A valuation model based on expected growth in book equity, the P/B-ROE model is one of a wide variety of valuation models derivable from simple economic assumptions. It is useful in practice because historical return on equity is a fairly good indicator of future return on equity and because the statistics required to estimate the relationship are simple. Furthermore, the P/B-ROE model appears to be superior (in terms of its mean squared error) to price-earnings ratio models when applied to historical data for the majority of companies covered by the
The P/B-ROE model allows estimation of investment horizon, required share-holder return and market consensus expected return on equity using historical data. Security analysts can thus compare their own estimates of future return on equity with the estimates of the market consensus. Their purchases should be biased toward stocks whose estimated average ROEs exceed the market consensus expected return on equity.
Finally, the P/B-ROE model leads to some interesting conclusions about the relations between stock price and earnings stability, dividend policy, leverage and beta. It suggests that stable earnings growth does not necessarily lead to higher prices, that dividends do matter, that leverage can be either good or bad, and that stocks with higher betas do not seem to have higher required returns in the way predicted by the Capital Asset Pricing Model.