Statistical Methods for Large Portfolio Risk Management

• Activity Number: 112
• Type: Topic Contributed
• Date/Time: Monday, August 4, 2014 : 8:30 AM to 10:20 AM
• Sponsor: Section on Risk Analysis
• Author(s): Jian Zou*+
• Companies: Indiana University-Purdue University Indianapolis
• Keywords: High-frequency ; volatility modeling ; portfolio allocation ; risk management

Abstract

The field of high-frequency finance has experienced a rapid evolvement over the past few decades. One focus point is volatility modeling and analysis for high-frequency financial data. Volatility analysis plays a central role in modern finance. The abundance of high-frequency financial data stimulates the study of price and volatility movements over a relatively short period of time to reflect the current market dynamics more accurately. It allows us to employ flexible nonparametric methodologies for the volatility study based on high-frequency returns directly. It makes possible to estimate the vast volatility matrix of a large number of assets. On the other hand, high-frequency financial data pose a set of new challenges for researchers and practitioners. Observed high-frequency financial data exhibit complex features and structures, such as micro-structure noise, non-sychronization, irregularly spaced times between observations, non-stationarity and jumps. In this talk, we introduce a new methodology to carry out efficient asset allocations using regularization on estimated integrated volatility via intra-day high-frequency data. We illustrate the methodology by comparing the re