Statistical and computational methods for credit portfolio loss and financial state-space models
Abstract/Contents
- Abstract
- Financial institutions are often exposed to credit sensitive assets such as loans and corporate bonds. Positions involving these securities could be entered into for hedging, speculation, or diversification purposes. An important consideration in the design and risk management of these credit-linked portfolios is the distribution of the portfolio loss. In the first part of this thesis, we develop an analytical approximation for the distribution of the portfolio loss due to defaults in a loan portfolio at a fixed time horizon. Our method is generic in that it can handle a large class of models of default timing, and addresses other important features of corporate loan portfolios, including stochastic volatility, and state-dependent jumps at and between defaults. A related problem which has been studied extensively in the recent filtering literature is that of joint online parameter estimation and latent state filtering in frailty models for credit risk, and other state-space models. In the second part of this thesis, we discuss a new methodological advancement. We introduce an adaptive particle filter that uses a computationally efficient Markov Chain Monte Carlo estimate of the posterior distribution of the state-space model parameters in conjunction with sequential state estimation. Our method can be widely applied to state-space models in economics, finance, engineering, and biostatistics to name a few. The superior numerical performance of our adaptive filter makes it a practical alternative to estimate parameters in a large class of dynamic state-space models in financial econometrics. We discuss one such application and develop an efficient sequential Monte Carlo method which updates market microstructure parameter estimates at each new quote or trade transaction for high-frequency transaction data. We introduce a dynamic non-linear microstructure model for the latent efficient price, which incorporates price discreteness and live market information from the limit order book.
Description
| Type of resource | text |
|---|---|
| Form | electronic; electronic resource; remote |
| Extent | 1 online resource. |
| Publication date | 2013 |
| Issuance | monographic |
| Language | English |
Creators/Contributors
| Associated with | Bukkapatanam, Vibhav |
|---|---|
| Associated with | Stanford University, Department of Management Science and Engineering. |
| Primary advisor | Lai, T. L |
| Thesis advisor | Lai, T. L |
| Thesis advisor | Giesecke, Kay |
| Thesis advisor | Rajaratnam, Balakanapathy |
| Advisor | Giesecke, Kay |
| Advisor | Rajaratnam, Balakanapathy |
Subjects
| Genre | Theses |
|---|
Bibliographic information
| Statement of responsibility | Vibhav Bukkapatanam. |
|---|---|
| Note | Submitted to the Department of Management Science and Engineering. |
| Thesis | Thesis (Ph.D.)--Stanford University, 2013. |
| Location | electronic resource |
Access conditions
- Copyright
- © 2013 by Vibhav Bukkapatanam
- License
- This work is licensed under a Creative Commons Attribution Non Commercial 3.0 Unported license (CC BY-NC).
Also listed in
Loading usage metrics...