The policy is spelled out in the article? Banks have strict regulations that mean they have to have a certain amount of capital backing loans, and by revaluing a building you lower the capital that backs the loan, thus raising its risk, and thus leading you to break the regulation around capital requirements.

Here the bank cares less about annual income than future income.

Keeping it vacant only impact current income, lowering rent impacts future forecasts.

The argument the article makes is that the bank doesn't want to admit the property is worth less than the mortgage because then they would "have to" foreclose.

The question is, why would they actually do that? The premise is that the landlord has to take out a new mortgage every few years and then the bank won't give them a new one if they're underwater. But that's only true if it's a different bank.

Let's take the same example. Building was expected to be worth $20M, landlord pays $4M down and takes a $16M interest-only mortgage. The only thing the bank ever expected from this was to collect interest on the $16M until it's paid back, which could be never and that's fine as long as they get to keep collecting interest.

Then we find out the building is maybe really only worth $14M. But the landlord is still making the interest payments on the $16M, and over time it will likely become worth more than $16M again due to inflation if nothing else, so why does the bank need to foreclose? The risk that they could "lose $2M" is by that point a sunk cost. It's the thing that happens if they do foreclose (or fail to renew the loan). They'd be calling in the note against an LLC that owns nothing but a building which is now estimated to be worth less than the loan principal. So the obvious thing would be to keep renewing it as long as the landlord continues to make the interest payments.

This feels like some kind of regulatory inefficiency or accounting scam where the bank is listing the mortgage lien as an asset and would have to take a write off if they valued it accurately and therefore transfer their perverse incentive to the landlord to prevent that from happening.

Notice however that doing that also hurts the bank. The landlord is collecting $500k/year at half occupancy, then paying the bank $640k and losing $140k/year to try to avoid the total loss of their $4M initial investment. Maybe they can do that for a year or three but the longer it continues the higher the probability that they run out of money. Whereas if they were collecting the $700k/year from renting out the entire building at lower rents then they could keep paying the bank its $640k/year forever, regardless of whether they're technically underwater. And if the landlord runs out of money then the bank has to take the $2M write off because they get a $14M building instead of collecting interest on a $16M loan. So the bank is really shooting itself in the foot.

The whole goal is not to write off the value of the property which you have to do if you rent it for less money than initially planned. That's not that difficult to understand is it?

How do you asses the value? You use the x last transactions. No transactions, no data, the last value remains.

Think of it that way - until you haven't climbed on the scale, you haven't gained weight, even if your pants are bursting at the seams.

Agreed. From the article:

Actual commercial real estate professionals could give you many more reasons than I can

I am so tired of listening to people with little to no experience with commercial real estate try and explain the vacant storefront thing. Maybe this explanation in the article is correct, but it raises more questions than it answers, and it’s unclear why we should trust this person’s explanation.