Time Series

Question 1

Dickey Fuller test

Answer 1

tests to see if |theta| is under 1

Question 2

Dickey Fuller test - problem

Answer 2

critical values must be found by simulation

Question 3

Augmented Dickey Fuller test

Answer 3

allows for more lags than the DF test

Question 4

Jarque Bera test

Answer 4

Tests null: returns are normally distributed against alternative: returns are not normally distributed

Question 5

Strong/strict stationarity

Answer 5

A process is strictly stationary if for any h,

Y₁, Y₂, Y₃

has the same distrubution as

Y_1+h, Y_2+h, Y_3+h

Question 6

Weak/covariance stationarity

Answer 6

A process is weakly stationary if

E(Y²_t) is finite for any t

E(Yt) = mu for any t

For any t, s, Cov(Yt, Ys) does not depend on t or s

Question 7

Any strictly stationary process

Answer 7

is also covariance stationary

Question 8

If a Gaussian process is covariance stationary

Answer 8

then it is also strictly stationary

Question 9

a covariance stationary process is white noise if

Answer 9

cov(Y_t, Y_s) = 0 for all t, s.