| SE&EM
Seminar on March 17th, 2008 (Monday) |
 | | ก@ |  |
Seminar
Department of Systems Engineering and Engineering Management
The Chinese University of Hong Kong
------------------------------------------------------------------------------------------
|
ก@ |
ก@ |
ก@ |
|
Title |
: |
Microstructure Filtering in Financial Market |
|
ก@ |
ก@ |
ก@ |
|
Speaker |
: |
Dr. Jianhui Huang, James |
|
ก@ |
ก@ |
Department of Applied Mathematics |
|
ก@ |
ก@ |
The Hong Kong Polytechnic University |
|
ก@ |
ก@ |
ก@ |
|
Date |
: |
March 17th, 2008 (Monday) |
|
ก@ |
ก@ |
ก@ |
|
Time |
: |
4:30 p.m. - 5:30 p.m. |
|
ก@ |
ก@ |
ก@ |
|
Venue |
: |
Room 513 |
|
ก@ |
ก@ |
William M.W. Mong Engineering Building |
|
ก@ |
ก@ |
(Engineering Building Complex Phase 2) |
|
ก@ |
ก@ |
CUHK |
|
ก@ |
ก@ |
ก@ |
------------------------------------------------------------------------------------------
Abstract:
ก@
|
We first propose a nonlinear filtering problem and then derive some novel stochastic filtering equation as well as Bayes factor equations. Next, we obtain robust versions of these equations and present a novel particle filtering algorithm to implement these equations. The second part of our work is concerned with the following problems: If stochastic volatility (SV) is present in the microstructure stock market? If so, which of the classical SV model best represents the stock prices? We first propose some novel microstructure model which is more general and reasonable than the existing models. Indeed, our model enables us to capture most statistical properties of the price process in microstructure market such as cycles, momentum, mean-reversion as well as discreteness and clustering (biasing). With the presence of microstructure noise, it is not clear if the SV plays an essential role when modeling the stock price in market. We use Bayes estimation to show that the SV remains important and model selection to establish that Heston\'s model represents our stock data significantly better than the other classical models. The prominent feature of our work is that we provide a common framework where different SV models could be tested and by Bayes factor, their performances could be evaluated in a consistent way.
|
-------------------------------------------------------------------------------------------
Biography:
ก@
|
Dr. Jianhui Huang, James got his Ph.D. in Mathematical Finance, Department of Mathematics and Statistics, University of Alberta, Canada. 2002-2007; Master of Science, Department of Probability and Statistics, School of Mathematics and System Science, Shandong University, China. 1998-2001; Bachelor of Science, Department of Probability and Statistics, School of Mathematics and System Science, Shandong University, China. 1994-1998.
|
************************* ALL ARE WELCOME ************************
*******************************************************************
|
|