In this article, we'll teach you how to use quantitiative analysis to discover the best stock-picking hedge funds.
As thousands of hedge fund allocators can attest, identifying skilled managers is a difficult process. It doesn’t help that most investors rely only on past performance and word of mouth as indications of how managers will perform in the future. But if you can’t rely on how a fund has historically performed to make a smart investment, what should you look at? According to Novus’ Chief Research Officer, Stan Altshuller, an ideal manager knows their strengths and has shown mastery in the fundamentals, including market timing, position sizing, and security selection, among others.
While there are many distinct skills that help managers generate alpha, this article will focus on what might be the most essential (and difficult) skill to master: security selection, or the ability to pick names that outperform their relevant benchmarks. Security selection is the first thing a discerning investor should consider when screening a prospective manager—high security selection alpha shows a fundamental understanding of the investment space the fund occupies, and implies that the manager has developed a thorough research and screening process.
Our research team, as well as our many institutional investor clients like state pensions, endowments, and foundations, use two metrics to determine a manager’s general stock-picking ability: batting average and win/loss ratio.
Batting average, or the percentage of total names that generate positive PnL, is a good measure of consistency. After all, total return alone says nothing about where or how the value was generated. Are they producing their returns through luck? A single position? Sector swings? Below you can see the batting average of Point State Capital from 2011 to present. About 62% of the securities they picked generated positive PnL, a good indicator of consistent security selection ability.
While a high batting average is a good sign, it is not a complete measure of security selection skill. Incorporate the manager’s Win/Loss Ratio, or how much an average winner contributes vs. how much a loser detracts, and you get slightly closer to the truth. For example, Coatue has a batting average of about 50%, less impressive at first glance than Point State’s 62%.
However, their Win/Loss Ratio is over 3x, compared to the former WLR of 1.5x. This means that Coatue’s winners drastically outperformed their losers, another broad measure of security selection ability. These two figures give a relatively complete assessment of a manager’s abilities, but ultimately they don’t tell the whole story.
In order to measure true security selection skill you need to know more than just what percent of total names did well, or by how much winners outperformed losers. You have to decompose returns into their contributing parts. There are many ways to do this, but at Novus, we use a system we call the Novus Framework. It breaks down total return into four principal components: Market, Sector, Security, and Trading.
By isolating the portion of returns that can be attributed to picking individual names within a sector, we get to the heart of stock-picking skill. It also allows us to perform the critical task of differentiating between beta and alpha.
What can you do with this type of analysis? For one, it allows you to investigate claims that managers make. Let’s say you get a pitch book from a manager with a 7% return over the last five years who claims they’re a sector specialist in Information Technology. If this was the case, you would expect to see high security selection alpha in that sector, and the majority of capital allocated to those names. By using Novus Attribution, you can not only check for style drift and an increase in market capitalization, but ensure that the manager really does create alpha in IT.
This type of analysis is essential when searching for new managers to invest in. By combining the ability to screen for fundamentals in a manager’s portfolio with public ownership data, you’re able to discover and assess managers using quantitative methods. Public ownership data allows an investor to accomplish two very powerful things: first, it enables them to evaluate the performance and behavior of their current managers; and second, as an investor searches the platform for that perfect healthcare or industrials fund to diversify their portfolio, it allows them to ensure the manager has demonstrated excellent security selection skills in the past.
As an example, we’ll take one manager from our Public Ownership platform. As you can see in the chart below, Coatue has shifted exposures slightly over the last three years between its two primary sectors, replacing Consumer Discretionary exposure with IT.
This observation begs the question, is the change promising or concerning for investors in the fund? Attribution analysis allows us to see that while Consumer Discretionary has been a true source of alpha, the manager’s security selection skill is in the IT space.
In other words, the move has likely been beneficial for investors. Coatue is clearly no sector generalist, but what matters is that they excel at picking stocks that outperform their relevant benchmarks in their specific areas of expertise.
An astute investor will use every possible resource to evaluate the abilities of their managers, be they current or prospective. Novus Attribution analysis combined with the right Public Data platform offers a powerful way to measure manager skill and ensure you’re making investments that make sense for your portfolio and your investment goals.Published on January 26, 2017