Why do Value Investors Like to Index?

I think I had some good things to say in the last post, but one commenter disagreed, and he had some good things to say:

Unfortunately, the premise of this article is completely flawed, as it assumes all “indexes” are simply S&P 500 or Total Stock portfolios. You are no doubt aware that Vanguard has Large/Mid/Small VALUE indexes as well, right?

Further, for someone wanting more pure, targeted, and consistent exposure to the lowest priced value stocks, the “enhanced” index funds from DFA are close to unbeatable. Sorting on a simple metric of price/book and holding approximately the cheapest 25% (as opposed to 50% for Vanguard and most Value ETFs) of stocks in the respective asset class (while trading patiently, using fund cash-flows to rebalance, lending securities to earn additional revenue), DFA’s large/small value funds in the US, Int’l, and EM markets have trounced their active manager competition. Here are the stats on the % of active value funds in each asset class over the last 10 years (through November) that have been OUTperformed by DFAs simple “structured” approach, which for all intents and purposes are index funds:

US Large Value (DFLVX) = 90%
US Small Value (DFSVX) = 83%
Int’l Large Value (DFIVX) = 91%
Int’l Small Value (DISVX) = 100%
Emerging Mkts Value (DFEVX) = 97%

And not that it matters much, as these percentages are so high to begin with, but these #s don’t include survivorship bias — something on the order of 40% of value managers that existed 10 years ago have gone out of business, so this outperformance is only measured as a % of those professional value managers that survived!

Somewhere, over some periods, I am sure we can find some value managers who have outperformed an intelligently structured value index portfolio, but the numbers are so small as to be almost irrelevant, and there is no persistence going forward in the # who have been able to pull off the feat.

No, the case is actually quite clear, “active” value investing is dead. All investors would be much better off simply holding broadly diversified, structured/indexed VALUE portfolios. And stop confusing “indexing” with “cap weighted total market index portfolios”. There are a lot of index funds beyond the Russell 3000 and S&P 500.

His main point is well taken.  Active long-only value managers have not done well versus the indexers.  I’ve stated that at other times.  This is also true for hedge fund managers, where survivorship bias is even worse.

That said, I have a few objections.

1) Index investing by its nature follows the return factors incorporated into their index (if cap-weighted) or enhanced index (if not).  Factors go in and out of favor.  Some factors are seemingly permanently in favor, like value, small size, low Net Operating Assets, and price momentum.

Occasionally, those factors can be overinvested, like in August 2007.  That doesn’t mean the factors should be abandoned — weak holders are getting shaken out.

2) Every investment strategy has a “carrying capacity.”  Indexed strategies are larger, as are value strategies.  But there is some point where value as a whole can be overinvested.  Value can become “too cool” for a time, and can get relatively overvalued.  Some market participants look at the range of P/E, P/B, or P/S ratios.  When they are thin, value is overpriced.  It’s like being a bond manager, and doing an up-in-credit trade, except that this is an up-in-growth trade: buy higher growth stocks when the difference in P/Es  and other valuation factors is relatively small.

3) Yes, I know about the many subindexes that underlie the whole of indexing.  That wasn’t my point in the prior article.  Indexers need some amount of valuation oriented  investors, whether they are portfolio managers, or that they investors willing to take the whole company private, or a public company that acquires it.  If everyone indexed to the market as a whole, there would be no price signals.  Yes, with subindexes, that is not so, but the more money you pour into a subindex, the greater the likelihood of overvaluation.

4) There is the possibility of an indexing bubble.  An indexing bubble would have a situation where stocks in major indexes are overvalued relative to companies that are not in indexes.  now, it’s hard to imagine an indexing bubble, because there is no leverage involved, and little speculation.  Now for subindexes, relative over- and undervaluation is normal.

Just as in commodity markets, you have commodities that trade on futures markets, and end up in indexes, and those that don’t, because they are less liquid, fungible, deliverable, etc.  Often the relative price difference between what is easily tradable can be an indicator of whether excess liquidity has warped prices beyond their fair value.  This can happen with stocks as well, where stocks less held by indexes those more held by indexes.

There will probably come a time when those that have invested in index funds have to liquidate to meet their long term goals, and there will not be enough new money to absorb the selling.  Unless this is disproportionately true of index investors (unlikely, but possible), this would be a whole market phenomenon.

The broader question is related to the markets as a whole.  Since most stock investing is done on an unlevered basis, the overall ability to hold a diversified stock portfolio comes down to time preference.  There is what stock investors as a group should have as their time-preference, which is largely based off of demographics.  The there is how they behave when markets get hot (optimism -> time preference lengthens) or cold (pessimism -> time preference shortens).  Note that this is the opposite of the way that absolute value investors behave.

Now, here is one problem with my thesis: DFA and Vanguard are clever traders.  In really small stocks, DFA is virtually a market maker and picks up some alpha doing it.  In my investing, I actually like it when Vanguard or DFA is a large holder, because it is an indicator of neglect.

Here’s the second problem with my thesis: the fees of active managers are too high.  Even if active managers can pick up on some inefficiencies, will it be enough to overcome their fees?  On average, impossible.  So I appreciate what the commenter said, even as I ply my trade as an active manager.  You need some degree of active management in the markets to keep prices in rough line with valuations.

And, maybe, just maybe this means that indexing will have to get to be a much larger proportion of the markets before active managers would have alpha after fees as a whole.  Until then, passive investing is a great way to go for most investors.

