Improving the performance of quant value portfolios

Value investing holds the promise of great returns over the long run and appeals to many investor’s innate sense of value derived from their personal experiences. But value investing has a couple of key downsides that make it very hard for many investors to stick with the approach long enough to experience the promised great long term out performance. Big drawdowns and long periods of under performance are two of the big downsides of value investing. That begs the question if these issues can be mitigated or at least improved with the addition of other factors. In this post I’ll look at the addition of one quality factor to quant value portfolios to see if there is any improvement.

A quick reminder on the great performance of value portfolios. As I first mentioned in my post on the Trending Value portfolio, value portfolios are great performers over the long run. The Trending Value portfolio starts off with a value portfolio based on a composite of value factors;

The first part involves doing a composite value screen on the universe of all stocks. By using the all stocks universe it opens the strategy to all companies with market caps of $200M or greater which provides a smaller cap benefit to this strategy. The value composite used for this screen, value composite 2 (VC2), is the same one I described in my last post on the Utilities strategy. It basically is a ranked list of stocks based on a composite score of P/E, P/B, P/S, P/FCF, EV/EBITDA, and Shareholder Yield. This part of the strategy is one of the most powerful strategies in the book. It’s called the VC2 high 25 strategy. It simply buys the top 25 stocks ranked by the VC2 score. From 1964 through 2009 this strategy returned 18% a year with a sharpe ratio of 0.7 and a max drawdown of 55.6%.

The problem with value portfolios is the existence of value traps which leads to large drawdowns and long periods of under performance. Value traps are stocks that are cheap because they should be and that is not captured in the simple combination of value factors. Can we reduce or eliminate these value traps from the screens by incorporating some quality factor? You probably won’t be surprised to hear that there has been at least a few quant studies done on incorporating quality metrics into quant portfolios. In What Works on Wall Street, O’Shaughnessy has an entire chapter on accounting ratios – which I think is just a way of using financial statements to pick out indicators of quality stocks. Wesley Gray, author of Quantitative Value and publisher of the Alpha Architect site is another great resource in this area. As an example I’ll use the issuance of debt as a quality screen to try and eliminate value traps.

In my new quality value quant portfolio I screen for value stocks ranked by the same value 2 composite used in the trending value screen. The difference is that I weed out the worst offenders in terms of debt issuance. I rank stocks by percentage debt change (net debt change in $$$ divided by total debt) and only keep the top 3 deciles (least change in debt) of stocks (30%). Annual portfolios are formed and held for a full year. Rebalance, rinse, and repeat. How does that compare to the standard value composite portfolios and to the index? Below are the results from a backtest on Portfolio123 going back to the beginning of 1999 (as far back as I could go).

Not bad. The Value 2 quant portfolio is good, handily crushing the SP500 over time. But adding a simple debt filter as a quality check improves the results significantly. The debt filter improves sharpe and sortino ratios by 50% and adds another 5% a year of Alpha to the already good results of the value composite screen.

In summary, adding a quality metric like debt change can significantly increase the returns and risk adjusted performance of quantitative value portfolios. Quality metrics such as this should be considered an essential part of any quant value portfolio.