The Problem With Designer ETFs

The exchange traded fund [ETF] world has blossomed with the introduction of designer exchange traded funds [ETFs] that apportion their holdings according to some statistical process other than mere market cap. There are ETFs from Wisdom Tree that apportion their holdings according to total dividend payments.

Along similar lines the SPDR Dividend ETF (SDY) invests in the 50 highest-yielding companies in the S&P Composite 1500 that have consistently raised their payouts for at least 25 years. On the other hand, the PowerShares FTSE RAFI US 1000 (PRF) evaluates the 1,000 biggest U.S. companies on the basis of average sales and cash flow over the previous five years, total dividends distributed over the previous five years and latest book value (assets minus liabilities) and apportions its holdings according to a formula based on these factors.

There is a basic philosophical problem with all of these funds. Apportioning according to market cap mimics and therefore accepts the actual distribution of money in the market. This distribution, just like the price of every stock, is that which those with the most money invested have made it. Those with the most money invested, unsurprisingly, have typically hired the best minds, with the most time and the best training to make the investment decisions that have resulted in the relative market caps of the various stocks. Using a statistical guide for the distribution of money into stocks suggests that this process has been inefficient.

The designer ETFs take a subset of the information that has been analyzed by the investors who have set the distribution of market caps, in an ever changing dynamic process. They then take this subset of available information and apply a rigid formula to it. And they appear to believe that this distribution, based on a subset of information and devoid of an ongoing application of human judgment can beat a process based on the full set of information and constantly informed and updated by skilled human judgment.

I don’t buy it.

Now I do swear to you, the reader, that when I began to write this journal entry I had not checked the performance of SDY or PRF. I picked them at random from a short list of designer ETFs. I have just checked performance and over the past two years they have both been beaten by the S&P 500 index. PRF was beaten by about 5%, whereas SDY was absolutely clobbered by roughly 20%. I did not take out the management fees of SDY or PRF, but also did not add in the dividends one would receive if holding the S&P 500. Clearly this is a quick analysis based on only two randomly selected designer ETFs. Nevertheless it tends to bear out my point. You can’t outsmart the market by basing your distribution of money between stocks on a rigid computer analysis of part of the data. It just does not make sense.

Now we all know that in March of 2000 the largest of the large cap stocks were overvalued. Since that time an S&P 500 index in which each stock was equally weighted has far outperformed the actual S&P 500 index in which each stock is weighted according to its market cap. But the designers of the designer index funds are not the only ones who have noticed this phenomenon. There is every possibility that the current market caps in the S&P 500 have been strongly influenced by this past overvaluation. Perhaps the market has actually overreacted. It is possible that the present valuations actually undervalue the largest large cap stocks. But let us theorize that there actually is some attribute of investor psychology which causes the largest large cap stocks to be consistently overvalued. If this were to be the case, surely a better way of addressing this issue would be to apply a secondary weighting to the market cap weighting, which would cause the largest large cap stocks to be weighted slightly lower than they would be in a straight market cap rating. I’m not recommending it, except for as a more sensible alternative to ranking by dividend payments or any other subset of the available financial information.

Market cap is a function of democracy, except for resulting from a process in which each dollar, rather than each eligible person, gets a vote. So Winston Churchill’s famous aphorism about democracy, that it is absolutely the worst system of government, “except for all the rest,” holds true in this context.

Comments

[Smart ETF]

You make an argument for market cap weighting but I question several of your assumptions: 1) that Wall Street hires the best minds, 2) market distributions are accurate, 3) and that market cap represents the market. I’ve met thousands of investment professionals and most follow the heard and few do their own homework. For example, look at the bulk of the asset allocation models that use off the shelf models, like Ibbotson. These models are ridiculously flawed; in fact the founder of CAPM, Bill Sharpe, even admits that. Just because everyone is making a bet that history will forecast the future (based on a regression to the mean) doesn’t make it right. This leads to point two, most distributions fall under the Normal Distribution category which by default ignores the fat-tails. But more to your point why is following the heard a good thing? Cap-weighting is a form of market-timing because it is momentum based, like it or not. Someone is over-paying for an asset because you have more buyers chasing an asset as the price is rising. Regardless of what you believe Market Cap is a strategy and who is to say which is right? I think Rob Arnott has more than proven his point. You discuss the statistical distribution of money but you are only looking at the effect and not the cause. Much more insight can be garnered by analyzing the distribution of institutional (block) vs. non-block trading volume. Doing so you soon realize you can have more buyers than sellers and yet the price can still fall if some of those sellers are big block institutions. Watch how often the big boys are selling into the strength of the smaller volume buyers; not a pretty sight. In other words, it’s the human judgment that you endorse (perhaps based off of investor sentiment) that leads to the over-bought or over-sold conditions. I have the most trouble with your comment ‘You can’t outsmart the market by basing your distribution of money between stocks on a rigid computer analysis of part of the data.’ I totally disagree; tell this to Simons, Tudor, Robertson or most of the quant hedge fund mangers. I build models that have out-performed for decades. Maybe you should have added: ‘…using traditional models like Modern Portfolio Theory’ or something to that sort. If it doesn’t make sense to you feel free to contact me and I’ll demonstrably spell it out. More confusing to me is that you defend market cap and then later in your article start making a case for a secondary weighting to adjust for human behavior (market sentiment); isn’t this contrary to your point? Market Cap is not a function of democracy; it is a greater fool theory. If it is a democracy it is one like most countries where a few fat cats pull the strings behind the curtains and influence the tide (institutional block money flow, or worse (if you’re into the conspiracy theorist thing)). My suggestion is to not drink the random-walk cool-aid. Maybe a nice book by Benoit Mandlebrot or Nassim Taleb or anyone else that can live outside the traditionalist will change your mind. Cheers.

