The Econometrics of High Frequency Data

FIN524 The Econometrics of High Frequency Data

  • Topics

    Topics

     Topics

     

    Introduction (Section 1-2.3 in lecture notes)

    Preliminary in Stochastic Calculus Section 2.3-2.6)

    Behavior of Estimators (Section 3)

    Asymptotic Normality (Section 4)

    Microstructure and Applications (Section 5)

     

    Each topic will be covered in approximately one day, in two sessions of two hours each, with lectures and problem solving/data analysis.

     

     

  • Teaching

    Teaching

     

     

     

     

     

  • Requirements for course approval

    Requirements for course approval

     

     

  • Assessment

    Assessment

    Course project (pass/fail)

     

     

     

  • Grading Scale

    Grading Scale

    Pass/Fail.

  • Objective/course outline

    Objective/course outline

    This is a course on modeling and estimation for high frequency financial data, i.e. data observed more frequently than daily and down to the individual transaction. It is designed for an audience that includes people interested in finance, econometrics, statistics, probability and financial engineering.

     

    In recent years there has been a vast increase in the amount of high frequency data available in finance. Their analysis may require methods different from the common ones for time series of regularly spaced data, and there has been an explosion in the literature on the subject. In this course, we start from scratch, introducing a probabilistic model for such data, and then turn to the estimation question in this model, with main emphasis on estimating volatility. Similar techniques to those we present can be applied to estimating leverage effects, realized regressions, semi-variances, doing analyses of variance, detecting jumps, measuring liquidity by measuring the size of the microstructure noise, and many other objects of interest. The applications are mainly in finance, ranging from risk management to options hedging, execution of transactions, portfolio optimization and forecasting. Methodologies based on high frequency data can also be found in neural science and climatology.

     

    Prerequisites: Basic mathematical sophistication, for example, courses in probability theory and stochastic process, in statistical inference, and finally a course in real analysis, or measure theory, or stochastic calculus.

  • Semester

    Semester

    Spring.

     

  • Literature

    Literature

    - The course will be based on lecture notes by Mykland and Zhang (2010). It assumes some literacy in probability and statistics (see prerequisites), similar to the basis for applying stochastic calculus in finance, but the notes is moderately self-contained. Some of the material is research front and not published elsewhere.

     

    - "A tale of two time scales: Determining integrated volatility with noisy high frequency data¿. Zhang, Mykland, and Aït-Sahalia, Journal of The American Statistical Association, 100 (472), 1394-1411, December 2005. (Downloadable from Per Mykland¿s website).

     

    - "Inference for continuous semimartingales observed at high frequency: A general approach¿. Mykland and Zhang, Econometrica, 77 (5), 1403-1445, 2009. (Downloadable from Per Mykland¿s website).

     

     

     

Overview

ECTS Credits
5
Teaching language
English.
Semester
Spring, Autumn

Course responsible

Per A. Mykland and Lan Zhang