Abstract
The Markowitz portfolio selection model is a quadratic programming model, which utilizes the variance-covariance matrix of asset returns per period of time as a measure of risk. The model minimizes the risk while trying to make expected return on the portfolio as large as possible. This assumes that we can find some assets with large expected returns but little effect on the variance. Such assets can be used to drive the return on the portfolio to the target return without increasing the variance. The coefficient of variation on the other hand, naturally decreases with increasing mean and therefore provides a way to formulate a portfolio of assets. The model belongs to the class of fractional quadratic programming problems. The portfolio optimization problem is formulated as a fractional programming problem where the squared coefficient of variation of the portfolio is minimized. The method is illustrated using some stocks from the Nigerian Stock Exchange, Lagos.
Keywords: portfolio optimization, Nigerian stock exchange, coefficient of variation