Quants

one of the question was about the implications of the suggessions made and one of the answer was “Beller model specification” I just blindly picked this one which was the first answer to the question. I am sorry that I couldn’t give moe info.

I forgot the other answers but the suggested reasoning was oil prices instead of oil changes…but there was something weird about oil prices…maybe it was non stationary? I think the answer was model misspecification … anyone remember the other choices?

I chose model misspecification because oil prices were not covariance stationary.

it was non economic sense , i guess, the oil price alone has nothing to do with the inflation, it is the change in the price only that makes economic sense

Yes … not covariance stationary … and since they didn’t give you anything on cointegration it would lead to bad t stat results

This is not time series so no requirement for covariance stationary. Am I missing something?

She said something like “since the levels of prices are rising over time, the trend is not covariance stationary” and that the model should be changed to a time series for that reason. I think that the point was that if it was changed to a time series model like she suggested that it would get invalid results.