FRM Part I · FRM Exam Part I · Nonstationary Time Series
An analyst applies the Augmented Dickey-Fuller (ADF) test instead of the basic DF test to a quarterly series. What is the purpose of the added lagged difference terms?
The lagged difference terms soak up serial correlation in the errors, which would otherwise distort the basic Dickey-Fuller test. The null is still a unit root and the nonstandard critical values still apply; the augmentation only corrects for higher-order dynamics.
- ATo absorb serial correlation in the residuals so the test statistic is validCorrect
- BTo convert the null hypothesis into one of stationarity
- CTo remove the need for special critical values
- DTo guarantee that the series is stationary after differencing
Explanation
If the series follows a higher-order autoregression, the DF residuals are autocorrelated and the test is distorted. Adding lagged ΔY terms captures this dynamic. The null remains a unit root and the nonstandard DF-type critical values still apply.
Did you get it right without looking?
One question tells you little. A timed set on Nonstationary Time Series shows your real accuracy, how long you take and where you lose marks.
More Nonstationary Time Series questions
- A quarterly series is modeled as Y_t = 50 + 0.5t + 8·Q2_t + 3·Q3_t − 12·Q4_t + e_t, where Q1 is the base quarter and t = 1 in the first quar…
- An analyst models a monthly price index as y_t = y_{t-1} + e_t, where e_t is white noise with variance 4. Which statement about this process…
- Two models are fitted to the same series. Model A: y_t = a + bt + e_t with e_t white noise. Model B: y_t = b + y_{t-1} + e_t with e_t white …
- An analyst runs an augmented Dickey-Fuller test on the level of a series and obtains a statistic of -1.50 against a 5% critical value of -2.…
- An analyst models monthly log sales of a firm as ln(Y_t) = 2.50 + 0.012 t + e_t, where t = 1 for the first month and e_t is a stationary, me…
- A researcher regresses one random walk on an independent random walk and finds R² = 0.85 and a t-statistic of 14 on the slope, with a Durbin…