Can Simple Quantitative Rules Improve the Performance of Fundamental Investors?
Simple quantitative rules achieve superior predictive accuracy than qualitative, clinical judgments. This is one of the most established findings in academic psychology, yet also one of the findings we humans have the most trouble believing. Surely, we think, the detailed and thorough analysis of financial experts should produce better results than simple quantitative rules?
This skepticism leads many investors to spurn investment strategies they perceive to be too reliant on quantitative rules, either because those rules seem too simple to work or because they appear to be an indecipherable “black box.”
But at Verdad, we believe that quantitative rules produce better results than human judgment — and that human judgment is more like a “black box” than quantitative rules that are clearly laid out and described, as we have done in our published papers on leveraged small value equities and predicting deleveraging..
This week, we wanted to show how quantitative rules can improve the performance of even some of the best fundamental investors — and show that the perceived tension between quantitative and qualitative approaches should give way to a more synthetic “we’ll use what works” approach.
To do this, I applied Verdad’s risk management rule to the long equity portfolio of Glenview Capital to test whether my rules enhanced or detracted from performance at a fund ranked No. 8 in the world by Institutional Investor and known for high quality fundamental research. I chose Glenview because I respect and admire the fund and I wanted to demonstrate that simple quantitative rules can increase performance even at a fund that is among the best of the best at fundamental research.
The Rules
We implement a simple risk management rule for Verdad’s portfolio: we do not invest in stocks with high short interest or with low-quality financials as measured by Piotroski’s F-Score. Academic research has found that short interest predicts significant negative abnormal returns. Piotroski’s F-Score is a systematic way of analyzing a company’s financial statements, scoring companies on basic tests like whether debt is increasing or decreasing, and whether net income and cash from operations are positive. Stocks that score low on this measure of financial health are at much higher risk of bankruptcy.
The goal of this rule is to help avoid the worst performing stocks that drag down overall portfolio returns. A recent JP Morgan report found that over long periods about 40% of stocks will suffer a catastrophic 70%+ loss, while only 7% will compound returns at an extremely above-market rate. Good defense is therefore more important than a good offense.
We believe these accounting and market-based strategies for avoiding junk stocks are powerful ways to improve investment returns that have been proven in academic literature across a broad cross-section. The question of this report is: are these rules helpful to sophisticated investors doing significant qualitative research, or should they be used strictly in screening or by quant-only investors?
The Methodology
We studied Glenview’s publicly reported quarterly holdings back to 2004. We assumed that Glenview held each quarterly portfolio for the entire quarter and then switched to the next quarter’s portfolio on the quarter end date — this simplifying assumption will provide some noise relative to the fund’s true holdings, which would have been bought and sold during the quarter. This methodology does not account for any large secret stakes, options and other derivatives, or short sales.
For this analysis, we considered any stock with short interest higher than 5% of shares outstanding or with a Piotroski F-Score below five as having failed the risk management rules.
The Results
These two rules would have had a significant positive impact on Glenview’s long equity portfolio.
Figure 1: Performance of Glenview Positions That Passed and Failed Risk Management Rules
Source: CapitalIQ, Verdad
We then compared the portfolio excluding the failed stocks with the performance of the total portfolio. The portfolio excluding the highly shorted names outperformed in 71% of quarters. The total return of the highly shorted portfolio is 12.2% relative to 14.1% for the portfolio excluding stocks that failed our tests (the difference in magnitude between the table above and the table below reflects the varying percentage of highly shorted stocks over time, with highly shorted stocks representing a heightened portion of the portfolio in the worst quarters).
Figure 2: Performance of Glenview Positions That Passed Rules vs. Total Portfolio
Source: CapitalIQ, Verdad
To illustrate this another way, we show the cumulative outperformance of the portfolio excluding stocks that failed our tests.
Figure 3: Cumulative Alpha from Excluding Stocks That Fail Our Risk Management Rule
Source: CapitalIQ, Verdad
As you can see, the rules have added value consistently over time.
To make this more tangible, we show below the best and worst performing investments Glenview has made in the past five years relative to the short-interest (as percent of shares outstanding) and the Piotroski F-Score at the beginning of the quarter. Seven of the 10 worst performers could have been avoided using this system, while only three of the best performers would have been eliminated. No rule is perfect, but this methodology works in helping investors avoid the steepest losses.
Figure 4: Biggest Winners
Source: CapitalIQ, Verdad
Figure 5: Biggest Losers
Source: CapitalIQ, Verdad
In sum, simple quantitative rules can help enhance performance even at the best, most sophisticated fundamental investment shops. These rules work consistently and can help investors systematically avoid stocks that are most likely to suffer catastrophic losses.