ECON 2310 Nonstationary Time Series

0.0(0)
Studied by 0 people
call kaiCall Kai
learnLearn
examPractice Test
spaced repetitionSpaced Repetition
heart puzzleMatch
flashcardsFlashcards
GameKnowt Play
Card Sorting

1/71

encourage image

There's no tags or description

Looks like no tags are added yet.

72 Terms

1
New cards

so far we have assumed the data are stationary


2
New cards

2 important types of nonstationarity


3
New cards

what is a trend

A trend is a persistent, long-term movement or tendency in the data. Trends need not be just a straight line

4
New cards

does this have a trend?



5
New cards

is this a trend?



6
New cards

is this a trend?



7
New cards

deterministic trend


8
New cards

stochastic trend


9
New cards

random walk


10
New cards

2 features of random walk


11
New cards

stochastic trend feature 1


12
New cards

stochastic trend feature 2


13
New cards

random walks diagram



14
New cards

random walk with drift model


15
New cards

the drift is…


16
New cards

The random walk model (with or without drift) is

The random walk model (with or without drift) is a good description of stochastic trends in many economic time series.

17
New cards

there is practical advice:


18
New cards

Upcoming specifics that lead to this advice


19
New cards

random walk with drift vs AR(1)


20
New cards

random walk with drift vs AR(1) notes


21
New cards

unit root terminology


22
New cards

unit roots in AR(2)


23
New cards

unit roots in AR(2) rearrange regression


24
New cards

unit roots in AR(2) subract Yt-1 from both sides


25
New cards

unit roots in AR(2) simplify


26
New cards

AR(2) can be rewritten as


27
New cards



28
New cards

unit root in AR(2) claim shows that


29
New cards

problem 1 caused by trends

AR coefficients are strongly biased towards zero leading to poor forecasts. Some t-statistics don’t have a standard normal distribution, even in large samples

30
New cards

notes on problem 1

If a regressor has a stochastic trend, then its usual OLS t-statistic can have a nonnormal distribution under the null hypothesis, and the estimate of the autoregressive coefficient is biased toward 0

– This nonnormal distribution means that conventional confidence intervals are not valid and hypothesis tests cannot be conducted as usual

31
New cards

problem 2 caused by trends

If two time series both have random walk trends, they can look related even if they are not – “spurious regressions”

32
New cards



33
New cards

How do you detect stochastic trends?

  1. plot data

  2. Dickey Fuller test


34
New cards

plot data

are there persistent long-run movements?

35
New cards

dickey fuller test


36
New cards

Detecting Stochastic Trends: Testing for a Unit AR Root


37
New cards

Detecting Stochastic Trends: Testing for a Unit AR Root notes


38
New cards

We need to use the table of Dickey-Fuller critical values.

There are two cases, which have different critical values:


39
New cards

dickey fuller table


40
New cards

dickey fuller table notes


41
New cards

When should you include a time trend in the DF test?

The decision to use the intercept-only DF test or the intercept & trend DF test depends on what the alternative is – and what the data look like.


42
New cards

a-intercept and b-intercept/time trend specifications


43
New cards



44
New cards



45
New cards

The second type of nonstationarity we consider is that


46
New cards

It is a problem for forecasting if the model describing the historical data


47
New cards

nonstationarity breaks outline


48
New cards

breaks can arise from


49
New cards



50
New cards



51
New cards

case 1 ADL(1,1) case


52
New cards

case 1 ADL(1,1) case notes


53
New cards



54
New cards

chow test stat notes


55
New cards



56
New cards



57
New cards



58
New cards

QLR stat notes


59
New cards

Quandt Likelihood Ratio (QLR) Statistic


60
New cards

Quandt Likelihood Ratio (QLR) Statistic notes


61
New cards

Quandt Likelihood Ratio (QLR) Statistic graph


62
New cards

Quandt Likelihood Ratio (QLR) Statistic second graph


63
New cards



64
New cards



65
New cards

dependent variables and regressors


66
New cards

notes


67
New cards

graph


68
New cards

QLR =


69
New cards



70
New cards

Detecting Breaks Using Pseudo Out-of-Sample Forecasts cont


71
New cards



72
New cards