Working Paper: CEPR ID: DP10444
Authors: Javier Menca; Enrique Sentana
Abstract: We compare Semi-Nonparametric expansions of the Gamma distribution with alternative Laguerre expansions, showing that they substantially widen the range of feasible moments of positive random variables. Then, we combine those expansions with a component version of the Multiplicative Error Model to capture the mean reversion typical in positive but stationary financial time series. Finally, we carry out an empirical application in which we compare various asset allocation strategies for Exchange Traded Notes tracking VIX futures indices, which are increasingly popular but risky financial instruments. We show the superior performance of the strategies based on our econometric model.
Keywords: Density expansions; Exchange traded notes; Multiplicative error model; Volatility index futures
JEL Codes: C16; G13
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| seminonparametric expansion of the gamma distribution + multiplicative error model (C46) | enhanced modeling of the conditional mean and distribution of VIX futures returns (C58) |
| enhanced modeling of the conditional mean and distribution of VIX futures returns (C58) | superior performance in dynamic asset allocation strategies (G11) |
| dynamic asset allocation strategies (G11) | improved risk-adjusted returns (G11) |
| seminonparametric expansion of the gamma distribution + multiplicative error model (C46) | higher conditional Sharpe ratios than existing ETNs (G19) |
| modeling approach (C50) | investment performance (G11) |
| flexibility of the model (C52) | better risk assessment (D81) |