Originally Posted by PUCCI GALORE
what always surprises me is that there is such variation in the forecasts given that they are - presumably - working with the same data?
actually no they arent
number of different models used, each with different granularity (level of detail) and different approach (some single forecast, others best forecast of a series using slightly altered data, some start at present time and forecast forwards only others start a few days ago and forcast forwards, etc)