Abstract
It is well known that the standard mean variance approach can be inappropriate when return distributions feature skewness, fat tails or multimodes. This is typically the situation for portfolios including derivatives. In this case, it can be necessary to come back to the basic expected utility approach. In this paper, an efficient portfolio maximizes the expected utility of future wealth. This paper presents an analysis of the efficiency frontier, formed by a set of efficient portfolios corresponding to a parameterized class of utility functions. First, we discuss the estimation of an efficient portfolio and introduce several tests of the efficiency hypothesis, depending on what is known about the utility function and the budget level. Next we analyse the shape of the frontier and develop a procedure for testing the separability of the efficiency frontier into K independent funds. The inference is semi-nonparametric because the return distribution is left unspecified. We illustrate our approach by an application to portfolios including derivatives.
| Original language | English |
|---|---|
| Pages (from-to) | 1-41 |
| Number of pages | 41 |
| Journal | Journal of Empirical Finance |
| Volume | 12 |
| Issue number | 1 |
| DOIs | |
| Publication status | Published - 1 Jan 2005 |
| Externally published | Yes |
Keywords
- Efficiency test
- Efficient portfolio
- Expected utility
- Fund separation theorem
- Risk monitoring
- Value at Risk