I am a huge fan of low-cost index funds. They provide investors with their fair share of market return in each investment category without paying high fees or dealing with the frustration of trying to guess which active manager will get lucky.
That being said, there’s always a category or two each year where the average active manager beats the index fund in the category. You’ll find them in the investment style that performs the worst relative to all other styles. Active managers don’t even have to be good to win in a bad sector. They just have to survive.
This phenomenon has a name. It is called Dunn’s Law, so-named after Steve Dunn by friend and cohort William Bernstein of Efficient Frontier Advisors. Dunn’s Law is as follows: “When an asset class does relatively well, an index fund in the asset class does even better. In contrast, when an asset class does poorly, the active managers do better in that asset class.”
Dunn’s Law can be extended into investment styles within an asset class. Mutual fund research companies sort mutual funds within an asset class into different investment styles for peer group comparison. These groupings typically are composed along two dimensions: a size axis and a style axis. The familiar Morningstar style box is one such methodology, as illustrated in Figure 1.
Figure 1: Morningstar Style Box
Style indices are composed of securities that fall within each of the nine boxes. Index funds that track a style hold securities that fall in that style box. For example, a large-cap value index fund holds only stocks in the large-cap value style box. This creates a “pure play” on that style.
A vast majority of actively-managed mutual funds are not pure style plays. Fund research companies such as Morningstar dumped all mutual funds into one category based on their weighted average size and style factors regardless of how messy the selection is. Each fund within a peer group is then ranked to find the best performing funds over various time periods.
Funds managers can select stocks in many different styles. Morningstar may classify a fund as large-cap value fund even though it includes mid-cap and growth stocks. That because its weighted average for all securities still falls inside the large-cap value box. See Figure 2 for an illustration of how a typical actively-managed fund may be spread across several different styles.
Figure 2: Typical Stock Picks of an Active Large-Cap Fund Manager
Figure 2 shows how the typical large-cap value active fund invests 75 percent of its securities. At the centers of the oval is the large-cap value core, while security selection bleeds in all directions. The final 25 percent is invested even further away from the core holdings.
Index funds outperform most active funds, in most styles, most of the time. On average, over a 5-year period, index funds beat two-thirds of surviving active funds in each style. This outperformance has been very consistent over the decades. See my latest book for detail, The Power of Passive Investing.
Active funds perform well when a style is out of favor. On average, about two-thirds of active funds beat index funds in the style with the worst short-term performance record. This occurs because the active managers are not style pure in their fund selections. Being messy boosts fund performance when the core style performs poorly. This is when active funds tend to outperform comparable index funds.
William Thatcher of Mercer Hammond has been quantifying Dunn’s Law for several years. He calls it the purity hypothesis. Thatcher first published research on this phenomenon in 2009 in the Journal of Investing and recently updated his findings in 2012 in The Journal of Index Investing.
Thatcher studied the purity hypothesis over 150 independent time periods and across three index providers: MSCI, Russell and S&P. He found a 0.88 correlation between how well a style performed and how well active managers in that style performed relative to a comparable style index.
Experienced fishermen know that fish tend to congregate in one section of a lake. The challenge is to figure out where that section is. The purity hypothesis challenges active investors in the same way. If you can figure out which investment style will deliver the worst performance, you have a high probability of picking a winning actively-managed fund in that style.
This all sounds interesting, and perhaps some people can use this information to pick a winning manager, but let’s not forget one important fact. We invest to earn money. Why would you invest any money in a losing style if you’re skilled at picking them?