IAI Actuarial Core Principles · Risk Modelling and Survival Analysis · Core concepts of time series models
A series shows a clear upward trend and a sample ACF that decays very slowly from values near 1. The first differences of the series have an ACF that is negligible beyond lag 1, with a decaying PACF. Which model is most appropriate for the original series?
The appropriate model is ARIMA(0,1,1). The slowly decaying ACF indicates one difference is needed, and the differenced series shows an MA(1) pattern, with ACF cutting off after lag 1 and a decaying PACF.
- AARMA(1,1)
- BARIMA(0,1,1)Correct
- CARIMA(1,1,0)
- DAR(1) with phi near 0.5
- ARIMA(0,2,1)
Explanation
The slowly decaying ACF suggests non-stationarity, so difference once (d=1). The differenced series has ACF cutting off after lag 1 and decaying PACF, which is an MA(1). The model is ARIMA(0,1,1).
Did you get it right without looking?
One question tells you little. A timed set on Core concepts of time series models shows your real accuracy, how long you take and where you lose marks.
More Core concepts of time series models questions
- A time series analyst inspects the sample ACF and PACF of a stationary series. The sample ACF decays gradually towards zero, while the sampl…
- The stationary AR(1) process X_t = 0.8 X_{t-1} + e_t has e_t white noise with variance 9. A stationary ARMA(1,1) process is not considered. …
- A stationary AR(1) model is X_t - 50 = 0.6(X_{t-1} - 50) + e_t, with white noise variance 16. The latest observation is x_100 = 60. What is …
- Let e_t be white noise with variance 9. A process is defined by Y_t = e_t + 0.4 e_{t-1}. Using the autocorrelation at lag 1 of this process,…
- For the MA(1) process X_t = e_t + 0.5 e_{t-1}, where e_t is white noise with variance sigma^2, what is the autocorrelation at lag 1?
- A stationary, zero-mean AR(2) process X_t = 0.5 X_{t-1} + 0.2 X_{t-2} + e_t is considered. What is the autocorrelation at lag 1, rho_1?