He describes his source of ideas from 4 situations: 1) value of sum of the parts greter than market price; 2) turnarounds; 3) undervalued growth stocks; 4) ‘broken IPOs’.
In his latest very interesting Bloomberg interview he underline the importance of being diversified (the number of stocks in his portfolio is more than 250 names and the 1st 3 names have around 1% weight) but what struck me was when he mentioned as his ‘core’ criteria in try to find undervalued stocks the price to book and the price to sales ratio. Every time I try to think like a value manager I forget the meaning of price to sales (see this interesting post for more informations: seekingalpha.com/article/128052-the-glitch-ten-best-price-to-sales-stocks).
I promptly went to my Bloomberg, searching how, in aggregate, his portfolio compare with market and competitors. I found very interesting to see that, in a universe of a bit less of 2,000 US mutual funds (exactly 1,966 mutual funds) with more than 100 mln $, focused on the US stock market and with a portfolio updated at least to the end of June ‘10, RYPNX ranked 42nd in my list with an average price to book @ 1.2 and 17th with an average price to sales of 0.49 (...). So, if I sum this two ratio to see a whole ranking (I know, it’s not the most sophisticated tool, but it’s simple) I found ‘my’ RYPNX ranking on the 19th, so, it seems coherent. But, interestingly, I found in the 1st 5 positions these names:
Aegis Value Fund, Nuveen Mid Cap Value Fund, Harbor Small Cap Growth Fund, Nuveen Large Cap Value Fund and Fairholme Fund
... it seems a very broad subset in terms of style and the median market cap was between under 200 mln $ (Aegis Value Fund and Franklin MicroCap Value Fund) and the 24 bln $ of Fairhome.
My conclusion? The presence of a large groups of investment style and one of the best portfolio manager (Bruce Berkowitz) has convinced me to deepen the validity of price to sales as an important valuation metrics when I choose stocks or funds for my investments.
Editor's notes: Interested readers should certainly check out GuruFocus screener of Predictable Companies that are sold at historical low P/S ratios. Also you can look at the current P/S ratio of stocks relative to the historical P/S ratios. Example in JNJ