“Smart beta” has become a buzzword in investing circles, especially among pension funds and other institutional investors. The term may be new, but the idea isn’t: it’s about looking for ways to capture the returns of an asset class with a strategy other than traditional cap-weighting. These alternatives include fundamental indexing, equal-weighted indexes, low-volatility strategies and a few more exotic techniques.
A growing body of evidence has highlighted the inherent flaws in cap-weighted indexes, which are undeniable. By their nature, cap-weighted indexes give the most influence to the largest companies, as well as any that happen to be overvalued. That’s a potential problem because these are companies that are most likely to underperform the broad market over long periods.
A second potential problem with cap-weighted indexes is concentration. This isn’t an issue in huge markets like the US, or in a multi-country index like the MSCI EAFE. But it’s a concern in small countries like Canada (where Nortel once represented about a third of our entire stock market) and in individual sector funds.
Dunce caps?
These flaws are real, but if you listen to the chatter of investment firms these days you’d think cap-weighting is a strategy for dunces. This spring the Cass Business School of City University London published a pair of research papers comparing traditional indexing to 13 alternatives. The results were jaw-dropping. Every single one of the alternative strategies produced better backtested results in U.S. markets from 1968 through 2011. During this period, a cap-weighted strategy delivered an annualized return of 9.4%, while all the others came in between 9.8% and 11.5%.
It gets worse. The papers also presented a simulation involving randomly generated portfolios of 1,000 stocks each, which were then equally weighted. “Effectively we programmed the computer to simulate the stock picking abilities of ten million monkeys,” the researchers wrote. And guess what? “Nearly every monkey beats the performance of the market-cap index.” Remember that old joke about active managers losing to monkeys throwing darts at the stock pages? Apparently the monkeys are beating the index funds, too.
I’ve been getting a steady stream of email from readers asking what I think of “smart beta” strategies. They want to know if it’s time to toss aside the cap-weighted Couch Potato portfolios and embrace these new alternatives? I have no plans to do so, and in the next couple of posts I’ll explain why.
No magic in the results
Let’s start by addressing the most obvious question: how is it possible that all of the alternative indexes in the Cass papers beat the cap-weighted benchmark? The answer is actually straightforward: the strategies all tap into “factor premiums” that have been known for decades. That is, they give greater weight to small-cap, value or low volatility stocks, all of which have tended to outperform over long periods.
Fundamental indexes have a built-in value tilt, since they give added weight to companies with low price-to-book ratios, higher dividend yields or other traditional value factors. Equal-weighted indexes give added influence to smaller firms. Other methodologies in the Cass backtests were based on concepts similar to those used by low-volatility ETFs that have appeared during the last year or so.
So there’s no magic here, and there’s nothing wrong with looking for added exposure to one or more of these factors. I’ve often written about Dimensional Fund Advisors, whose funds are designed to capture the value and small-cap premiums. If you understand the risk factors, you’re willing to pay the higher costs, and you have the discipline to stick with the strategies during their inevitable periods of underperformance, I won’t try to talk you out of it.
But for the majority of DIY investors, I’ll continue to recommend simple portfolios with low-cost, cap-weighted index funds. And that’s the way I invest the my own savings. In my next post, I’ll look at whether “smart beta” strategies are really as appealing as they appear on the surface.
A very timely article as I am interested in some of the alternatives indexes. There is also a video of an interview with the author and a reporter :
http://www.cassknowledge.com/video-podcast/video/episode-155-monkeys-beat-market-cap-indices-cass-knowledge
I am still digesting the information by reading the original papers.
What I am taking away from this is the following:
1) An index strategy is still better than an active strategy.
2) The real return is still index return minus fees, so costs still matter
However, I am still pleased to see research being done on indexing in general. This represents the next stage in the evolution of passive management.
Just my 2 cents.
rob…
An article I read this weekend pointed out that “low-volatility” stocks currently have much higher valuations than the rest of the market. There is some overlap between those and dividend-paying stocks which pushes up the prices. If it had any value in the past it’s probably too late to get in on it.
