1 / 9

Econ 201

Econ 201. David Kim 3.23.2011. Heterogeneous Autoregressive (HAR) Class of Models. Represent expectation of future variance if all non-variance data is ignored Recently, it has been found that, when forecasting variance, simple models often outperform more sophisticated parametric models

jared
Download Presentation

Econ 201

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Econ 201 David Kim 3.23.2011

  2. Heterogeneous Autoregressive (HAR) Class of Models • Represent expectation of future variance if all non-variance data is ignored • Recently, it has been found that, when forecasting variance, simple models often outperform more sophisticated parametric models • Use averaged future RV as dependent variable and use averages of past values of variance measures as independent variables

  3. HAR-RV • Multi-period normalized realized variation over h discrete periods • HAR-RV model • 1, 5 and 22 are used • Refer to daily, weekly and monthly frequencies • Associated with long-memory models

  4. HAR-RV-CJ • Regression of RV on the lagged averaged normalized continuous and jump components

  5. HAR-RV-RAV • Shown to be superior to HAR-RV and HA-RV-CJ • Realized absolute value (RAV) is substituted in as the regressor for realized variation

  6. HAR-RV-RAV (cont’d) • Advantage of using RAV • Highly robust to jumps and sampling error • Jumps do not affect RAV asymptotically • RAV is in different units than RV

  7. Combined Forecasts • HAR-RV-CJ with IV • HAR-RAV with IV

  8. Robust Regression vs. OLS Estimation Methods • Robust method consistently outperformed OLS method over all models and time periods • Offers way to mitigate effects of confounding factors • Robust estimation has three goals • Good efficiency for assumed model • Small deviations from model assumptions should not have small effect on performance • Larger deviations from model should not ruin the model • Robust regression may be better at forecasting variance • Achieves estimates that are closer to the original model

  9. To do: • More focus on out-of-sample forecasting results

More Related