As a value investor one should qualitative look into the future and avoid stocks in definitely dying industries such as the video rental industry or the paper phone book industry (however just maybe, remember that Berkshire was a cash-cow in the dying textile industry when Warren Buffet started buying stocks in the company during the 60s). If one is buying a really cheap troubled company based on trailing multiples, one should also have a feel that even if the company currently is in trouble, it will get out of trouble at some point.
Using trailing multiples instead of multiples based on future estimates of fundamentals have historically yielded superior investment returns as frankly no one can precisely predict the future.
According to the article "Analyzing Valuation Measures: A Performance Horse Race over the Past 40 Years" from The Journal of Portfolio Management (Wesley R. Gray and Jack Vogel as main authors) using a forward looking analyst consensus earnings multiple to systematically invest in the quintile cheapest portfolio of stocks from 1982 to 2010 resulted in a CAGR of 8.63 % vs. the overall market CAGR of 11.73 %. Trailing multiples such as Price/ Book and EV/EBITDA earned 13.63 % and 16.73 % CAGR respectively over the same time period. In other words you will under-perform the market as a value investor if you trust earnings forecasts. What we do know is that asset prices have a tendency to be mean reverting.
