Is Being Different Better? Dispersion and Active Management

Beating the market very obviously requires performance that is different than the market, which in turn requires portfolio holdings that are themselves different from the market. “Active Share”–or the percent of your portfolio which does not overlap with your relevant market benchmark (say, the S&P 500 or the Russell 2000)–has become a buzz word in the industry thanks to a paper showing that managers with higher active share (i.e. more unique portfolios) have delivered superior performance through history. The concept is under a bit of fire lately (from AQR among others), but I think the concept is useful. To be clear, being different (or having high active share) is not some silver bullet. All being different does it ensure that you have a better chance of earning more unique returns—either uniquely good or uniquely awful.

One area of investigation in this field guide will be portfolio concentration: how many stocks should you own, and how top-heavy should your portfolio be?

At one extreme is a Warren Buffett-style portfolio. I love the idea of “focus investing” as described in Robert Hagstrom’s book The Warren Buffett Portfolio: Mastering the Power of the Focus Investment Strategy.

Focus investing is the antithesis of a broadly diversified, high-turnover approach. Among all active strategies, focus investing stands the best chance of outperforming an index return over time, but it requires investors to patiently hold their portfolio even when it appears that other strategies are marching ahead. In shorter periods, we realize that changes in interest rates, inflation, or the near-term expectation for a company’s earnings can affect share prices. But as the time horizon lengthens, the trend-line economics of the underlying business will increasingly dominate its share price.

As Buffett himself has said, “The goal of each investor, should be to create a portfolio (in effect, a `company’) that will deliver him or her the highest possible look-through earnings a decade or so from now.” For adherents of this style, diversification is a four letter word, because too much of it dilutes potential performance.

At the other end of the active spectrum are portfolios which “tilt” a market index towards particular factors (say, small value stocks), but earn returns more similar to the overall market. In this camp fall many quantitative/smart beta/fundamental index managers who believe that factors offer a strong and enduring advantage but should be applied broadly in a very well diversified portfolio. People that like this style think of it as a better beta; people that don’t like this style think of it as closet indexing: a portfolio that looks a lot like the market index and delivers market-like returns, but charges active management fees.

Both have their merits (which we will explore), but regardless of where you fall on the spectrum of active management portfolio concentration, the opportunity set from which you pick your stocks is very important. You can own one stock, or you can own everything.

A “Stock Picker’s Market”

If an active manager is skilled at picking stocks, he or she would hope for a market with lots of dispersion (difference in performance between the best and worst performers). So what areas of the market have offered such dispersion through history?

First, let’s explore the sectors in which stocks have delivered widely dispersed returns. The numbers in the table below represent the average standard deviation of one-year (twelve month) returns within a sector. These numbers include all stocks with an inflation adjusted market cap higher than $200 (equally weighted), and are therefore skewed higher by small cap stocks. The higher the number, the greater the differences, on average, of stock returns with that sector. I show the results for the full period (1963-2015), the last ten years (2005-2015), and last year.

ssector dispersion

Health Care and Technology stand out as sectors where a perfect stock picker would do extremely well. But notice how much the technology sector has changed in recent years. Once the home to some of the highest fliers and worst performers, the technology sector has recently had dispersion that looks just like the rest of the market.

tech and market dispersion 5 years

Digging one level deeper, here are the same results broken out by industry group.

ind group dispersion

Here we find that biotechnology stocks drive the high dispersion in in Health Care results, especially in the more recent periods. Utilities, on the other hand, seem to be a stock picker’s nightmare, offering by far the most clustered returns.

It is important to note that while we always root for a “stock picker’s market,” it does not follow that in such markets active managers will do well in aggregate. This study from S&P Dow Jones undermines the idea (at least for the recent past) that higher dispersion leads to better results for active management as a whole. I think this may have something to do with the fact that so many more managers are “closet indexers” in this sample of years than in the past, but the results are still interesting. Josh Brown has also explored this same topic.

active managers and dispersion

All that higher market dispersion means is that the spread between best and worst will likely be higher, which is confirmed by the finding in the same study.

active managers and spread

Market Cap Segments

What I find even more interesting than sector dispersion is the dispersion of returns within various market cap segments. Below I compare the average excess returns for the best and worst performing 10% of stocks in different market capitalization ranges.

dispersion of returns by cap space

These excess returns are based on the last 50 years, and are measured against an equal weighted benchmark. So, for example, the 58% result for the best large stocks means that, on average, the top 10% (based on total return) of large stocks in any given year outperform an equally weighted basket of all large stocks by 58%.

What jumps out immediately is that as you move down towards smaller stocks, returns are much more dispersed, on average.

This means that in small cap, a more unique/concentrated portfolio is more likely to earn very unique returns relative to market benchmarks. Small cap is a paradox for big asset managers: it offers the most opportunity for differentiated returns, but is also the least lucrative because of very limited capacity (it would be very difficult to manage more than a few billion in a small-cap strategy, or a few hundred million for micro-cap).

Does Dispersion Matter?

Since our goal is earning the highest returns possible above the market, it stands to reason that we should prefer parts of the market which offer the greatest chance of earning huge returns. This may tempt us to think that we should all be trying to pick micro-cap biotechnology and software stocks. But we can’t make that mistake. It is very difficult to consistently pick winners in hyper-competitive areas like biotech and technology—indeed anyone that has done so was likely more lucky than skillful.

The results from sectors and industry groups are interesting but may not be useful because, as the change in technology dispersion makes clear, high dispersion in a given part of the market may not persist.

I do believe that these results highlight an opportunity in micro- and small-cap investing. This is especially true for smaller, individual investors who don’t have to worry about market impact when trading. Large institutions simply can’t run meaningfully large micro-cap portfolios, and so many small stocks go unnoticed. For institutions/fund managers hunting in these spaces, two important considerations are turnover and assets under management. Lower levels of each would reduce market impact. There may be a “paper portfolio” or “backtested” advantage (i.e. returns on paper without any of the real costs accounted for–taxes, commissions, market impact) to being a small-cap value investor instead of a large-cap value manager, but the market impact costs associated with small cap trading can erode that advantage quickly, so beware.

In future posts I will further explore the relationship between dispersion, value, momentum, and active management. I welcome ideas and sources on the topic.

I’ll leave you with five more ideas/questions:

  1. Does the value factor work better among less covered stocks? These could be stocks with zero or minimal sell-side analyst coverage, or those not represented in a major index.
  2. Have you read studies on dispersion and factor returns? I’ll be doing my own but curious if the ground has been covered yet.
  3. Stock selection factors (value, momentum, etc) are like risk: they need to be tied to an investment horizon to be understood. Momentum is short term, value longer term, perhaps quality is the longest term. We will explore these horizons in great detail.
  4. Why isn’t there a company that allows you to use your fingerprint for EVERYTHING? I feel like we should already be at the point where your fingerprint is your house key, your car key, your ID, and your entire wallet. What am I missing?
  5. Just read Oldman’s Guide to Outsmarting Wine: 108 Ingenious Shortcuts to Navigate the World of Wine with Confidence and Style by Mark Oldman, anyone have any hidden favorite wines?