@Robert_M: “This represents the next stage in the evolution of passive management.” I would have to disagree with that. Alternative indexing, or smart beta, or whatever you want to call it, is the next stage in the evolution of active management.
While stock picking is still common among retail investors and mutual fund managers, it’s not appropriate for institutional investors because of their scale. Many institutional managers are looking for ways to capture the returns of entire asset classes, but they want to use something other than cap-weighted index strategies so they can deliver higher risk-adjusted returns. That seems to be what’s driving this research (and marketing).
OK, so when these other methods are back-tested they do well, not surprising, since they’ve been optimised based on past events… but how have they done since then, in 2012 and the first half of 2013 ?
I have been considering adding low volatility funds to make up about 1/3 of the equity side of my portfolio (which right now is split in 1/3rds between canada, US, and euro/EM). This would not be for the benefits of low volatility but purely for sector diversification, to decrease weighting in financials and energy and to increase staples, pharma, telco, etc. Any feedback would be appreciated. The funds would be ZLB, VIG (not low vol. but sector diversication is nice, and the US low vol is almost 1/3 utilities which would be crazy) and XMI.
Dan offers wise words of caution regarding alternative indexing portfolio strategies!
I would agree that this represents an evolution in active management. It is what has also been referred to as “tilt investing”: moving in and out of currently favoured market factors to gain outperformance. “Tilting” into alternative indexes begs the question of how and when to “tilt”! Market timing becomes critical to gain the most outperformance and so it requires much research or a clever advisor to hire to manage the complexity. For passive investors who hold simplicity and cost in high regard then sticking to Dan’s model portfolios is wiser for those focused on goal directed investing rather than the reckless quest for outperformance!
See:
https://www.spdrs.com/library-content/public/A%20Case%20For%20Tilt%20Investing.pdf
@CPP: O.K. Now I am confused. If an index is based on objective criteria that can be easily verified and is audit-able, than it would be classified as a passive strategy. Since these alternative strategies are based on objective and measurable criteria, then wouldn’t they be an extension of passive investing rather than active?
I am having a hard time categorizing a strategy as active or passive. If the strategy can be automated without human intervention, than I consider it passive. If a human has to make a decision within the overall strategy, than it is active.
That is essentially why I lumped the alternative indexes into the passive camp. I was under the impression that no human intervention (or analysis) was required to generate the index.
So what did I miss?
cheers, rob…
@Robert_M: The difference is that an alternative index is designed specifically to deviate from the market in some particular way. Cap-weighted indexes also have rules, of course, but they have far fewer, and these are there for practical reasons: it would be impossible to invest in any index unless it had some screens for liquidity, for example.
At some point this is semantics: I like to think of active versus passive as a continuum, rather than two black and white groups. A fundamental index is clearly a lot less active than a stock-picking fund manager. But it still assumes an active decision to exclude some stocks and include others based on decisions imposed by humans who are trying to outperform the market.
What makes a cap-weighted portfolio unique is that it’s the only one that could, in theory at least, by held by all investors at all times.
I have a friend who’s been asking me a lot about fundamental indexing recently and my response is always twofold: 1) Like you said, it sounds mostly like a tilt towards small and especially value, and 2) If they really found a “secret sauce” for beating the market, that secret sauce will be useless in very short time with all of the exposure it’s getting. I personally stick to total market funds and am very comfortable with the decision.
@Robert_M: an index that only held stocks with prices between $30 and $40 would also be objective, and measurable :) You could call it passive since it’s an automated rule but it wouldn’t be particularly smart.
A passive index as most people understand it is usually the average of all other investors’ holdings within certain boundaries. By owning a minimal-cost passive index you can’t under-perform the average investor. If it’s as broad as possible you aren’t affected by any individual stock.
The main reason to own it is the belief that most stock prices are correct most of the time and most errors will cancel out, so trying to focus on certain stocks that are underpriced would be a waste of time.
Another thing that makes a cap-weighted index passive is that you don’t need to do any trading as the weights change. So a cap-weighted passive index is the ultimate buy-and-hold tool.
