Applied Econometrics global edition_Chapter 2.pptx

Econometrics

1. Dimitrios Asteriou and Stephen G Hall

2. What is Econometrics? The Stages of Econometric Work INTRODUCTION

3. What is Econometrics? Econometrics means measurement (metrics in greek ) in economics. The importance of applied work in economics is increasing constantly. Theory suggests that X affects Y but is this true or not? This is the work of the applied econometrician.

4. What is Econometrics? Examples of problems that may be tackled by an Econometrician Modelling long-term relationships among prices and interest rates. Examining the effect of inflation in unemployment rates Examining the effect of disposable income on consumption

5. What is Econometrics? Examples of problems that may be tackled by an Econometrician (continued) Determining the factors that affect GDP per capita growth Testing the validity of the CAPM and APT theories Forecasting the correlation among the returns and the stock indices of two countries.

6. The Stages of Applied Econometric Analysis Economic Theory Econometric Model Data Model Estimation Specification Testing and Diagnostics Is the Model Adequate? No Yes Test any hypothesis Use for Predictions and Policy Making

7. The Structure of Economic Data There are three different types of economic data TIME SERIES CROSS SECTIONAL PANEL DATA

8. The Structure of Economic Data Time Series Data Examples: GDP, Unemployment, Inflation, Stock Prices, etc Vectors or Columns (like in a spreadsheet) Frequencies: Yearly, Bi-annually, Quarterly, Monthly, Weekly, Daily, Hourly. Lots of different values for different time periods for one country, state, city, market.

9. The Structure of Economic Data Cross-Sectional Data Examples: GDP, Unemployment, Inflation, Stock Prices, etc. Vectors or Rows (like in a spreadsheet) Frequencies: Yearly, Bi-annually, Quarterly, Monthly, Weekly, Daily, Hourly. Lots of different values for different countries, states, cities, markets, but for one time period only.

10. The Structure of Economic Data Panel Data A combination of time series and cross-sectional data. Examples: GDP, Unemployment, Inflation, Stock Prices, etc. Matrices (columns and rows to make an n times m matrix) Frequencies: Yearly, Bi-annually, Quarterly, Monthly, Weekly, Daily, Hourly. Lots of different values for different countries, states, cities, markets, and for different time periods.

11. The Structure of Economic Data - Notation Time series: Y t , t=1990, 1991, …, 2012 Cross-Sectional: Y i , i =1, 2, 3, …, 40 Panel Data: Y it , i and t defined as above. It is common to denote each observation by the letter t and the total number of observations by T for time series data, and to denote each observation by the letter i and the total number of observations by N for cross-sectional data.

12. 12 The Structure of Economic Data – Quantitative vs Qualitative The data may be quantitative and qualitative. Quantitative (e.g. GDP per capita, exchange rates, stock prices, unemployment rates) Qualitative (e.g. Day of the week, gender, level of education)

13. Basic Data Handling Looking at raw data Graphical Analysis Summary Statistics Components of a Time Series Indices and Base Dates Data Transformations

14. Basic Data Handling - Looking at raw data Before getting into the statistical and econometric tools, a preliminary analysis is extremely important “Get the feel” of your data Look at the numbers on a spreadsheet . Note number of series/end and start dates, range of values etc. Outliers, discontinuities, structural breaks etc.

15. Basic Data Handling – Graphical Analysis Graphs facilitate the inspection process See the “big picture” Histograms: give an indication of the distribution of a variable Scatter plots: give combinations of values from two series for the purpose of determining their relationship (if any) Line Graphs: facilitate the comparisons of series Bar Charts: good for comparisons Pie Charts: good for percentages/portions

16. Basic Data Handling – Histograms Histograms: give an indication of the distribution of a variable Command in Eviews (View\Histogram and Stats)

17. Basic Data Handling – Scatter Plots Scatter Plots: give combinations of values from two series for the purpose of determining their relationship (if any) Eviews command (open the two series together in a group and choose View/Graph/Scatter)

18. Basic Data Handling–Line Graphs Line Graphs: facilitate the comparisons of series Command in Eviews : Plot X Y

19. Basic Data Handling–Bar Charts Bar Charts: facilitate the comparisons of series Command in Eviews : View/Graph/Bar

20. Basic Data Handling–Pie Charts Pie Charts: Good for proportions Command in Eviews : View/Graph/Pie

21. Basic Data Handling – Summary Statistics Summary statistics provide a more precise idea of the distribution of a variable (mean, variance, st. dev. etc ) For comparisons open the variables in a group.

22. Basic Data Handling–Components of a Time Series An economic or financial time series consists of up to four components Trend (smooth, long-term/consistent upward or downward movement) Cycle (rise and fall over periods longer than a year) Seasonal (within year pattern seen in frequency data) Irregular (random component, episodic – unpredictable but identifiable – and residual – unpredictable and unidentifiable)

23. Basic Data Handling–Components of a Time Series An economic or financial time series consists of up to four components

24. Basic Data Handling–Components of a Time Series Seasonal (within year pattern seen in quarterly, monthly or weekly data)

25. Basic Data Handling–Components of a Time Series Trend (smooth, long-term/consistent upward or downward movement)

26. Basic Data Handling–Components of a Time Series Cycle (rise and fall over periods longer than a year)

27. Basic Data Handling–Components of a Time Series Irregular (random component, episodic – unpredictable but identifiable – and residual – unpredictable and unidentifiable)

28. Basic Data Handling – Indices and Base Dates Splicing two indices and change the base date 218 is 100 258 is X? X= 100*(258/218) or X=258*(100/218) or X=258/2.81

29. Basic Data Handling – Data Transformations Nominal vs Real data (use appropriate price deflator)

30. Basic Data Handling – Data Transformations Natural Logs ( ln ) have three advantages: Natural logs linearize the exponential trends in the series (makes the graph look smoother). Natural logs linearize a model that is non linear in parameters (consider Cobb-Douglas Production function). The OLS coefficients are elasticities .

31. Basic Data Handling – Data Transformations Differencing In case we want to remove the trend component from a time series entirely (i.e. in order to make it stationary) First differences: Δ Y t = Y t – Y t-1 Second diffferences : Δ 2 Y t = Δ ( Y t – Y t-1 ) = Δ Y t – Δ Y t-1 = = ( Y t – Y t-1 ) – ( Y t -1 – Y t - 2 ) = Y t – 2 Y t-1 + Y t - 2

32. Basic Data Handling – Data Transformations Growth rates Discretely compounded growth rate of Y t =(Y t -Y t-1 )/Y t-1 Continuously compounded growth rate of Y t = ln (Y t -Y t-1 ) = ln ( Y t )- ln (Y t-1 )