Comparing bonds, prefs and shares

Preference shares
The return on a preference share is actually pretty simply, simpler even than bonds, assuming you hold it forever. Provided the issuer doesn't default, then annual return = yield * par /price.

Bonds
If a bond trades at par, then annual return = yield. Otherwise it needs to be adjusted for its maturity date and price.

Shares
If you ignore the problems of reinvestment, then the return on a share should be straightforward to calculate, provided the RoE of that share remains constant.

For instance, Company A has a RoE of 12%. It distributes all earnings as dividends. Company B has a RoE of 12% and reinvests all earnings. You purchase both at book value (100). Company A pays you 12 per year in perpetuity - a return of 12%. Assuming you can always buy shares in Company A at book value, then you could reinvest each year and compound that 12% growth - Company B simply takes care of that reinvesting for you. The return on both is the same.

As a second example, what if you buy shares in Company A and B at 50% of book value. Company A distributes 24% of your purchase price per year in dividends. You earn a 24% return. Company B reinvests its earnings at 12% RoE, which is 24% return on your investment. If the share price recovers to book, then the owner of A or B is getting a 12% on his newly valuable portfolio. He is still earning 24% compounded on his original investment of 50% of book.

What if you buy shares at twice book value? Now Company A is paying you 6% per year, and Company B is reinvesting at 12% RoE. But if this is the most attractive investment opportunity around, and you own Company A, then you are stuck reinvesting dividends for a 6% return. Company B is reinvesting dividends at a 12% return.

Company B clearly seems a more valuable company. When Company B's RoE exceeds the dividend yield available on Company A, i.e. when companies trade above book value, an investor has a dilemma as to where they can reinvest their dividends. The most attractive option is therefore to invest in a company with high RoE, which has opportunities to productively reinvest earnings, i.e. the Buffett approach.

When companies typically trade at below book value, then it is much less important whether they have opportunities to invest earnings. The biggest driver will be their P/E ratio and P/B value (i.e. the Graham approach). This probably reflects the different investing environments that Buffett and Graham have worked in.