A comprehensive analysis of CAPM (Capital Asset Pricing Model) was carried out. Our analysis within CAPM for several dozen listed companies from different countries, as well as numerous results of other authors, showed that the CAPM results differ significantly from the real returns of companies. The reasons for this discrepancy are discussed, which are related to the internal properties and shortcomings of the model, as well as its possible modifications that can bring the model closer to real life. Among the latter, accounting for financial risk in CAPM along with business risk, various amendments such as the Fama-French corrections, Arbitrage Pricing Theory (APT) and the incorporation of CAPM into modern capital structure theories. Accounting for financial risk in CAPM along with business risk was made by Brusov, Filatova and Kulik recently analytically (CAPM 2.0 model), in contrast to the phenomenological model of Hamada, and a large difference in these results was emphasized. In addition to the renormalization of the beta–coefficient, obtained in the Hamada model, two additional terms are found: the renormalized risk–free return and the term dependent on the cost of debt kd. Disadvantages of Hamada's model are discussed. Two versions of CAPM (market and industry) are considered and it is shown that the latter is closer to real life.