Where alpha lives: measuring conviction in hedge funds

Where is the skill?

Given the last five years’ underperformance of hedge funds, investors are wondering whether paying high fees to active managers still makes sense. Is this an anomalous period or have they lost their mojo for good? To unravel the 2 trillion dollar question, we began by analyzing public ownership data for signs of persistent skill in managers. What we found is that the alpha machine is not broken; indeed it is humming along quite nicely. The main reason it’s been difficult to spot active management skill in recent years is the market has roared upwards leaving many hedged strategies behind, weighed down by their short books. But one place where skill is still evident –perhaps even on the rise– is in what we call conviction trades: trades that multiple managers size heavily in their books.

Measuring conviction

To measure trades with conviction, we took an aggregated view of public equities within the hedge fund industry, scrubbing over 1,000 managers’ historical public filings. For each quarter, we took 20 stocks with the highest conviction from multiple managers and simulated daily P&L tracking it like a live portfolio. We call this our Conviction portfolio and talk about it in our prior research. This approach takes advantage of two critical manager skill sets revered by investors: stock selection and position sizing skills. Evidence points to these skills playing a more prominent role lately. Some shrewd investors use public data to tap these specific skill-sets directly, bypassing the high fees associated with hedge funds. Of course such a strategy would miss out on most other skill-sets like trading, exposure management, and shorting for alpha among many others. Nonetheless, studying this portfolio gives investors a good sense of where managers are placing their big bets. We used the Novus Alpha platform to run our analysis. Let’s talk about some basic properties of the so-called conviction trades.

Market capitalization

Looking at the portfolio over time we can observe some clear trends. For instance, most conviction bets are in large and mega cap stocks, and that trend has been increasing over the past decade at the expense of smaller capitalization stocks. This implies a couple of things: First, it makes the portfolio fairly liquid and mitigates the risk of crowding present in more thinly traded securities. Second, it shifts managers to an area where they have been successful historically: mega caps represent the highest win/loss ratios and batting averages of all buckets (not counting micro caps of which there was only one stock). Looking at the table below we see that 75% of all mega cap conviction stocks over a 10 year period have been winners. More importantly, each winner contributed over three times the P&L that each loser detracted, translating to a 3.09 win/loss ratio.

Sector bets

Looking at conviction trades by sector, we can see an overweight conviction bet in Information Technology and more recently in Healthcare. While managers have historically enjoyed a high win/loss ratio in conviction Technology bets, this has not been the case with Healthcare.

This is also evidenced by the portion of P&L attributable through stock selection skill in each sector. Technology leads the other sectors while Healthcare is a laggard – most of the contribution hedge funds enjoyed in that sector came from the upwards move in the general market and the sector itself. Another sector where hedge funds are overweight is energy. Recently the bet has hurt managers amidst falling oil prices slamming energy companies. This is partially the reason for the sector’s poor win/loss and batting average numbers. It should be noted, however, that security selection in the sector has been positive historically as conviction names outperformed the sector. Most of the pain came from the sector-specific bet alone.

In the next installment we will talk about the relative performance of conviction trades and recent changes to the make-up of the portfolio.

‍