Stationarity And Moving Averages Flashcards
Stationary def
In the same shape/ distribution for long time - 10 years
Week stationary properties
Constant mean
Constant lag
Constant variance
White noise process
Constant mean and variance
No pattern
Box and pierce
Test for joint hypothesis that the lag of the t coefficients are equal to 0
Using q statistic formula
Difference between box and pierce and ljung box
Ljung box- bigger sample properties
Moving average
Used to structure returns where there’s autocorrelation
Autoregressive
Contains lags of dependent variable
Has stationary condition- roots lie outside the unit circle
Arma
Mix of moving average and autoregressive models
How to build arma
Use box and Jenkins- identify, estimate, check
Test for autocorrelation
Think lagged:
Ut= put_1 +vt