I believe the text says:

firms should not diversify;
diversification does not add value to a firm;
combining assets is additive for the present value of their expected returns;
present value of a stock is present value of expected dividends.

Isn't this counter-intuitive with real life? Don't most firms diversify - big firms like GE, and big insurance companies? Monoline, mono-state insurance companies are the exception. Even the mom and pop store sells more than just one item.

Doesn't diversification by the firm potentially reduce the variability of the firm's returns if there is negative correlation, or at worse if there is zero correlation permit growth while not increasing the variability? Shouldn't the Florida property insurer diversify by adding California liability (or California earthquake) insurance, because these have zero correlation?

Isn't reduced variability an input to the determination of the dividend? If two insurers each have a 20% profit this year, but one is a property catastrophe only insurer and the other is a diversified insurer, isn't the second insurer likely to pay a larger dividend? Perhaps buried in that decision is that the catastrophe insurer is earning the same IRR on the retained earnings.

I would appreciate some discussion.

[NEAS: Firms diversify for many reasons. Senior managers want to diversify to protect their jobs. They do not want to be in single product industries, since unexpected changes in demand might cause the firm to fail. Senior managers also want to build larger, more poweful firms, which increases their pay and power. So firms are always looking for ways to diversify and to grow. But shareholders can diversify on their own by holding shares of different companies. They can diversify more eficiently than the company's managers can. Shareholders prefer firms that are highly specialized with high returns in niche markets.]