The Heterogeneous Autoregressive Model; the Role of Long Memory in Realized Volatility

Richard T. Baillie, Department of Economics, Michigan State University, USA

The presence of long memory in Realized Volatility (RV) of asset prices is a widespread stylized fact. The origins of RV have been attributed to jumps, structural breaks, nonlinearities, or pure long memory. The alternative economic explanations are extensions of the Heterogeneous Autoregressive (HAR) model with jumps and good volatility. This paper assesses the contribution of these rival explanations through the use of fractional long memory models combined with extended HAR models and random coefficient extended HAR models. We find evidence that the statistical modeling of long memory is generally important, in addition to more structural model explanations.

Please direct all questions to the following contacts - Your question will be routed to get you an answer as quickly as possible.

• MSUStatSymposium2018@stt.msu.edu
• Michigan State University
  Department of Statistics and Probability
  Room C413 Wells Hall
  619 Red Cedar Road
  East Lansing, MI 48824