Continuing off of the previous post, it’s important to note that while CAPM serves as a foundational model in financial theory and practice, it is not without its shortcomings. Despite its widespread application, criticisms of CAPM highlight significant challenges and limitations in its assumptions and empirical evidence.

For one, the Capital Asset Pricing Model assumes an idealized financial market without taxes, where all investors are rational, have homogeneous expectations, and exhibit risk-averse behavior. These assumptions are highly unrealistic. In reality, financial markets involve hidden information, income taxes, transaction costs, etc. Hence, the assumption of a perfect capital market by CAPM is disconnected from reality. Additionally, not all investors hold homogeneous expectations. For instance, CAPM assumes that all investors share the same outlook on Samsung Electronics’ future stock prices, which is hardly plausible.

Roll demonstrated that if an arbitrary efficient portfolio is considered as the market portfolio, then the expected returns and beta are linearly related. This implies that empirical evidence supporting CAPM’s linear relationship between expected returns and beta is insufficient. The empirical analysis by Black, Jensen, and Scholes showed that there is no or even a weak negative correlation between beta and returns. Thus, their findings indicated that CAPM does not accurately reflect reality.

Furthermore, to rationalize historical market returns, a much larger risk aversion than empirically observed among individuals is required.

The conditions for CAPM to hold are very stringent compared to Arbitrage Pricing Theory (APT). CAPM requires the pre and post returns of the market portfolio to follow a normal distribution, investors to exhibit at least second-order risk aversion in their utility functions, absence of information asymmetry among investors, and adherence to the Semi-Strong Form Efficient Market Hypothesis (EMH), among other unrealistic assumptions.

A significant amount of information is required to calculate individual asset values. When the capital market consists of n assets, it requires n × (n+2)/2 variances and covariance values. Although Professor Sharpe’s Index Model significantly reduced the information required to 3n+2, it still remains substantial compared to APT, which allows for individual asset valuation with partial analysis alone.

Despite the numerous limitations mentioned above, CAPM is overwhelmingly used in financial practice for the following reasons:

It is the most familiar and easy-to-understand valuation model for practitioners across various industries, including accounting, taxation, insurance, banking, and credit rating, not just finance majors. Individual asset values can be evaluated based on the returns of index funds (≒market portfolio). Systematic risk-return relationships can be linearly represented through regression analysis to derive Security Market Line (SML) and Capital Market Line (CML). It is useful as a first-order approximation (1st order approximation) in investment evaluation, providing excellent short-term explanatory power and the ability to reject most null hypotheses. Existing company operating and financial information (such as alternative beta) can be used to evaluate new investment opportunities.

The process of valuing a company using CAPM involves the following steps:

Calculate the required rate of return by estimating beta (β) based on market volatility and the stock’s volatility. (End of CAPM’s role) Predict the company’s growth based on its past and present situations (as reflected in its financial statements) and its current situation (including its structure, management, tangible and intangible values, etc.). Calculate the value expected to be obtained from the company in the future based on its growth. Discount the calculated value at the required rate of return. (The higher the required rate of return, the lower the fair value.)

In financial management exams, like Hamada’s model, when you calculate the beta value of a debt-free company and then provide the beta value reflecting the financial risk premium of the debt, your supervisor might question you with an incredulous look, “Who are you to ignore the beta value on Bloomberg and value the company as you please?” However, the Bloomberg beta is merely based on five years of monthly returns. The raw beta is nothing more than the calculation of five years of monthly returns, and the adjusted beta is raw*2/3+1/3.