Abstract
In this article we propose two new methods of portfolio allocation which are applicable for all return distributions. The properties of these new methods are compared with that of Markowitz's mean-variance method using extensive simulation. It is found that the new methods perform appreciably in terms of growth of wealth as well as protecting against the downside risk, in situations where the return distributions of one or more of the stocks is heavy-tailed. These methods can be effective substitutes for the mean-variance method which is not applicable for return distributions with heavy-tails having infinite expectation or variance.
Notes
1 The usual definition of VaR of a given portfolio of instruments is the maximum expected loss for a given time horizon and for a given confidence level, attributable to changes in the market price of financial instruments.