest stocks have significantly lagged the equal-weighted market. In particular, the biggest group of stocks has lagged the equal weighted opportunity set by almost 2% per year. Here’s the problem: 78% of the weight of the S&P 500 is in this largest decile today. The top 50 stocks by market cap have done even worse historically.

In contrast to market cap, there are a host of other rules that have worked for selecting stocks. A brief list includes: value, momentum, low volatility, quality (earnings quality, profitability), and small cap. The effectiveness of these factors varies. Below is the value example I’ve used before: value is much more powerful than small cap.

Malkiel is right that these factors (on their own) can have long periods of underperformance: momentum did poorly in the 2000’s because of the ’09 momentum reversal. But typically, as the holding period lengthens, the “risk” of losing to the market fades. Here is the percentage of time that cheap large stocks (using earning/price) outperform all large stocks over the past 50 years. These factors (and any smart beta strategies using them) only work if you stick with them.

Smarter selection criteria have been proven to work over very long periods of time, arguably because these factors identify stocks that are systematically mispriced by the market; and the market is comprised of people who make systematic behavioral errors. Value and Momentum are the two that work the best. We’ve known about each “anomaly” for a long, long time, and yet they’ve continued to deliver outsized returns.

Bottom line: there are smarter ways than market cap to build a portfolio (before costs, we’ll get to those in a moment).

#3—Concentration

I loved Howard Mark’s latest shareholder letter, which is essentially an ode to “active share.” If your goal is to beat cap-weighted indexes, then your odds go up if you are very different from the index that you are trying to beat. There’s really no reason to be an “active” investor if you are just going to hug/slightly tilt an index.

Of course being different, or “daring to be great” as Mark’s calls it, brings the highest odds of ruin, too. There are smart ways to be different and dumb ways. The very long-term historical success of certain factors suggest that cheap stocks with good momentum is a smart way, and building a unique portfolio using value and momentum is the best way to “dare to be great.”

One of the main arguments against “active management” is that the average active manager loses to index funds—but this is like saying the Pope is Catholic or bears crap in the woods. Because of fees charged, active management is a negative sum game. The entire group of investors will earn the market rate of return, and the average will be negatively offset by active management fees that are higher than index fund fees. As this paper on active share demonstrated, the group of funds with the most unique holdings (highest active share) has historically beat market-cap weighted indexes by 1.1% after costs/fees. It has paid to be different than the market.

Bottom line: if you don’t buy an index, you’ll want to focus on strategies that have unique holdings and high active share.

#4—Implementation Costs (Trading (Turnover), Taxes, and Fees)

Turnover and taxes are hard to evaluate. Broadly speaking, the best strategies will have holding periods long enough that most gains are taxed as long-term capital gains. Trading/turnover get a bad rap, but turnover (and the taxes it brings) is really just the flexibility to move into more attractive opportunities. No turnover means you never adapt to changing market opportunities. I think the best strategies balance trading costs and turnover with the need to buy newly attractive stocks (e.g. undervalued, good momentum) and sell stocks that look less attractive (e.g. overvalued, bad momentum).

Fees are most straightforward category. Lower is better, and here indexes have a massive advantage. But it will thin out in the years to come. Smart beta fees (and traditional active fees) will keep coming down. Lots of smart beta strategies are in the 0.50% range right now, which is a disadvantage versus the 0.07%-0.09% range for index funds, but many smart beta strategies have historically delivered much more than 0.41% outperformance.

Bottom line: any strategies edge (its potential outperformance) needs to be greater than the total cost differential versus a cap-weighted index. The best deciles of value and momentum stocks have outperformed the market by between 4%-6% annualized since the early 1960s[ii]. That is more than enough to overcome cost differentials. Will these excess return number persist over the next 10-20 years? Of course it is impossible to say. But these factors (and others) sync nicely with the latest research in the behavioral economics field, have long and proven track records, and have continued to work despite widespread awareness of their existence and effectiveness.

Summing It Up

The funny thing about smart beta is that it’s gone by the name “quantitative investing” for two decades. Big names like Josef Lakonishok and Cliff Asness were writing papers about value and momentum in the early 90’s, and the first edition of What Works on Wall Street (which chronicles many “smart beta” factors) came out in 1996.

There are many successful quant firms that have posted consistent, market-beating results net of costs. Many of these firms are very well known, have great strategies, and manage taxes and turnover as best they can.

Cap weighted indexes do have big advantages. A consistent strategy, low taxes, low trading costs, and low fees is a great combination. When looking for strategies other than cap-weighted indexes, look for ones that are rules based (consistent), use rules/factors that have worked well over long periods of history (like value and momentum), are different from the market, balance taxes and turnover with expected outperformance, and charge reasonable fees.

If you are investing in the stock market for long term success, then smart beta strategies with the right strategies can be a great alternative to cap-weighted benchmarks, but each should be evaluated on its own. So far, the leaders in the online financial advisory space (especially Wealthfront) have avoided smart beta. I will be very curious to see if it lasts.

[i] Market Cap above $200MM, then inflation adjusted back in time

[ii] Complete stats for these any many other factors available in the latest edition of What Works on Wall Street