Some Graham and Dodd Type Thoughts on Stocks vs. Bonds

by: Tom Au, CFA

The essence of the Graham and Dodd philosophy is to buy stocks almost as if they were bonds. The investment value metric that we use (book value plus ten times dividends) is basically a bond metric.

Bill Gross, the bond guru at Pimco, has observed that the investment return on stocks in the 20th century was dividends, plus inflation, plus 0.6% a year real. Put another way, U.S. blue chip stocks are glorified TIPS (Treasury Inflation Protection Securities).

Around the turn of the century, stocks' dividend yields of less than 2% were less than real TIP yields of nearly 3%. That's why stocks have done nothing for a decade, because they weren't competitive on a risk adjusted basis.

The relationship is somewhat more reasonable now, with stock yields just "wide" of Treasury yields, and way above TIP yields. The respective yield premiums are now meaningful (many stocks beat straight bonds on inflation protection, and TIPs on yields).

Even so, we expect U.S. stocks to eventually lose 70% of their 2000 peak in real terms as they have done in the past. (The 2007 peaks were nominally higher than in 2000, but actually lower, after inflation.)

This could happen in one of two ways. The 1930s model involved nominal declines above 80%, which went back into the 70%s after the accompanying deflation.

The 1970s model involved inflation. Nominal stock values stayed flat for 18 years, during which time the subsequent inflation eroded the real value by more than 70%.

The market goes up about four times (in real terms) over a good 18 year cycle, and then loses most of that back (almost 75%) in the bad 18 years.

So expect a continued erosion of most of the gains of 1982-1999. A "strong" stock generation such as that is being followed by a weak stock generation from 2000-2018(?) to pull back the average annual real capital gain to Bill Gross 0.6%. The point is, you get to keep any dividends that are paid in the meantime.

The U.S. blue chip market is like a sine wave, not an upline. It basically hovers around 0% (or if you prefer, 0.6% a year) real. But values range from -1 to +1, causing booms and busts.

Therefore, we feel much better with a stock that has a dividend, providing some income. Then it is much like TIPS (Treasury inflation protected securities). If dividends are non-existent, then our proprietary investment value metric is just equal to book value. Then we try to buy such a stock as if it is a zero coupon bond; at as deep a discount as possible...

Bank stocks often meet this criterion, at least nominally. But we'd be very careful with them, because we're not at all sure of the sustainability of their dividends, or for that matter, of their book values, given the large write-ups and writedowns of assets.

We were more successful with chemical stocks earlier this year. Ashland Chemical was at half of book value in the mid-20s, when we sold it, but we were able to buy it for less than 8. Dow Chemical was also in the single digits when we bought our last batch. Dupont sometimes sells for investment value; $8 a share of book value, plus ten times its dividend takes you to the low $20s.

Other areas of interest are tech stocks. From time to time, at the bottom of their cycles, they sell for less than book value. Current and former holdings include Agilysis (NASDAQ:AGYS) and Celestica (NYSE:CLS).

Disclosure: Still long AGYS. No longer long ASH, CLS, DD, or DOW.