[Panskeptic]

What a lot of flapdoodle!

"A process based on the full set of information and constantly informed and updated by skilled human judgment."

Is he kidding? A cuddlesome dogma, but demonstrably false. How about intellectual laziness, herd mentality and fear of independent thought? Is Mr. Siegel against the American Revolution because George III was anointed by God to be our leader? That orthodoxy makes just as much sense.

"A full set of information?" Then how come everybody was blindsided by Bear Stearns?
Yeah, right. That's the kind of thinking that got us into this mess.

"Market cap is a function of democracy?" What addlepated nonsense! Market cap is a function of market cap, and there's more than one path to stock market Nirvana. It's Mr. Siegel's kind of chanting blind, lemming-like sloganeering that gives economics a bad name.

This article belongs on a religious or archeological website, because on a scientific basis it's dead, over, obsolete, extinct, kaput.

[Dividend Growth Investor]

Interesting article. You might think that these ETF's mostly based off of a backtested index are just full of hot air. I do think,however that over the long-term companies that could increase their dividends consistently over time will do at least as well as the stock market averages.

[Timothy Siegel]

Smart ETF

Most of what you say, I agree with. But remember: I say that Market Cap is the worst system, except for all the rest (credit: Churchill). With respect to Simons, Tudor and Robertson, we really don't know how they beat the market, and to what extent they rely on weighting models, because they are extremely secretive. I accept that you construct models that beat the market for decades, but my guess is the market performance that you beat is pre-model construction. I can certainly accept that with a rearward view, you can construct models that beat the market for many decades of the past. But you are not the only one who is observing the past. Others will also respond, and this will tend to destroy those same relationships in the future. Also, my suggestion for weighting the market caps is highly qualified with, "let us theorize that there actually is some attribute ... ." I only suggest that as a superior alternative to weighting by other alternative weighting schemes that have been developed.

Also, with respect to my democracy statement, I state, "except for resulting from a process in which each dollar, rather than each eligible person, gets a vote." So of course it is one in which the big money players, or "fat cats," are weighted much more than the average investor.

My basic point is that every observation that you make about the market is made by others, too. So if you construct an index that takes these observations into account, your index must contend with other indexes and market actions of others who have made the same observations.

I would like to make one qualification, however. Rather than saying "the best minds" I should have said "the best minds that the big money players have identified." There could be way better minds out there, that have just not been identified by the big money players.

Also, if you constructed an index, decades ago, or even a single decade ago, that has consistently beaten the market since then, I would be very, very interested in seeing it and seeing how you did that.

[Timothy Siegel]

Panskeptic

"Flapdoodle"...

I just have one basic answer, and that is: We are talkin' money here. Yes, market cap is flawed, it is horribly flawed. It has all the flaws you list. But we are talkin' money and so there will be a constant, continuing financial incentive for all players to break away from dogma and to see the falsehoods and respond. And, Timbo's rule 54.28.3, states that "People respond to financial incentives."

Market cap is the worst, except for the rest.

[Timothy Siegel]

Dividend Growth Investor

I greatly appreciate that you did not call my posting "flapdoodle."... But, if I understand it correctly, I disagree with your statement that "companies that could... ." At any moment we only know what companies have done in the past, and so a company that in the past has consistently increased its dividend, might not in the future. If enough investors start to invest based on consistent dividend increases, those companies that are forced to refrain from a past pattern of dividend increases could really get clobbered when all the dividend increase investors sell en masse when a good dividend pattern is discontinued. Also, high yield stocks tend to be vulnerable to rising interest rates. My thought is that an investor is always better off looking at everything. The best way is to observe some relationship that others have missed. I just don't see how you can hope to beat the market by being part of some huge crowd of dividend growth investors, as it seems to me that this investment philosophy has really grown in popularity over the last 10 years, and might be cresting.

