I am thinking about long-term impact on consumption by Covid in terms of loss of taste and smell.
I read medical study that sees up to 18% of patients do not recover normal taste and smell even long after recovery. Many even cannot stand the smell of coffee anymore and thus stop/reduce drinking it.
I wonder what will these potentially millions of people, who can no more enjoy cup of coffee, do with consumption in 2021.
For sure consumption forecast is very difficult to make for this year.
However, it is in my opinion very intelectually-lazy to just say that all scenarios are equally possible, because IMO, the probability distribution is different.
For proper modeling, a probability should be always estimated for each prediction.
Of course, that is the tricky part, but still, I would like to see what are the probability guesses when scenarios are used (if nothing more it reveals the bias).
My take is:
Consumption decrease by 2% ..... probability 35%
Consumption stagnation .... probability 40%
Consumption rise by 2% .... probability 20%
Consumption rise by 5% .... probability 3%
Consumption rise by 9% .... probability 2%
Estimation calculated based on above probabilities: rise by 0.03% (basically stagnation).
There is discussion about the number of scenarios to be made, some suggest that 3-4 scenarios is enough, others say that the more scenarios the better grasp of the complexity there is.
But, that is basically how modeling and quantifying of various scenarios is done.
Of course, it is just an exact calculation of some guestimates, but at least it provides one final number that reflects the bias of the creator.
(Sure, all the partial numbers are debatable)
Disclaimer, I tend to be cautiously long for near future, but short long-term.
« Back to index