When a BRRRR purchase is really underwritten like a flip

I'm starting from zero and I'd genuinely never seen the two-tests thing explained that way. So the appraisal can be great and you still get a small loan. That reframes the whole strategy for me.

That's the part worth absorbing early. Take a property where rent comps say $1,450 and a payment on a $150k loan puts debt coverage under 1.2. In that situation a lender writes whatever loan size gets coverage back to 1.2, and the borrower eats the difference in trapped cash. The only real levers at that point are renegotiating the purchase price or bringing more cash to close; there usually isn't a way around the lender's coverage math once the rent number is fixed. The upstream fix is screening for rent support before the rehab budget gets set, not after.

Points can sometimes buy down the rate, which lowers the payment and improves coverage. That costs money at close to get money at close, so it is rarely a clean win, but it is a lever worth pricing out before assuming the refinance does not work.

The rehab side of your problem: on a rental I'd spend the money on things that reduce future calls and let me hold a higher rent, and on a flip I'd spend it on what photographs well. Different scopes. If you scoped deal two like a flip you've probably got 8 or 10k in finishes doing nothing for your DSCR. Not recoverable now, but worth knowing for deal three.

Pulling rent comps before offering is such an obvious step and I've watched three people in my area skip it because the ARV spread looked so good.

There's a version of this where you're not wrong, quarry. If your market has strong appreciation and weak rents, chasing spread and accepting trapped capital is a defensible choice. You just have to admit that's what you're doing rather than believing you're running the refinance play.

Does the lender count all the rent or do they haircut it? I've heard something about them only using a portion.

Beacon, the coverage calculation typically uses net operating income, so rent minus taxes, insurance, management, and a vacancy allowance. Some lenders use a simplified version. It's not gross rent divided by payment, which is what people usually assume when they run it in their head.

Following up on my own comment, so the vacancy allowance means they assume the place is empty part of the year even if it isn't. That's another thing I'd have missed entirely.

Every lender I've talked to while studying uses a slightly different net calculation, which makes comparing quotes harder than comparing rates.

Going by this thread, the version that works seems to be buying in a place where rent is high relative to price, so the two tests give similar numbers and neither one surprises anyone. That's close to the core of it, though it's worth pressure testing against a market where the spread runs the other way before treating it as a rule.

Verity, that's most of it, and the trade is that high rent-to-price markets usually have weaker appreciation and rougher tenant economics. Something is always paying for the ratio. Quarry, one number worth checking before deciding a screen was wrong: what's the gap? If coverage lands at 1.14 against a 1.20 requirement, that's roughly 12k of loan away, and a slightly longer amortization or a small rate buydown might close it. If coverage is at 0.95, the property doesn't support the debt and no structuring fixes that.