1

ECON 424 Fall 2016

Econ 424: Time Series Economics, Fall 2016

Prof Martin Evans

Overview:

The course introduces students to the econometric techniques needed to study time series data. The emphasis of the course is on the practical application of techniques rather than econometric theory. To this end, students will study how the econometric techniques have been used to study a variety of topics in macroeconomics, international finance, and financial economics.

Pre-Requite:

All students must have completed ECON 122 before taking this course.

Textbook and Readings:

The textbook for the course is Applied Econometric Time Series by Walter Enders. Students may use either the 3rd. or 4th. editions.

In addition, some students may find an unpublished textbook by John H. Cochrane (2005) “Time series for macroeconomics and finance”, useful. It is available at:

Readings associated with specific topics will be made available during the course.

Software:

The course will make extensive use of a programming language called Gretl. This open-sourcesoftware allows the user to perform a huge range of tasks in time series econometrics. It can be downloaded at: how to program in Gretl will be an integral part of the course. There are a very useful set of notes describing how to use Gretl by Lee C. Adkins available at:

Evaluation:

The grade for the course will be based on bi-weekly homework (50%) and a take-home final exam (50%). Homework will usually involve programming in Gretl. Students working in teams of two should complete homework. Teams will be asked to present their homework to others in class. Each student must complete the take-home final exam individually. Grades will be awarded in a manner consistent with the Economics Department grading recommendations.

Honor Code:

Students are expected to strictly adhere to the Georgetown University honor code. Any student found violating the code on either their homework assignments or on the final exam, will receive zero points for the homework or final exam, respectively.

Provisional Outline

  1. Overview:
  2. Stochastic Processes
  3. Time Series Models
  4. Univariate/Multivariate
  5. Linear/Nonlinear
  6. Uses:
  7. Impulse Responses
  8. Forecasting
  9. Testing Theory

Reading: Enders Ch. 1, Cochrane Ch 2

  1. Introduction to Gretl (review ECON 122)
  2. Reading Data
  3. Transforming Data
  4. Running Regressions
  5. Testing Hypothesis

Reading: Adkins Chs. 1, 2, (not 2.5) 3 (not 3.2).

  1. Programing in Gretl
  2. Basic matrix Algebra
  3. Logic control Loops
  4. Bootstrapping sampling distributions
  5. Finite power and Size of Tests

Reading: Adkins Chs. 2.5 and 3.2.

  1. Properties of Univariate Time Series
  2. White Noise, Autocovariance, Autocorrelation
  3. AR, MA, ARMA, ARIMA
  4. Estimation (Kalman Filter)
  5. Forecasting

Reading: Enders Ch. 2 (Cochrane Ch. 3 optional)

Application: Estimating the process for US GDP

  1. Regression Models with Time Series
  2. Estimation Theory
  3. Short-run vs. Long-run effects
  4. Serial Correlation

Reading: TBD

Applications: Rational Expectations, Term Structure Models, UIP Models

  1. Multivariate Time Series Models (VARs)
  2. Estimation, Lag length
  3. Identification. (short-run, long run)
  4. Impulse responses, variance decompositions, historical decompositions.
  5. Forecasting, Granger Causality

Reading: Enders Ch 5 (Cochrane Ch. 7 optional)

Applications: Drivers of Business Cycles, Present Value Models

  1. Trends:
  2. HP filter
  3. BN decomposition
  4. StructuralTS models
  5. Unit root Tests

Reading: Enders Ch 5 (Cochrane Ch. 10 optional)

Applications: Secular Stagnation

  1. Cointegration and Stochastic Trends
  2. Tests
  3. Cointegrating Regressions
  4. Error-Correction Models

Reading: Enders Ch. 6 (Cochrane Ch. 11 optional)

Applications: Fisher Hypothesis, Present Value Relations

  1. Nonlinear Models (time permitting):
  2. ARCH/GARCH
  3. Regime Switching
  4. Threshold Models

Reading: Enders Ch. 3

Applications: Optimal Portfolio Allocation