Please use this identifier to cite or link to this item:
http://hdl.handle.net/11375/25544
Full metadata record
DC Field | Value | Language |
---|---|---|
dc.contributor.advisor | Maheu, John | - |
dc.contributor.author | Li, Chenxing | - |
dc.date.accessioned | 2020-07-20T09:23:36Z | - |
dc.date.available | 2020-07-20T09:23:36Z | - |
dc.date.issued | 2020 | - |
dc.identifier.uri | http://hdl.handle.net/11375/25544 | - |
dc.description.abstract | This thesis examines three important topics in finance: co-jumps among risky assets and portfolios, proposing a new Bayesian semiparametric stochastic volatility model with Markovian mixtures, and whether better return density forecasts lead to economic gains in portfolio allocation practice. First, jump models are useful in capturing skewed and/or leptokurtic financial returns. So far, research has focused on jumps in a single asset, including co-jumps between return and volatility. On the other hand, co-jumps among assets are also important especially in practices such as beta dynamics and portfolio allocation. I propose a parsimonious yet flexible multivariate GARCH-jump mixture model (MGARCH-jump model) with multivariate jumps that allows both jump sizes and jump arrivals to be correlated. The model identities co-jumps well and shows that both jump arrivals and jump sizes are highly correlated. The model also provides better prediction and better investment outcomes as opposed to the benchmark multivariate GARCH model with normal innovations (MGARCH-N model). Second, I extend the Bayesian semiparametric stochastic volatility (SV-DPM) model of Jensen and Maheu (2010). Instead of using a Dirichlet process mixture (DPM) to capture return innovations, we use an infinite hidden Markov model (IHMM). This allows for time variation in the return density beyond that attributed to latent volatility dynamics. The new model (SV-IHMM) also nests the SV-DPM as a special case and improves the density forecast from the SV model with Student-t innovations, SV-DPM and other benchmark models. Our model is applied to equity stock returns, foreign exchange rates, commodity prices and industrial growth rates. The model demonstrates robust and consistent gains in out-of-sample performance against other benchmarks. Furthermore, predictive densities from the SV-IHMM exhibit clear distributional change over time. Third, I investigate the relationship between statistical improvements in density forecasts of returns and actual economic gains in portfolio allocation for a risk-averse investor. To aid this investigation, this chapter proposes a new multivariate Bayesian semiparametric model that has better out-of-sample density forecasts than benchmark models. Results show that this more sophisticated econometric model does provide positive economic gains whether the investor’s utility is CRRA, CARA or quadratic. The economic gain diminishes when the investor is more risk-averse because she is moving away from risky investment positions. | en_US |
dc.language.iso | en | en_US |
dc.title | Three Essays on Forecasting Return Distributions with Mixture Modelling | en_US |
dc.type | Thesis | en_US |
dc.contributor.department | Business | en_US |
dc.description.degreetype | Thesis | en_US |
dc.description.degree | Doctor of Philosophy (PhD) | en_US |
Appears in Collections: | Open Access Dissertations and Theses |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
Li_Chenxing_202007_PhD.pdf | 3.65 MB | Adobe PDF | View/Open |
Items in MacSphere are protected by copyright, with all rights reserved, unless otherwise indicated.