Image credit: Windows theme

Statistical Learning of Financial Markets spring 2020

This course is designed for Ph.D. students in Analytics and Data Science at Kennesaw State University. We study statistical learning methods and their applications in financial data analysis.

Textbook

  1. The Elements of Statistical Learning by Trevor Hastie, Robert Tibshirani, and Jerome Friedman. Springer, 2nd edition 2009.
  2. Analysis of Financial Time Series by Ruey S. Tsay. Wiley, 3rd edition 2010.

The first book (ESL) discusses the fundamentals of statistical machine learning theory, and it is a popular textbook used in many universities. The second book (FTS) provides a clear explanation of many standard methods in financial econometrics, making it a useful reference for financial data modelling. Combining the two will, hopefully, give students necessary technical skills to conduct research in financial data modelling.

Tentative schedule

Date Topic Chapter

week 1

  • optimization
  • numerical optimization
  • time series basics

FTS Chapter 2
lecture notes

week 2

Overview of Statistical Learning

ESL Chapter 2
slides pdf
Tradeoff between linear and fully nonparametric methods; function approximation; bias-variance tradeoff

week 3

Empirical Asset Pricing

FTS Chapter 9
slides pdf
Asset pricing test; single and multi-factor models

week 4

Empirical Asset Pricing

FTS Chapter 9
slides pdf
Macroeconomic factor pricing; fundamental factor pricing; statistical factor pricing

week 5

Linear Methods for Regression

ESL Chapter 3
slides pdf
Linear model; subset selection; ridge regression

week 6

Linear Methods for Regression

ESL Chapter 3
slides pdf
lasso; least angle regression; partial least squares method; applications in asset returning modelling

week 7

Linear Methods for Classification

ESL Chapter 4
slides pdf
Linear and quadratic discriminant; logistic method; separating hyperplanes; applications in bond credit rating

week 8

Basis Expansion and Regularization

ESL Chapter 5
slides
Polynomials and splines; selection of smoothing parameters

week 9

Model Selection

ESL Chapter 7
slides
Model complexity; in-sample prediction error; effective number of parameters; cross-validation

week 10

Volatility Models

FTS Chapters 3 and 10
slides
ARCH and GARCH modeling; stochastic volatility model; multivariate volatility model

week 11

Model Inference and Averaging

ESL Chapter 8
slides
Bootstrap; Bayesian methods; the EM algorithm; bagging

week 12

Additive Models and Trees

ESL Chapter 9
slides
Generalized additive model; tree-based methods; applications in asset return modeling

week 13

Boosting and Additive Trees

ESL Chapter 10
slides
Boosting linear models; boosting trees; model regularization

week 14

Value at Risk

FTS Chapter 7
slides
RiskMetrics approach; econometric approach to multi-period VaR forecast; quantile estimator

week 15

Support Vector Machines

ESL Chapter 12
slides

week 16

Random Forests

ESL Chapter 15
slides
Algorithm description; overfitting; variance and de-correlation effect; application in financial time series modeling

week 17

project presentation

Problem Set

Problem sets are assigned on a regular basis and they are graded.