[falcon2382]

Timothy, I also have some objections to much of what you said, but I don't read these articles to waste my time arguing or proving that I am right, I do so in order to challenge what I take for granted and to share my thoughts with others willing to respond to comments so that we can become better investors, and as you have illustrated more than your fair share of tolerance, I want to thank you for your time and energy. But I also want to hear what you have to say about the following response to your article:

What interests me most about your thesis is how you seem to advocate for a more "human" managed selection of stocks than for a selection of stocks chosen by a computer algorithm while admitting that market cap itself is an emotionally charged way of valuing a company (all with which I totally agree). Yet, you seem to ignore the time-oriented part of the investing process that is, by its very nature, about human decision to allocate and or reallocate funds according to perceived changes in the market. I don’t think the ETF’s take away from this human element. The formulaic, non-human, objectivity of the ETF is precisely why it is a good investment tool—it is a “control” in a world that is more than uncertain at times. In the case of SDY, which I recently purchased at $43/share, the unchanging formula is what attracts me to it. I know what I can count on. The ETF is more heavily weighted in financials than any other sector and this is why it has greatly underperformed the greater market. However, if you share my prognostication for the market moving forward, I believe many other sectors will soon be catching up with financials. This judgment is the “human” part of my investing strategy, and since I am only 25, I have a time horizon that provides ample opportunity for this imbalance to work itself out—meanwhile the dividends will compound the total number of shares I own. Further, because of the “rigid” formula of the ETF, if a company can't afford to keep its dividend or has to cut it after 25 years of increases, then it gets booted automatically. This is good because something fundamental in the company’s health has changed and unlike humans, the computer algorithm can’t find ways to justify holding on to a loser that has “broken the rule.” At the same time, the computer program also doesn’t get scared and sell a good company at a loss simply because it was taken down with its sector. But the best part is that I didn’t have to pay a transaction fee to get rid of the loser, or to buy its replacement. Yet, if there are any capital gains I am able to receive them (again without a transaction cost). And on that note, how can you not consider the dividends and management fee structure in your analysis? It’s crucial! The AVERAGE mutual fund (which is generally not actively managed either) is over 1.2%. SDY only has a management fee of 0.35%.

Another point I’d like to address is your very comparison of SDY with the S&P 500 strictly from a capital depreciation standpoint. It was only recently that I chose to invest in SDY, and it was BECAUSE it was down so much in relation to the greater market. Or rather, it was because I realized that the ETF is more heavily weighted in financials than other broader market ETF's, but without the concentration of a strictly financial ETF, and without all the “bad eggs” (i.e. indymac, Washington Mutual, etc...). So after deciding that the financial companies that are included in SDY were among the healthiest in the industry, I welcomed the extra exposure, expecting that at some point the banks that weather this crisis will be handsomely rewarded, not only because they once again proved themselves (all of the Banks held by SDY made it through the S&L crisis and the fall of LTMC), but because they will gain market share simply by not going bankrupt. Moreover, if you want to make a comparison you should be aware of what the comparison is really telling you. What good is it to blindly compare the S&P 500 to any ETF if you don’t consider the underlying stocks? If, for example you had compared SDY to XLF you would have seen just why SDY is a compelling investment right now. At the time of this writing, the SDY boasts a dividend yield of 4.87% while XLF (a strictly financial ETF) is 4.55% (the S&P is only 1.67%). Thus, one might conclude that with SDY you get diversification with a huge dividend, exposure to the financials once they do bounce back, and peace of mind knowing that the fifty companies you own in this ETF have not only been around for over 25 years, not only paid a dividend for over 25 years, but have INCREASED that dividend during that time period. Again, I think it is relevant to remember the other crises we have been through during the past 25 years.

I don't think it is fair to talk about ETF's as if they all have the same purpose, function, and risk/reward. If that were the case then we wouldn't need different ETF's; SPY or DIA would serve everyone's needs. I have a long time horizon and I welcome the extra dividend that SDY provides over SPY as we work our way through the current mess. In fact, this is a very very important reason (there’s the “human” part again) to consider SDY since most bear markets turnover slowly and through a basing process. When we reach that point—and I think we are there or at least very close—I expect the compounded dividends to pay me better than both treasuries and the greater equity market as I wait for the economy to find support. The S&P contains a boatload of discretionary retailers and companies that earn their money from consumer services and products, leisure activities and is it that hard to see that the energy companies which have become quite bloated recently might also drag on the S&P. The future losses of these sectors over the next year or two (or three) won’t affect SDY nearly as much as the greater market at which point SDY, I believe, will seem like a golden egg. Any thoughts?

-Mike