If I were draft such a document it would probably include some of the following: > Previous historical STR business success/numbers (or other relevant REI experience)> Exit plans for property if they were restrict STR's and ability to liquidate or convert to standard rental> Revenue and Occupancy rate projections using industry 'approved' sites like Airdna.co>Cash flow projects with proof of expense amounts citing publicly available data In the end, the lender wants something predictable as a lease would give them on a standard rental purchase. 

--An algorithm will generally try to fit a distribution as a whole rather than predict each of its parts individually.

Huntsville is predicted to surpass Birmingham as the largest city in Alabama by 2024.

Your prediction (no matter how well founded) must come to pass if you're to make money.

Does anyone have any actual data to evaluate that might predict this? 

The other alternative in that range would be muni bonds.FYI...people have been predicting a recession on these forums for the past five years.

That in itself is unsustainable so something will change.Some predict that it is an indication of recession - however, per IBD, the last 8 times this has occurred 4 times resulted in recession, 2 times the market remained flat for over several years, and 2 times the market continued with a bull run.A predictor that SOMETHING will change - yes.

What happens if the storm clouds move in and all the predictions made in years past said only sunny weather?  

How can I predict changes in the laws and changes in the area and any type of changes in a neighborhood?

Hope it changes sooner than what I've predicted but I dont think it is likely.