Portfolio , Quant Investing


After a few introductory posts ( part 1 , part 2 , part 3 ) on quantitative investing it’s time to get down to some real current portfolios. In this post I want to look at the first half one of the best risk adjusted return strategies presented in What Works On Wall Street and how an investor can implement it in their own portfolio.

The Combined Consumer Staples/Utilities Strategy is one of the best strategies ranked by risk-adjusted return with annual returns of 16.56%, a Sharpe ratio of 0.84 and also with a maximum drawdown of -34%. This strategy leverages investing in the two lowest risk sectors of the market; the consumer staples sector and the utilities sector. The consumer staples half of this strategy is a pure value strategy that invests in an equal weighted portfolio of the top 25 stocks in the consumer staples sector ranked by shareholder yield. That’s it. Hold for a year and re-balance. Rinse and repeat. It doesn’t get much easier. I ran the screen based on Friday closing prices and below are the results.

Cons Staples Screen June 10 2013

Note that the list contains 26 stocks. The second stock on the list Smithfield Foods (SFD) recently received a buyout offer from a Chinese company. One of the rules I discussed in my pervious post is one for acquisitions. You sell or don’t buy if the share price reaches 90% of the buyout offer. SFD is trading within 90% of the $34 buyout offer so we add the next stock on the ranked list, Con Agra Foods (CAG). I’ve usually put the group of stocks into an equal weighted portfolio in the portfolio tracker at finviz.com.

But this does highlight a couple of important topics I want to highlight. First, its important to note that the consumer staples is a value strategy that uses shareholder yield as its value measuring stick. Shareholder Yield is a metric that starts with dividend yield but adds buyback yield to it, i.e what percentage of shares is the company buying back every year. It does not use more traditional value metrics such as PE ratio, PB ratio, EV/EBITDA, etc… It turns out that in different sectors the value metrics that yield the best performance vary. In one sector shareholder yield may outperform, while in a different sector other value metrics do better. As you’ll see in the Utilities strategy sometimes its a combination of value factors that does best. For this strategy dividend yield is the second best value factor. You can use it without losing too much in returns or risk. By using dividend yield you can implement it in the free finviz screener.

The second point I want to highlight is that the more uncommon the metric the harder it is to implement for the individual investor. There are a lot of screeners that let you sort by P/E ratio but try to find one that even has shareholder yield in it. The best full function low cost stock screener I have found is Stock Investor Pro from AAII. It’s a PC based application and its $198 a year. Not too bad considering what you can do with it’s but not free by a long shot. Personally, I have found it well worth the money. The learning curve is not too bad either.

How is that in terms of effort to implement a strategy that has returned over 16% a year since 1965 with risk adjusted returns over 3 times the market? Once you’ve learned how to implement this strategy it takes less than 10 minutes to run and maybe an hour making sure none of the portfolio management rules have been violated and that’s about it.

In the next post I’ll cover the second half of the strategy, the Utilities portfolio.

Tagged , , ,

18 thoughts on “ Quantitative Investing – Consumer Staples Strategy

  1. Another great article in the series Paul. Whenever I see a new opportunity such as this my first question is “how to begin?”. In other words in this case would seasonality enter into the picture? I suppose over time it would not matter whether the investor rebalanced but It would not be fun to start out now at market highs and observe a big summer pullback. (sorry if you covered this somewhere and I missed. After 60 my comprehension and retention has gone all to hell! 🙂

    1. Hey Doug, I was waiting for someone to ask that question. All the return figures for all the quant strategies from the book are an average of 12 annual portfolios. In other words,a portfolio is formed in Jan and tracked for 1 yr periods over the whole range of years. Another portfolio is formed in Feb and tracked for one year periods over the whole range, etc…for all 12 months. Then the compound annual returns of the 12 portfolios are averaged.

      In other words, this method eliminated any seasonality bias in the data. The seasonality is still there, Jan is still the best month, but the persistence of the factor(s) tested remains.


      1. Great minds think alike. ha ha…Actually referring to great minds (yours and Mebane Fabers)…did you know that he recently kicked off an etf that focuses on shareholder yield? SYLD

        1. Yeah, I saw that. Have the book. I value his research a lot. He charges too much for his ETFs though, and now he is charging for his research as well.
          The SYLD book is good though and worth the price.


  2. Wow! Such a well done and useful explanation – clear, concise, thorough, minimum acronyms. No fluff and BS. A joy to read and use your articles. It will be a sad day when a financial house bribes you away from this passion of yours. Thanks again.

  3. I’m reading through this whole series of posts on quant strategies and taking it all in! How disheartening to see so many of the stocks on this sample list are tobacco, soda and other consumer goods I avoid like the plague! I’ll keep reading and hopefully will find a strategy I can follow with less indigestion.

  4. Hey Paul,
    Ever since you began the series on quantitative stock screening I’ve been exploring the possibility of implementing it into my plan.
    After looking at different ways to go (AAII pre-defined screens, P123 pre-defined and R2G screens, learning to develop my own) I still have not decided.
    I do have a question for you about AAII Stock Screener Pro. One of the screens that is of particular interest to me is a free R2G screen offered at Portfolio123 that limits its search to David Fish’s CCC lists that he publishes for free every month. Hemmerling is the guy who created it and he has a few other screens there but it appears he is a income investor. My question is whether the AAII screener will allow screening from a specified list of stocks provided by the user?
    Have you put any real money behind your screens yet?
    As always thanks for all you do and best regards,

    1. Doug,

      Yes, SI Pro can screen from a specified list of stocks. You just need to define the portfolio of stocks and go from there. I have the CCC list of stocks as portfolios in my SI Pro.

      As for the way I use quant screens, I said in my first qaunt post: “I’ve studied many of the strategies, learned how to implement the strategies with available low cost tools, paper traded the top strategies for over a few years, and have committed real money to a few of them.”

      Most of my individual stock picking has gone by the way side and that portion of my portfolio is now in quant screens, this is my second year with real money invested – mainly the Trending Value screen but I also use the Consumer Staples screen and a variant of Trending Value for foreign stocks. Come Jan I will up my allocation again to the quant screens.


  5. Hi Paul,

    Thanks for the great post.

    I’m trying to reproduce your screen in SI Pro, however the software follows a different sector classification that the one used in the WWOWS book. (GICS)

    How did you build your SI Pro screen to capture all the Consumer Staples stocks including ADR’s as defined by the GICS classification?



  6. Thanks for the quick answer Paul.
    However GICS sectors do not map neatly vs. SIP’s sector breakdown…
    I think you need detailed GICS stock list to precisely reproduce the WWOWS strategy.


    1. True. But it doesn’t need to be exact for still great results. It was one of the reasons I switched to P123 as my data source.

      1. Hi Paul,

        Could you kindly share how you implement the “Trending Value” strategy in P123?

        I am going back and forth between P123 and SIP, but I have not found a good way to implement the “trending value” strategy in P123.

        Thanks in advance, and please keep posting to this very informative blog!


        1. Hey Trevor,

          Implementing Trending Value is a bit tricky. First you need to create a custom stock univierse to match the O’Shaughnessy All Stock section of the market. Then you need to create a custom rank that includes all the value parameters P/B, P/S, etc… This requires two custom formulas as well, EV/EBITDA and SHY. Then you can create a screen to look for stocks in the top decile of the value composite and sort that top decile by 6 month return.

          Make sense?


Comments are closed.