Econometric
Autor: Hussain133 • October 10, 2016 • Exam • 1,543 Words (7 Pages) • 668 Views
SAMPLE QUESTION
- Given below is the ACF and PACF plot of Apple stock returns estimated by a researcher
[pic 1]
[pic 2]
- what should be the order of ARIMA model
Though ACF and PACF suggest either AR(11) or AR(24), but based on principle of parsimony one may consider AR(4).
- The researcher estimated AR(p) Models as follows for Apple Stock returns:
AR(1) | Coefficients: ar1 intercept -0.0056 0.0011 s.e. 0.0254 0.0006 sigma^2 estimated as 0.0005566: log likelihood = 3612.84, aic = -7219.68 Box-Ljung test data: ar1$residuals X-squared = 39.4241, df = 20, p-value = 0.005902 |
AR(2) | Coefficients: ar1 ar2 intercept -0.0059 -0.0498 0.0011 s.e. 0.0254 0.0254 0.0006 sigma^2 estimated as 0.000555 log likelihood = 3614.76, aic = -7221.53 Box-Ljung test data: ar2$residuals X-squared = 32.654, df = 20, p-value = 0.03681 |
AR(3) | Coefficients: ar1 ar2 ar3 intercept -0.0055 -0.0497 0.0075 0.0011 s.e. 0.0254 0.0254 0.0254 0.0006 sigma^2 estimated as 0.0005552: log likelihood = 3614.81, aic = -7219.61 Box-Ljung test data: ar3$residuals X-squared = 32.6176, df = 20, p-value = 0.03715 |
AR(4) | Coefficients: ar1 ar2 ar3 ar4 intercept -0.0059 -0.0464 0.0079 0.0678 0.0011 s.e. 0.0253 0.0253 0.0253 0.0254 0.0006 sigma^2 estimated as 0.0005527: log likelihood = 3618.35, aic = -7224.7 Box-Ljung test data: ar4$residuals X-squared = 22.2278, df = 20, p-value = 0.3283 |
Without calculating t-statistics, answer intuitively which coefficients are significant across all the models
...