INTRODUCTION to TIME SERIES ANALYSIS
MASTER in INDUSTRIAL ECONOMICS 2009
Question: Is time important? Can time explain anything?
"Once upon a time....." "From time to time...." "El tiempo lo cura todo (time cures everything)..." "What time is it?" "Time is money (el tiempo es oro)"..."El tiempo dira (time will say)..." And now it is time for the course to start: On your marks!!, Ready!!!, Goooo!!!!
Econometrics Uncertainty Principle:
- In order to study causality we need to keep certain things constant ("ceteris paribus")
- In order to study causality we need time to pass (there is not causality between simultaneous events)
- Nothing is constant through time
- Therefore ...........
|
|
|
A Guide to Write an Empirical Project Does Everything Grow? It Seems So: USTrendsPopulation USTrendsGovCrimeTransport USTrendsMoneyPolitics USTrendsEducationReligion USTrendsBusinessCommunication USTrendsLeisureHealth USTrendsWork [From "The First Measured Century" by T. Caplow, L. Hicks and B. Watenberg, PBS 2001] During the course we will play with two E-views working files: (1) GDP growth and interest rates: USA ( GDPIrates.WF) y UK (GDPIrates.WF) (2) Global warming: (AnnualGDPTemperatureUK.WF) (3) Maddison GDPpc: (GDPpcMaddison.WF) AND A GREAT PLACE for US DATA is the S.Louis FED (http://research.stlouisfed.org/fred-addin/install_windows.html) |
| Introduction Stochastic Processes Examples |
GapMinder
SnapShots ( Charles Jones ) Some thing to do: Invent a transformation that makes growing time series to look stationary. |
Reading
1 (Ergodicity) (from Breiman (1969) "Probability and Stochastic Processes: With a View Toward Applications") |
| Arma Models | It is very Important to understand the Wold Decomposition: Wold Decomposition (E-views.prg) Applets for Arma processes GenerateARMA (E-views.prg) |
Reading
2 (ARMA models)
(from Pollock "Lecture Notes in Time Series Analysis and Forecasting") Reading 3 (ARMA models) (from W.Wei "Time Series Analysis: Univariate and Multivariate Methods") |
Estimation and Inference Model Selection |
Estimation
(E-Views) MA-estimation (E-views.prg) Simulation Estimation AR (E-views.prg) Simulation Estimation MA (E-views.prg) |
Reading 4 (Estimation)
(from W.Wei "Time Series Analysis: Univariate and Multivariate Methods") Reading 5 (Identification) (from W.Wei "Time Series Analysis: Univariate and Multivariate Methods") |
| Forecasting
I and Forecasting
II Applied Example |
Forecasting
with E-Views A bit of humor Forecast Program (plug-in versus direct method) |
Reading
6 (Classical forecasting methods) Reading 7 (Jim Stock's paper on forecasting) Reading 8 (Forecasting) (from W.Wei "Time Series Analysis: Univariate and Multivariate Methods") |
Regression with autocorrelation (chapter 13 Jeff Woolridge's book) HAC Standard Errors |
Intervals Mean.prg, MonteCarlo Intervals Mean.prg HAC-comparisons via simulations |
| The
Land of Unit Roots Graphs1 Graphs2 Brief Introduction to Structural Breaks |
Applets on: Brownian Motion I Brownian Motion II Brownian Motion III Brownian Motion IV DF-TEST Program Structural Break Program NelsonPlosserExtendedData.WF1 ShillerYearlyData.WF1 |
Reading 9 (ARIMA models) (from W.Wei "Time Series Analysis: Univariate and Multivariate Methods") Reading 10 (D.S. Pollock on Trends) Reading 11 (Herman Bierenes on Unit Roots) |
| VAR
Models Nexus I (notes of Chris Sims on VARs) Nexus II (notes of Guido Kuersteiner on VARs) |
VAR
Models with E-Views Stock-Watson JEP2001 VARS Stock-Watson Data Set (Eviews) |
Reading
12 (VAR notes by Mark Watson) Reading 13 (Structural VAR notes by Eric Zivot) |
| Spurious Regression and Cointegration | Examples of Spurious Regression Spurious Regression Program Cointegration Program Engle-Granger Program GoldSilver.WF1 |
Reading
14 (Spurious Regression in Finance) Reading 15 (Spurious Regression with I(0)) |
| Cointegration
and Common Factors + Class-notes |
Cointegration-Examples
(Andrew Buck- Temple University) |
Reading
16 (Cointegration by Dolado, Gonzalo and Marmol) Reading 17 ("Common Long-memory Components") (J. Gonzalo and C.Granger, 1996) |