While reading these interviews he did between age 79 and 82, I got the feeling that Graham was enthusiastic as he described a new mechanical formula he had nearly finished testing. He believed his formula was the simplest way for both seasoned analysts and layman investors to find undervalued stocks and outperform the Dow (the performance metric used during his time). After 60 years of analyzing financial statements and managements, Graham said this about projecting earnings, evaluating market share, and analyzing individual companies:
“Those factors are significant in theory, but they turn out to be of little practical use in deciding what price to pay for particular stocks or when to sell them. My investigations have convinced me you can predetermine these logical “buy” and “sell” levels for a widely diversified portfolio without getting involved in weighing the fundamental factors affecting the prospects of specific companies or industries.”
Graham further recommended building a portfolio of 30 diversified stocks meeting such criteria. His study employed strict sale rules that required selling the stocks after a 50% gain or after a two year holding period, whichever came first. Graham noted that in the market downturn of 1973-1974 investors using the formula would have shown paper losses but would have been rewarded soon thereafter for sticking with the formula. Thus, to allow time for the program to work he recommended a minimum of five years. In other words, patience was a requirement for success.
Graham back tested the period from 1926-1976 with his refined formula and concluded that such a program would have earned 15% or more, not including dividends, and would have beaten the Dow by twice as much. He was so excited by the study results that he contemplated including them in the 5 th edition of Security Analysis.