Term 2 lecture notes Co-integration Flashcards

1
Q

How is a stationary series integrated?

What does it mean if the change in Zt = I(0)

A

it means the series is I(o)
and integrated of order 1 means it needs to be differenced 0 times for it to be stationary.

It means the original series was I(1)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What is the definition of co-integration?

Why does this work?

A

Co-integrates if yt and xt sepeartely are cointegrated of order 1

Then the linear combination of them is integrated of order 0

-As long as they share the same stochastic trend.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

What is a method for testing co-integration?

What are the notes?

A

Engle and granger

Step 1 regress long run equation then save residuals

Then save residuals and use ADF test

if gamma = 0 non-stationary —– no cointegration as the linear sum is not I(0)

If gamma is less than 0 stationary and therefore it is stationary.

Find correct p as normal
-You must justify if you use a trend or not.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

How do you use the mckinnon tables differently when doing cointegration

A

As you are working out the difference in stationarity between residuals you therefore use the n = the number of I(1) eqautions.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

What does it mean if something is stationary?

A
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

What does the evidence in using static and dynamic models to test for cointegration

A

in finite samples it is better to solve for a dynamic model

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What happens to OLS estimate when you have cointegration compared to stationary series?

A

The OLS estimates are super consistent if they are co-integrated.

Stationary approaches real parameter at rate root T

Whilst cointegrated reaches at rate T

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Why is super-consistency in cointegration useuful?

A

If there is any mispspecifciation as t goes to inifinity it makes no difference.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

What is the error correction model?

What does it mean?

A

SR changes are a measure of disequilibrium last period

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

What is the issue with unrelated non-stationary series and why is this?

A

77% of the time you will find a correlation when there is no relation.

This is because of the stochastic trend.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Graphically what is the difference between sprurious and cointegrated

A

Spurious the residuals have random
Cointegrated they are stationary looking.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly