Suppose we calculate g as ROE (1–p)/(1–ROE (1–p)) and the Ke by the CAPM. We replace both values into the formula PER = (ROE (1+g) – g)/ROE (Ke-g) but there PER we obtain is fully different from the one we get by dividing the quotation of the share to the earnings/share. Is this possible to undervaluation of that share on the market or interpret that difference as an overvaluation?