A log-linear trend model is estimated on annual euro-area GDP using data from 1995 until 2018. The estimated model is lnRGDPt=−18.15+.0136t+ϵt, and the estimated standard deviation of ϵt is 0.0322. Assuming the shocks are normally distributed, what are the point forecasts of GDP for the next three years? How do these compare with a linear model RGDPt=−234178.8+121.3∗t+ϵt?
选项:
解释:
In this case,
ET[YT]=exp(ET[lnYT]+σ2/2)
And the error bounds on the ln are +/-1.96*0.0322, so the bounds are given in proportional terms rather than fixed values.