market projection under uncertainty

[Charles Grace]

Hi everyone,

Please excuse this intrusion and ignore me if I'm becoming a pain.

I was recently trying to project market size for a retail inventory
factoring firm in NYC. We were trying to project the market opportunity
for them in men's clothing stores, where the inventory mix varies store
to store and season to season, but with some predictability. I also
thought it might be a easy case for a class I'm teaching in the Fall. I
can't share those numbers, but I can share a similar one.

The Crop Protection market in a Romania (formerly called Pesticides).
The market size is function of:
- arable land area in hectares,
-the hectares adjusted to consider only larger farms that are core customers
- % hectares for each of top 7 crops in Romania, the Crop Mix, which
varies year to year but is predictable a few months ahead via a normal
distribution for each crop.
- and a fixed price per hectare of crop protection product for each crop

Sounds easy enough to model with MonteCarlo and Yasai. The problem
arises when you GenNormal the crop area estimates for each crop and add
them up. The Sum of those estimates needs to be constrained somewhere
close to 100%, otherwise you get an unnecessarily wide standard
deviation in the Simoutput.

I've tried a couple of solutions in the attached spreadsheet (Input
Page) with no joy. They are described in detail on the page. The Output
pages match to each equation that I tried.

Any suggestions are welcomed. Note that the Simoutput formulas are
presently 'labels' and you will need to remove the apostrophe to make
them active. Thanks in advance for your interest.

Charles Grace

[Jonathan Eckstein]

I am busy with final exams etc., but will try to look this toward the
end of next week.

J E

[Charles Grace]

Thanks, Take your time. I'm out of pocket next week cycling across the
Dakotas.

CG