Name
Abstract | ||
|---|---|---|
We present a new approach to portfolio optimization by separating asset return distributions into positive and negative half-spaces. The approach minimizes a so-called Partitioned Value-at-Risk (PVaR) measure by using the statistical information from the two half-spaces respectively. We show that the proposed PVaR approach is a signiflcant improvement in several important aspects when compared to Markowitz mean-variance optimization approach. First, our approach, which accom- modates ambiguous asymmetric return distributions and captures portfolio risk in higher moments, does not require asset distributions being elliptically symmetric or multivariate normal. Second, using simulated and real data, our approach generates better risk-return tradeofis in the optimal portfolios. The difierence between the two approaches increases in the degree of asymmetry in the underlying asset distributions. Third, when given the support of asset returns, our PVaR measure becomes a coherent risk measure proposed by Artzner et al. (1999) whereas conventional risk measures such as variance and VaR fail to do so. Moreover, our PVaR measure is an asymmetric risk measure, which is difierent from symmetric risk measures like variance and worst-case mean-covariance VaR (WVaR). Therefore, our proposed PVaR is a signiflcant addition to the existing portfolio risk mea- sures. We believe that the PVaR approach can be very useful for better portfolio allocations than the mean-variance or other symmetric risk-metrics approach during market downturns when asset return distributions are often fat-tailed or skewed. |
Year | DOI | Venue |
|---|---|---|
2012 | 10.1016/j.ejor.2012.03.012 | European Journal of Operational Research |
Keywords | Field | DocType |
Risk management,Asymmetric distributions,Partitioned value-at-risk,Portfolio optimization,Robust risk measures | Mathematical optimization,Project portfolio management,Robust optimization,Portfolio,Efficient frontier,Risk management,Portfolio optimization,Risk measure,Value at risk,Mathematics | Journal |
Volume | Issue | ISSN |
221 | 2 | 0377-2217 |
Citations | PageRank | References |
26 | 1.08 | 11 |
Authors | ||
4 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Joel Goh | 1 | 164 | 9.07 |
| Kian Guan Lim | 2 | 60 | 5.35 |
| Melvyn Sim | 3 | 1909 | 117.68 |
| Weina Zhang | 4 | 28 | 1.46 |