4 Comments

  • bbarberayr says:

    Active, long-only value managers can beat the overall indexes, but the industry works in a way which makes this difficult. Many studies, such as the Tweedy Browne “What works in Investing” have shown that buying cheap stocks will outperform the market over long periods of time, so logically it follows that managers following these types of strategies should outperform. The reasons they don’t generally include fees being too high, having to maintain a cash cushion for redemptions, being forced to buy high and sell low due to fund flows, copycat funds, short term focus of investment holders, fund size getting too large, etc.

    For fund managers who truly are concerned about their clients outperforming the market, the solutions to these issues like restricting fees and fund size, limiting investors to those who will lock in for longer periods, etc. mean that the fund manager or company will make less money and most are not willing to do this. I have seen it many times where a new fund manager comes out and outperforms the market while managing a smaller amount of money. The success becomes known and the fund then gets flooded with funds and the outperformance goes away.

    • eric@servo says:

      bbarberayr,

      I think your comment misses the point of my post above. Giving active managers credit for beating broad market averages (S&P 500/Russell 3000) that are inherently focused more on growth companies (when you cap-weight the entire market or subset of the market, you are naturally going to have the largest % in the highest priced “growth” companies) is flawed.

      We don’t give active equity managers credit for beating bonds, why would we give them credit for beating other asset classes they don’t hold?

      When we control for their overall value exposure, we do find that fees, excessive, taxes, and cash balances lead the universe of active value managers to underperform a value index of similar geographic and size (large/medium/small) orientation. Sure, some “win” over some periods of time, but there is no evidence of persistence of these winners out of sample from one period to the next (Standard and Poor’s Index vs Active scorecards have detailed data on this subject).

      The easiest way to check for the persistence of “alpha” from a value manager is to regress their historical returns on the dimensions of expected return: beta (TSM minus t-bills), size (small companies minus big companies), and value (high book to market stocks minus low book to market stocks) — the dimensions being available at Ken French’s website which is updated monthly.

      You see, if this were the 1960s and we operated in a “CAPM world” where we believed expected return was simply a function of sensitivity to the overall market (beta), then comparing everyone to broad market averages would be acceptable.

      However, since the late 1970s, we’ve evolved to understand that relative size and price are additional priced factors that determine returns. Active managers shouldn’t get credit for simple exposure to priced factors we can target at lower cost and more consistently via a passive/index fund. These “3 Factor” regressions can help us understand how much of a manager’s return is attributable to “manager skill” (alpha) vs dimensions of expected return (stocks over bonds, small over large stocks, and value over growth).

      One should be willing to hire an active value manager (at any cost) if it can be determined that said manager has produced statistically significant alpha in excess of exposure to dimensions of return, and they can be relatively certain that said outperformance will persist.

      Unfortunately, the # of managers with positive alphas net of fees is slim (especially after taxes), and by the time we have a track record long enough to rule out “noise” as a possible explanation of positive alpha the manager is probably closer to retirement than not (or is no longer accepting new $). Then, we are left with the issue that even in the presence of statistically significant positive alpha, there is very little evidence that it will persist (see Bill Miller).

      Sorry to be Debbie Downer! :)

      • eric@servo says:

        By the way, I am not saying there is no way a portfolio manager/investor can add value. As was mentioned in the original post, paying attention to trading and execution matters a great deal (especially in smaller and Int’l stocks).

        The closer one is able to capture the bid/ask spread in transactions (sell above the bid and buy below the ask), avoid the cost of momentum (negative momentum in purchases, positive momentum in sales), and delay or avoid taxes (tax-loss harvesting and at least avoiding ST gains), the more of your returns you’ll get to keep. But, of course, we are discussing basis points and not percentage points here, so any value contained in these approaches is totally subsumed by 1%+ fees.

        Finally, we are left with the giant anomaly that is gross profitability. All things equal, more profitable stocks (gross profits/assets) are at the same time healthier and more successful companies that also have higher future expected returns–which is not what we’d expect as these companies also have higher valuation ratios (not typical “value stocks”). Some will point to the fact that highly profitable stocks already have high prices and “risk” not being able to live up to future profitability forecasts (therefore a return premium is warranted to take this risk), but of course that is a bit dubious.

        In what is probably a deviation from pure indexing or “enhanced” indexing, holding stocks with high price/book values (or other value metric) that simultaneously are not unprofitable companies (low gross profitability) is a substantially better approach to buying value stocks than simply low-valuation sorts by themselves (readers should see Robert Novy Marks “The Other Side of Value” for more on this subject).

        Further, in the face of high gross profitability, “growth stocks” are no longer such a bad deal as one would presume simply by observing the returns of cap weighted growth indexes that sort stocks based on high valuation ratios. To the extent that a portfolio holds the most highly profitable stocks that AREN’T the also the most expensive companies (P/B, P/E, etc.), one can fashion a GROWTH portfolio with similarly high expected returns as a value portfolio.

        Combining these two approaches and rebalancing periodically has at the same time low tracking error to broad markets and higher expected returns than a stand-alone value approach (something approaching the holy grail of investing).

        Of course, even this approach is better attempted in a passive/index fashion with an attention to costs and implementation.

  • bbarberayr says:

    Eric,

    Thanks for your comments and the info on the Ken French web site. A few comments.

    I guess the question then is who do the above market-average returns accrue to? Is it the long-term holders, insiders, etc. who do not trade? Or is it individual investors like myself who outperform pretty much every year by looking at small stocks mainly excluded from ETFs? 50% of stocks have to provide returns above the average and given that active funds underperform, someone must be getting that outperformance or it is all being chewed up by fees which would be pretty sad.

2 Trackbacks