Chapter 3 came with a warning; Beware of valuation models. They are not a solution. They are enticing, tempting the investor to plug into a formula to get to a valuation. But they typically don’t work for four reasons:
First, they are sometimes based on assumptions that are not realistic. The CAPM is an example, derived assuming returns follow a normal return distribution: they don’t in practice. And CAPM does not work.
Second, the assumptions may be palatable, but they are incomplete. The Dividend Discount Models is an example. It assumes that investor invests to get dividend returns and that is not only palatable but also consistent with no-arbitrage valuation theory. However, for forecasting dividends over a given horizon, it’s inconsistent with theory: Dividend payout is irrelevant to value.
Third, the inputs required cannot be estimated with any precision. The CAPM, where a beta and the market risk premium is required (what is that?), is again an example. Introducing imprecision invites speculation.
Fourth, most models are impractical. They require forecasts over a very long horizon, too much for us limited humans and also for AI with all its supposed intelligence. Formulas play a trick by inserting a growth rate in a terminal value, but what is this? It is just an estimate of the long term, a speculation.
Formulas are for fools.