Statistical Modelling of High Frequency Multi-dimensional Stock Data | Department of Mathematics

Statistical Modelling of High Frequency Multi-dimensional Stock Data

We aim to develop univariate and multivariate stochastic models for high-frequency data of the Indian stock market. These models will be used in development of risk management and regulatory strategies.

The main problems we are considering are:

  1. To establish generalizations of concepts like Value at Risk to many dimensions as well as for heavy-tailed distributions.
  2. To find corresponding optimal hedging strategies.
  3. To develop regulatory strategies to mitigate the `flash crashes’ that have been observed in recent years as a consequence of high-frequency trading.

The analysis will be based on tick-by-tick data purchased from the National Stock Exchange, Mumbai.

Principal Investigator: Charu Sharma

Co-Investigators: Amber Habib, Sunil Bowry

Funding Agency: National Board for Higher Mathematics

Faculty

Dean Undergraduate Studies, HoD Mathematics, HoD Computer Science & Engineering, Director IIIMIT
Assistant Professor