Introduction to Time Series Econometrics with Eviews UNIT

  • Slides: 25
Download presentation
Introduction to Time Series Econometrics with Eviews * UNIT ROOT Tests * [Mainly compiled

Introduction to Time Series Econometrics with Eviews * UNIT ROOT Tests * [Mainly compiled from GP]

TS Data 2

TS Data 2

Note: A random or stochastic process is a collection of random variables ordered in

Note: A random or stochastic process is a collection of random variables ordered in time. 3

4

4

RW 5

RW 5

6

6

7

7

EXAMPLE from Towards a More Realistic Time Series Analysis of Unemployment Dynamics: Dynamics the

EXAMPLE from Towards a More Realistic Time Series Analysis of Unemployment Dynamics: Dynamics the Dual Adjustment Approach* By M. İsmihan * Paper presented to the Fourth Nordic Post-Keynesian Conference-Economics at the Edge", 20 th - 21 st of April 2017, Aalborg University, Aalborg, Denmark. 8

Natural Rate vs. Hysteresis H. • Hysteresis (in economics) generally • Milton Friedman “argued”

Natural Rate vs. Hysteresis H. • Hysteresis (in economics) generally • Milton Friedman “argued” that there ‘used to mean that, “after a shock, a is a unique natural rate of variable does not return to its initial unemployment which is consistent value, even when the shock has gone with price stability. away”. ’ BUT there are different • Natural Rate is determined as “the definitions (Setterfield, 2010) level ground out by the Walrasian • Accr. to Blanchard and Summers (1988) : system of general equilibrium “What is required is a theory of equations. . . including market unemployment in which unemployment, imperfections. . . the cost of gathering far from returning to a stable information about job vacancies and equilibrium -or "natural rate"- over time, labor availabilities, the costs of is instead strongly dependent on mobility, and so on” (Friedman 1968) history. We refer to such an equilibrium [ “Similar” arguments from Edmund as a "fragile equilibrium, " to highlight Phelps] the sensitive dependence of unemployment on current and past They earned Nobel Prize events. ” “FIXED” or ahistorical EQUILIBRIUM HISTORY -DEPENDENT EQUILIBRIUM

Main focus shifted towards univariate TS properties of U [UNIVARIATE APPROACH] Most of the

Main focus shifted towards univariate TS properties of U [UNIVARIATE APPROACH] Most of the empirical studies in the existing literature have focused on the univariate time series properties of the unemployment rate. While the presence of unit roots in the unemployment rate series are considered to be evidence in favour of the hysteresis hypothesis, its rejection (i. e. stationarity of unemployment rate) is taken as a validation of the unique natural unemployment rate.

Univariate [Unit Root] Approach

Univariate [Unit Root] Approach

Does unemployment rate have a unit root ? Ut = 1. 60 + 0.

Does unemployment rate have a unit root ? Ut = 1. 60 + 0. 97 Ut-1 - 0. 25 Ut-2 Yes/No! 1%/5% Dickey Fuller

Does unemployment rate have a unit root ? Answer is usually “yes” for many

Does unemployment rate have a unit root ? Answer is usually “yes” for many countries

Univariate Approach (G 6 – evidence)

Univariate Approach (G 6 – evidence)

Stochastic Vs Deterministic Trends 15

Stochastic Vs Deterministic Trends 15

I (d) 16

I (d) 16

17

17

18

18

19

19

DF Tests: Alt. Spec. 20

DF Tests: Alt. Spec. 20

LGDP: DF Tests 21

LGDP: DF Tests 21

22

22

23

23

LGDP: ADF Tests 24

LGDP: ADF Tests 24

25

25