Time Series Analysis for Finance
From classical methods to TVP-VAR models
This module covers the full spectrum of time series methods used in financial econometrics, from classical decomposition and exponential smoothing through ARIMA, GARCH, VAR, cointegration, and time-varying parameter models.
Sessions¶
| # | Topic | Key Concepts |
|---|---|---|
| 1 | Introduction to Time Series | Stationarity, autocorrelation, white noise, stylized facts |
| 2 | Decomposition and Smoothing | Trend-cycle-seasonal decomposition, moving averages, HP filter |
| 3 | Exponential Smoothing (ETS) | Simple, Holt, Holt-Winters, state space representation |
| 4 | ARIMA Fundamentals | AR, MA, ARMA, integration, Box-Jenkins methodology |
| 5 | Advanced ARIMA | Model selection, diagnostics, forecasting, SARIMA |
| 6 | GARCH Models | Volatility clustering, ARCH, GARCH, estimation |
| 7 | Advanced GARCH | Asymmetric models, GJR, EGARCH, forecasting volatility |
| 8 | Vector Autoregression (VAR) | Multivariate systems, Granger causality, impulse responses |
| 9 | Cointegration and VECM | Engle-Granger, Johansen test, error correction models |
| 10 | TVP-VAR Part I | Time-varying parameters, Kalman filter, state space models |
| 11 | TVP-VAR Part II | Estimation, stochastic volatility, connectedness measures |
| 12 | Integration and Applications | Combining methods, applied financial forecasting |
How to use these notebooks¶
The course follows a natural progression from univariate to multivariate methods, and from constant to time-varying parameter models. Each session includes mathematical foundations, Python implementations, and empirical applications to financial data.