Bill, thanks so much for the input. The replacement value could be great.
After speaking with my agent and a couple of mentors as well, I have the following answers:
The property tax assessment is usually very conservative, especially in high value places like SoCal where some of my properties are. They'll put it at 5% building value which is not quite reasonable (ex. a $1m property would be $50k value in the building? no...), so that may be difficult to rely on in these scenarios.
Defaulting to something like 80/20 (building/land) could be a good approach, but you need to modify it depending on location. Somewhere like SoCal again would be too much there, (ex. $1m property also isn't $800k in building value).
The 2 recommended options are DIY modeling or Cost Segregation:
DIY Modeling: Come up with some expression you could reasonably defend in county court if you get audited somehow. One idea is 80/20 which could apply, but maybe in your area it's 70/30. Another option could be coming up with the building's replacement value. Another option could be looking up comps for raw land and new construction and combining that somehow. Again, just make it something you could earnestly try to defend in a court.
Cost Segregation: This is probably the best option, especially if you are likely to be selling properties every 5-10 years. See here for a definition but the TlDr is that a professional firm values each component of the building and land and helps you define fine grain depreciation schedules for each piece of the property. This isn't cheap, but 15k today to save an extra 30k in taxes over the next couple of years could be very worth it!
My team is planning on using a DIY model for a year or two, especially as we are doing some improvements to the properties. Then we will get a cost seg study.
