In the last year of his life, Ben Graham gave an interview to a trade journal–Medical Economics–in which he outlined a very simple quantitative method for selecting stocks.
In fact, the method was simple to the point of simplistic and quite provocative in its total disregard for growth prospects. Graham, in his crustier years, seemed to have become quite dismissive of the idea of even attempting to determine an appropriate price to pay for growth stocks, something that he considered too difficult to produce consistent results. He also considered it impractical to attempt to engage in micro-analysis of a company’s projected earnings or prospects.
So, what did he recommend to the good doctors?
- Make a list of stocks selling at seven times or less their latest–not projected–earnings.
The magic number seven was relevant at the time of the interview (1976), but his rationale was as follows: If bond yields are high, you should look for a low multiple at which to buy stocks, and if they are low then you can pay a higher multiple. His rule-of-thumb was that the earnings yield of a share (the inverse P/E) should be at least twice the average current yield on AAA corporate bonds. - Identify the companies within the above subset that have a satisfactory financial position.
His recommended test for this was to look for companies that own at least twice what they owe. To do this the ratio of stockholders’ equity to total assets (the equity ratio) should be at least 50% or .5, which would indicate that at least half of the total assets are financed by stockholders and not creditors. - Buy a basket of at least thirty stocks that meet these criteria and hold each with a profit objective of 50%, before selling. If it doesn’t reach that target by the end of the second calendar year after purchase, sell anyway. If you are not finding many candidates with which to replace sold stocks given these criteria, then sadly it is a sign that the market is overpriced and you should buy government bonds. Note: Graham here remains true to his Intelligent Investor principle for the defensive investor that no more than 75% of a portfolio (and no less than 25%) should be held in stocks, with the balance always held in government bonds.
BG boldly states that, in the long run, this method should average a return of 15% a year or better. Unfortunately in today’s miserable market climate, anybody touting a potential return of 15% p.a. would either be committed or arrested, but 1976 America was a happier place with more normal bond yields and fewer battle-scarred investors.
Anyway, as usual, I thought I would run a few of my current holdings and possible future purchases through this short and sweet test to see how things appear. Rather than use a P/E of seven, I will apply his rule of thumb to calculate an equivalent figure for today, where Aaa corporate bond yields are about 3.8%. He recommended an earnings yield (inverse of the P/E ratio) of at least twice the quality bond yield, which would give us a target of about 7.5%.
Rolls Royce: RR’s earnings yield (earnings/price x 100) is (65.59/1,076) x 100 = 6%, which is a first hurdle fall. Just for academic purposes, Rolls’ equity ratio is (total shareholder’s equity/total assets) 3,782/15,422 = 0.245. Another failure.
Diageo: Diageo’s earnings yield is (104.4/1,905) x 100 = 5.48% and its equity ratio is 7,036/25,077 = 0.28. BG would not bite.
IBM: IBM’s earnings yield is (14.63/184.29) x 100 = 7.9% and its equity ratio is 22,792/126,223 = 0.18. One out of two isn’t too bad although, if the US AAA corporate bond yield were used (2 x 4.3%), then this would fail too.
JPM: JPM’s earnings yield is (4.03/57.9) x 100 = 6.9% and its equity ratio is 211,178/2,415,689 = 0.08. Fortunately the equity ratio is not relevant here because it is a bank, something that I am fairly sure would have ruled it out for Graham regardless.
Not encouraging. Perhaps government bonds aren’t such a bad deal after all?
References:
- If you are interested in the full interview, then you can find it here.
- Here is a link to historical long-term (20-30yr) Aaa US corporate bond yields.
- All figures from Digital Look or Morningstar.
Disclosure: Long RR., DEO, IBM, JPM.
Disclaimer: This post is not a recommendation to either buy or sell. Please consult your investment advisor.