Most research shows that 90 to 95 percent of the risk-return profile of an investment portfolio comes from the asset allocation decision, with things like tactical rebalancing and manager selection combining to make up that last meager 5 to 10 percent. For someone who has spent the vast majority of his professional career — literally thousands of hours each year — researching and selecting asset management firms, those numbers have always been a source of consternation and more than an occasional pang of self-doubt.
There’s not a lot of impact from all that effort.
And I’ve even written about the folly of picking managers in large-cap U.S. equities, where roughly 80 percent of active mutual funds trail their benchmark year in and year out. The market return, which is pure luck, is going to dominate any effect from choosing managers. The limited, and on average negative, effects of managers’ stock-picking decisions make it difficult to justify dedicating so much time and so many resources to such a seemingly futile effort.
And still, we try.
Recently, I heard behavioral finance researcher and author extraordinaire Michael Mauboussin — at an Institutional Investor conference, of course — discuss how to untangle the effects of luck versus skill, based on his book The Success Equation. In his speech Mauboussin laid out a framework for assessing the relative impact of luck as compared with skill on the outcome of an activity. I’ve found his comments to be immensely helpful in thinking about this conundrum.
Mauboussin starts with the concept of the inside view versus the outside view, borrowing from the grandfather of behavioral finance, Daniel Kahneman. Kahneman defines the inside view as how we see our own situation. In this view we rely on our own limited experience to make decisions and we overestimate our own ability, ascribing any success to our unique skills and knowledge.
Conversely, the outside view is how we perceive others. Here we tend to be more objective, searching for patterns and situational context, which allows us to make more-informed decisions that are based on prior probabilities — the so-called base rate — as opposed to merely our own anecdotal experience.
Now successful outcomes always combine some percentage of both luck and skill, but figuring out their relative contributions is critical for effective decision making. Mauboussin argues that it’s a spectrum, where on one end success comes 100 percent from luck, as when playing the lottery or a slot machine, and on the other end skill is the dominant, if not sole, factor in determining success, as when playing chess.
Investing in general probably falls somewhere in between, but depending on the specific characteristics of an asset class or investment strategy, figuring out exactly where is the challenge.
In investing, market efficiency is usually defined as the degree to which all publicly available information is already reflected in the clearing price of a security. In theory, the more efficient a given market is, the harder it is to outperform a passive basket of all the securities in it. But perhaps another way to think of market efficiency is to consider where a given asset class falls on Mauboussin’s spectrum of luck versus skill.
Looking at manager dispersion, or the spread between manager returns, in two asset classes — private equity and public equity — we can clearly observe different competitive dynamics at play. Using 20-year return data from Burgiss and eVestment, we can see that the difference between top- and bottom-quartile managers in private equity is nearly 20 percent per annum, whereas in public stocks the top managers outperform the bottom 25th percentile by just 2.6 percent annually.
Not only is the gap between the best and worst performers markedly different for these asset classes, but the percentage of return of the top managers that can be explained by the average is far different as well. The average public stock manager’s return of 8.8 percent is a full 85 percent of the return of the 10.3 percent of top-quartile managers. (The scale of this economic effect is so large that I doubt it becomes insignificant even if properly adjusted for the effects of market beta or differences in volatility — as opposed to this dirty math.)
Just 15 percent comes from selecting the best manager. However, in private equity only half of the return of top-quartile performers comes from just being average; actually picking the best managers in the space drives the other half of the 22.5 percent.
In more efficient markets — like public equity, where there are 16,000-plus mutual funds picking through fewer than 4,000 listed securities — the market return swamps the effect of skill. You should spend more of your time calculating the base rate or determining capital markets assumptions. That’s to say it’s an asset allocation decision.
In less efficient markets — like private equity, with perhaps 5,000 active private equity funds and more than 500,000 private companies to choose from — skill actually matters a whole lot more. The base rate isn’t as important as how you actually implement the asset class.
This is an interesting theory — but is it borne out by empirical research?
In one of the more unusual looks at the effects of due diligence that I’ve come across, A.J. Watson, at startup funding platform Fundify, compared hundreds of angel investments that were made on its platform. Angel investments are very early and typically very small private investments in new businesses — the earliest-stage and riskiest form of private equity. Because of the nature of the information collected by Fundify, Watson was able to track these investments by the amount of time each investor had spent on due diligence as well as what the eventual return was.
Watson bucketed these deals into research quartiles: those where the investor had spent less than five hours on due diligence before investing; five to 15 hours; 15 to 40 hours; and more than 40 hours. Obviously, hours spent on due diligence served as a proxy for the quality of research performed. And in my opinion, that’s a skill.
Watson then separated these four diligence-based quartiles into three performance groupings, reporting returns as a multiple of invested capital, or MOIC. MOIC calculates how many times the investment multiplied the initial capital; 2x means you made $2 for every $1 you put in. Investments were grouped into those deals that lost money by returning less than 1x, those that fell between 1x and 5x, and those that made greater than 5x capital invested.
How did the various quartiles perform?
The results speak for themselves. Sixty-nine percent of those investments that were preceded by just one hour or less of due diligence lost money. Conversely, 58 percent of angel deals made after a full week of research were winners, and these heavily researched investments were more likely to be among the top deals than those in any of the other categories. Twenty-six percent of angel deals consummated after 40 hours of due diligence returned 5x their original investment or more.
And the averages were even more instructive. Angel investments made after 40 hours of diligence generated an average MOIC of 7.1x. Those done with less than one hour of work made a mere 0.8x, losing money on average.
The trend is crystal clear. The more due diligence performed, the better the returns.
Perhaps institutional investors with limited time and resources can use this framework to prioritize where to allocate manager selection efforts and active risk taking. For a small team with a large portfolio and liquidity to spare, it may make sense to go completely passive in public equities, prioritizing limited manager research efforts exclusively for illiquid markets. Conversely, if you cannot bring the appropriate resources to bear in selecting investments in private markets, you’re probably better off not doing them at all. Skill matters.
And yes, maybe now I do feel a little bit better about all those long hours spent in manager meetings.