To provide further evidences on jumps in asset prices, in this thesis we conduct an empirical analysis on high-frequency data from a stock index and consider the problem of identifying jumps at intraday intervals. Our approach generalizes two existing methods in the literature in terms of estimating spot volatility and of correcting for the spurious rejection problem due to multiple testing. The proposed procedure directly depends on a credible volatility model that we specify and calibrate from the index data. By simulating the volatility model, it is shown that a relevant parameter which governs the shape of the generalized extreme value (GEV) distribution determines the critical regions of jump tests. Empirical sizes of jump tests can then be held at nominal level approximately when the testing procedure is applied to high-frequency returns. We also study the dynamics of detected jumps and model their time-varying intensities with a linear self-exciting point process.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:663227 |
| Date | January 2013 |
| Creators | Tsai, Ping-Chen |
| Publisher | Lancaster University |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
Page generated in 0.0117 seconds