Historically, the foundations of mean-variance portfolio theory were laid by Hicks , Markowitz [1952; 1959] and Tobin , who developed a rigorous model of individual portfolio choice within a `mean-variance' world, where investment opportunities are evaluated solely in terms of the first two moments of their distributions. Both Markowitz and Tobin demonstrated that mean-variance preferences could be reconciled with the von Neumann-Morgenstern  theory of choice by assuming quadratic utility or the multivariate normal distribution of returns. An important result obtained within this context was Tobin's separation theorem; that individual portfolio decisions can be separated into two distinct phases. The first involved the determination of the optimal combination of risky assets, the second the choice of the optimal allocation between risk-free and risky assets.
This micromodel of choice was aggregated into an equilibrium framework by Sharpe  and Lintner [1965a]. Under the additional assumptions that investors share homogeneous beliefs, and that borrowing and lending opportunities are available at the risk-free rate, they showed that the portfolio of risky assets held by investors in equilibrium must coincide with the market portfolio. The risk premium on any asset was, therefore, a linear function of the asset's contribution to the risk of the market portfolio i.e. the asset's β.