With the notable exception of price to cash flow, very few valuation metrics identified stocks that have outperformed the market so far in 2013. The equities that were cheapest in terms of trailing price to earnings and price to book value (the usual suspects of stock selection) were actually detrimental to portfolio returns for the year.
To assess the relative success of conventional stock picking methods, I grouped all S&P/TSX Composite members into quintiles ranked by the December 31, 2012 values for each of the following metrics: stock price to trailing 12 month earnings; price to forward earnings estimates; price to book value; trailing 12-month profit growth; enterprise value to EBITDA; and price to cash flow. Then I calculated the average year-to-date simple return for each quintile.
The table below shows the results. The trailing P/E ratio, which is the most popular valuation tool by far, was more or less useless this year. The cheapest quintile by trailing P/E was actually the second worst performer and the top performing quintile was the second most expensive quintile.
Price to book value was an even worse indicator of performance. The less attractive (ie. expensive) stocks were by book value at the end of 2012, the better they performed.
Stock selection by price to cash flow worked this year – kinda, sorta. The cheapest, most attractive quintile by P/CF was the best performing and the most expensive was the worst. If we overlook the fact that the fourth quintile outperformed the second (it wasn't by much), we can call price to cash flow a successful stock picking tool. It was clearly the most effective.
There's a limit to what can be learned from this exercise. Grouping by quintiles removes a lot of nuance and hides legitimate differences between companies. But it does support my belief that cash flow should be used as an important part of stock analysis whenever possible.