You raise a good question since you need to ask yourself what market you are trying to get an index for. The criteria above are a good guide for me. A low-volatility index is still an index of a specific type of stocks. But it assumes that they are usually underpriced, otherwise no one would care. It also limits the number of companies you own and your exposure to various sectors which means you are less diversified.
In a similar way someone who buys a value stock index is assuming they are underpriced. They probably do get more diversification there since I don’t think value stocks are concentrated in any specific sectors. Small-cap indexes are in part a bet on underpricing as well. But they also have a historically low correlation with other indexes.
That can be useful since a reliable portfolio should be built with indexes that have good performance on their own and are not correlated with the other indexes in the portfolio to keep it diversified. The common indexes do a good job of this and most attempts to come up with an improved index have failed because of their returns, correlations, fees, or other difficulties in actually following the plan.
@CCP and @Richard,
Thanks for the explanation. Being a former technology geek, I tend to think of things as black and white, especially when the costs of a 2.5% MER actively managed portfolio is compared against a 0.37% MER ETF portfolio.
I am very happy with my move to the passive “dark side”. The only one shedding a tear is my “soon to be ex” financial advisor.
I would like to learn more about the different indexes in general. Is there a goto reference on the subject?
cheers,
rob…
@Robert_M: Probably the best place to look for articles on alternative indexing is Index Universe: http://www.indexuniverse.com.
It’s interesting to read all these alternative indexes that beat the traditional cap-weighted ones. But I suspect once people start jumping into these alternative indexes, their premium will disappear, will start to lag the traditional index.
@CCP,
XIC (ZCN) of the simple Global Portfolio has approx. 230+ holdings on the TSX. Is this sufficient diversification away from the Large Caps? Do you think XIC (ZCN) should be substituted from the Global Portfolio in light of this information?
@CD: I think a broad market ETF for Canadian equities is just fine. Some people may want to split their holding to get more small-cap exposure, but that just introduces other problems: namely, small caps in Canada are dominated by energy and mining.
This is one of the problems you encounter when you use alternative indexes: you often trade one flaw for another.
@CPP: I read you current article about smart beta on Moneysense, and have a question I thought I’d ask here. My understanding about smart beta products is that they get their higher returns from accessing known factors such as small and value, but costs are high because of the higher degree of activity, and that the same higher returns can be accessed more efficiently with cap weighted small and value funds. In other words, for a value fund, the stocks are screened for P/B etc, then cap weighted in the fund. Is this correct? I remember reading on Bogleheads a while back, Larry Swedroe commented that his definition of active management is anything other than cap weighting Thanks..
@Tristan: You’ve definitely go the right idea, but there there is such huge variation in factor-based funds that it’s hard to make generalizations. The small-cap factor, by its nature is easily obtained with a cap-weighted index fund. But with value (or many other factors) how do rely only on cap-weighing? Do you simply start with the whole universe of stocks, declare them either “value” or “not value” and then cap-weight the former? Doesn’t it make more sense to weight companies according to their relative value? You also need to decide which specific value criteria you are going to use (P/E, book value, dividend yield, etc.). So it’s not a simple question.
Regarding the idea that anything other than cap-weighting is active management, I don’t think it’s helpful. Indexing works because of low cost, broad diversification and discipline rebalancing, not because cap-weighting is a magic formula.
@CCP: I looked up how the CRSP indexes are constructed, the index that many Vanguard cap weighted ETFs track. From what I understand they exclude the half of the market that is non value, then weight the value half by market cap. So a large value ETF would take large cap stocks, exclude the non value half then weight the value half by market cap. They use four value metrics, but I am not sure what weight they place on each one. I think the advantage of this approach is that it is inexpensive with minimal turnover so less tax drag. Other methods using fundamental weighting are accessing the same factors, but are more active with higher turnover so tend to have higher costs. I like the idea of cap weighting as we are letting the market decide how much of a given. market should be included in an index, and it also comes with likely the lowest cost which, as you say, is an important part of why indexing works.