Forecasting Zoom link - Shared screen with speaker view
Who can see your viewing activity?
sorry I had problem signing in fully. I could just listen. so be it.
Good to have you here Kajal - feel free to chat if you have comments/questions
To Prakash: When you talk about forecast efficiency, do you refer to average forecasts or individual forecasts? This distinction is very important.
To Prakash: Two questions: (1) Are you capturing more that the pandemic recession is different than past recessions or are you saying that forecasts are different this time even conditional on the recession type? (B) your historical analysis ignored the ‘WEO updates’ so the June outlier would not be captured as much if the historical analysis was identical to current
To Prakash: Great talk, and thanks. I tried to connect your findings to my REStat paper on how professional forecasters respond to large shocks, such as natural disasters. What we find is consistent with your results. During normal times, average forecasters do not pay much attention, leading to underreaction to news. Following large shocks, however, they pay close attention and their forecasts are close to FIRE after the largest shock.
Prakash - a thought provoking evidence on overshooting.
The optimism seems to be associated with countries that had major outbreaks late in the year… Especially for India
including next year forecast dispersions in addition to current year.
Speaking of real-time estimates, is there any concern that there may be large revisions in the actual GDP values (more so than in the past)?
@Andrew - that's exactly the debate Ana Galvao and I have been having about if the revisions will be similar in size to the past (linear) or should we expect them to be scaled by the preliminary estimates