Three Essays In Computational Finance PDF Download

This dissertation contains three essays.The first part studies the continuous-time dynamics of VIX with stochasticvolatility and jumps in VIX and volatility. Built on the general parametric affinemodel with stochastic volatility and jumps in the logarithm of VIX, we derive alinear relationship between the stochastic volatility factor and the VVIX index. Wedetect the existence of a co-jump of VIX and VVIX and put forward a double-jumpstochastic volatility model for VIX through its joint property with VVIX. Usingthe VVIX index as a proxy for stochastic volatility, we use the MCMC method toestimate the dynamics of VIX. Comparing nested models of VIX, we show thatthe jump in VIX and the volatility factor are statistically significant. The jumpintensity is also stochastic. We analyze the impact of the jump factor on VIXdynamics.The second part establishes a forecast framework for the bond excess return basedon macroeconomics fundamentals. Empirical evidence has suggested that excessbond returns are forecastable with macroeconomics fundamentals. In our study, webuild new links to tie the forecastable variation in excess bond returns to underlyingmacroeconomic series. Based on two types of models, the linear model and additivemodel, and utilizing different combinations of screening methods, nonlinearizationtechniques and regularization techniques, we extract different factor combinationsfrom 131 macroeconomic series, including employment, housing, financial, andinflation factors. This approach results in stronger forecast power for the excessbond returns compared with existing macro-based return predictors. The nonlineareffect of the macroeconomic predictors on the excess bond returns is recovered ifwe incorporate nonlinearized macro data in the analysis. A horse race comparingdifferent variable selection approaches allows us to propose a robust model thatgenerates highly accurate predictions of bond risk premia. Finally, we perform acomprehensive analysis of risk premia with an ETF dataset.The third part of this dissertation is a summary of traditional asset allocationmethods performance on Chinese market. Since traditional asset allocation methods are well analyzed in US capital market, similarly, we want to conduct a comprehensiveanalysis of asset allocation techniques on Chinese market. Based on a horseracecomparison among the trading performance by different asset allocation approacheswith investment universe of Chinese capital market indices and the associatedETFs, we achieve a clear understanding on the relative ranking of different methods,finding the link between trading performance with different parameter estimationtime windows and different investment universe as well. To explain the differencein the trading performance of several methods, we perform a simulation study andattribute bad performance as the inaccuracy of return estimation.