Yes, the CAPM is Absurd : OLS is Misunderstood and Incorrectly Modeled Mathematically

The Capital Asset Pricing Model (CAPM) purports to mathematically relate the expected return of an asset with its risk. Two types of risk are addressed: systematic risk, which is a function of the market; and unsystematic risk, which is independent of the market and is asset specific. The market risk is described as Beta (β), which linearly relates the asset return with the market return. The unsystematic risk is described in terms of errors which are deviations from the linear relationship between the market and asset. The β is estimated by ordinary least squares (OLS). In system theory, OLS may be modeled as a linear signal processing system with inputs and outputs (i.e., a "black box" with "gosintas" and "gosoutas"). Samples of signal and errors are treated as the input to the OLS "system" which transforms the samples into an output in the form of a prediction along with its expected value and variance. The purpose of OLS is to reduce the prediction variance by "filtering out" (a term from target tracking) error samples (described as noise in tracking) to improve future estimates. Unfortunately, the CAPM represents a devastating failure to understand, model, and use the OLS correctly while, as a detrimental myth, it continues to be perpetuated by the uninformed. Yes, the CAPM is patently absurd as Fernandez explains. In contrast, proven and long established engineering techniques are applied here to derive the empirical statistical OLS asset return prediction, expected value, and variance. This engineering model suggests that the CAPM should be completely scrapped. Instead, the well established and appropriately suited OLS should be correctly used in its place rather than be cannibalized by obsessively attempting to justify the preposterous delusion that α must be zero and then ridiculously force-fitting the mismatched β onto the fatally defunct CAPM