Range-Based Estimation of Quadratic Variation

Arbeitspapier

This paper proposes using realized range-based estimators to draw inference about the quadratic variation of jump-diffusion processes. We also construct a range-based test of the hypothesis that an asset price has a continuous sample path. Simulated data shows that our approach is efficient, the test is well-sized and more powerful than a return-based t-statistic for sampling frequencies normally used in empirical work. Applied to equity data, we show that the intensity of the jump process is not as high as previously reported.

Object Details

• Language: Englisch

References and Relationships

• Bibliographic citation: Series: Technical Report ; No. 2006,37

Classification and Topics

• Classification: Econometric and Statistical Methods and Methodology: General ; Data Collection and Data Estimation Methodology; Computer Programs: General ; Single Equation Models; Single Variables: Time-Series Models; Dynamic Quantile Regressions; Dynamic Treatment Effect Models; Diffusion Processes
• Subject: Bipower Variation ; Finite-Activity Counting Processes ; Jump Detection ; Quadratic Variation ; Range-Based Bipower Variation ; Semimartingale Theory

Contributors, Places and Time

• Event: Geistige Schöpfung
• (who): Christensen, Kim ; Podolskij, Mark
• Event: Veröffentlichung
• (who): Universität Dortmund, Sonderforschungsbereich 475 - Komplexitätsreduktion in Multivariaten Datenstrukturen
• (where): Dortmund
• (when): 2006

Further information

• Handle: http://hdl.handle.net/10419/22681
• Last update: 10.03.2025, 11:42 AM CET