CBE Forecasting 4

Question 1

Which of the following best describes a stationary time series?

Accepted Answer

A series with constant mean, variance, and autocovariance over time

Answer

A series whose mean and variance change systematically over time

Answer

A series with a deterministic linear trend

Answer

A series that exhibits strong seasonality

Question 2

A company's sales forecasts show that actual sales are always below the forecast. This suggests the forecasts are:

Accepted Answer

Overestimates (positively biased)

Answer

Efficient

Answer

Unbiased

Answer

Underestimates (negatively biased)

Question 3

Holt's double exponential smoothing extends simple exponential smoothing by also smoothing:

Accepted Answer

The trend component

Answer

Seasonal indices

Answer

Irregular fluctuations

Answer

Cyclical components

Question 4

Theil's U statistic is used to evaluate a forecast by comparing it to:

Accepted Answer

A random walk (naïve) forecast

Answer

An OLS regression benchmark

Answer

The Delphi panel consensus

Answer

A seasonal index-adjusted baseline

Question 5

In regression forecasting, multicollinearity primarily causes problems by:

Accepted Answer

Making individual coefficient estimates unstable and hard to interpret

Answer

Inflating forecast intervals for in-sample data

Answer

Producing biased forecasts on average

Answer

Eliminating autocorrelation in residuals

Question 6

Which approach is most appropriate for forecasting a non-stationary economic series with a stochastic trend?

Accepted Answer

ARIMA with differencing (d ≥ 1)

Answer

Simple moving average

Answer

Seasonal index decomposition only

Answer

A deterministic trend regression

Question 7

Prediction intervals in forecasting are wider than confidence intervals for the mean because they:

Accepted Answer

Also account for the variance of future individual observations

Answer

Ignore parameter uncertainty entirely

Answer

Use a lower significance level

Answer

Assume the model is perfectly specified