Market Efficiency: A Theoretical Distinction and So What?

longer follow. The capital asset pricing model (CAPM) is an elegant theory. With the aid of some simplifying assumptions, it comes to dramatic conclusions about practical matters, such as how to choose an investment portfolio, how to forecast the expected return of a security or asset class, how to price a new security, or how to price risky assets in a merger or acquisition. The CAPM starts with some assumptions about investors and markets and deduces its dramatic conclusions from these assumptions. First, it assumes that investors seek mean–variance efficient portfolios; in other words, it assumes that investors seek low volatility and high return on average. Different investors may have different trade-offs between these two, depending on their aversion to risk. Second, the CAPM assumes that taxes, transaction costs, and other illiquidities can be ignored for the purposes of this analysis. In effect, it assumes that such illiquidities may impede the market’s approach to the CAPM solution but do not change the general tendency of the market. A third CAPM assumption is that all investors have the same predictions for the expected returns, volatilities, and correlations of securities. This assumption is usually not critical.1 Finally, the CAPM makes assumptions about what portfolios the investor can select. The original Sharpe (1964)–Lintner (1965) CAPM considered long positions only and assumed that the investor could borrow without limit at the risk-free rate. From this assumption, and the three preceding assumptions, one can deduce conclusions of the sort outlined in the first paragraph. The assumption that the investor can borrow without limit is crucial to the Sharpe–Lintner model’s conclusions. As illustrated later in this article, if we accept the other three CAPM assumptions but assume limited (or no) borrowing, the Sharpe–Lintner conclusions no longer follow. For example, if the four premises of the Sharpe– Lintner original CAPM were true, then the “market portfolio”—a portfolio whose amounts invested are proportional to each security’s market capitalization—would be an efficient portfolio. We could not find a portfolio with greater return (on average) without greater volatility. In fact, if the four premises of the Sharpe–Lintner original CAPM were true, the market portfolio, plus perhaps borrowing and lending, would be the only efficient portfolio. If, however, we assume the first three premises of the Sharpe–Lintner CAPM but take into account the fact that investors have limited borrowing capacity, then it no longer follows that the market portfolio is efficient. As this article will illustrate, this inefficiency of the market portfolio could be substantial and it would not be arbitraged away even if some investors could borrow without limit.