What I actually look at before I touch a BRRRR deal from the lending side

The misconception I want to correct is that 75 percent LTV on the refi is a lender preference. Most DSCR lenders treating this as a long-term rental will underwrite to 75 percent LTV as a hard ceiling, and the appraisal they order will use rental income methodology, not just comp sales, so an ARV built on sales comps alone can come apart the moment the rent schedule hits their desk.

What I spend time on before the refi number even matters is the spread between the DSCR lender's floor rate and the actual market rent, because that spread is where deals quietly die in a way the sponsor never modeled. Take a property with a stabilized rent of $2,200 and an ARV of $340,000. At 75 percent LTV the refi loan is $255,000. At a 7.5 percent DSCR note rate on a 30-year amortization, principal and interest runs roughly $1,785. Add taxes, insurance and any HOA and you are at, say, $2,100 all-in before vacancy and maintenance. That pencils at a 1.05 DSCR, which most DSCR lenders reject at that LTV. The sponsor needed rent closer to $2,500 or an ARV of $380,000 to make the refi actually retire the bridge note and leave the lender comfortable. The question I always ask the sponsor is: what is the rent comp you are using, and is it a current executed lease or a Zillow estimate from six months ago, because those two numbers are not the same thing in this market.

The assumption doing the most work in your framework is the ARV. At 75% LTV on a DSCR refi, the appraised value controls almost everything: how much capital comes back, whether the debt service is serviceable at 1.25x, and whether the sponsor traps equity or recycles it. Sponsors who model from purchase price forward tend to anchor their ARV to the comp that justifies the deal rather than the comp a conservative appraiser would use. You're running the math against a number that hasn't been stress-tested yet.

The risk I'd add to your framework is appraisal gap timing. In Snohomish specifically, comp velocity can shift between the day the sponsor runs their model and the day the DSCR lender orders the appraisal, sometimes 90 to 120 days later after rehab. A refinance underwritten at 75% of a stale ARV can still blow up if the appraiser pulls comps from a softer quarter. The sponsor's conservative ARV estimate needs to account for that lag, not just market conditions on the day they underwrote.

The other thing worth noting: DSCR thresholds vary by lender and loan structure, and the 1.25 floor you're using may not match what the refinance lender applies. Confirm with the DSCR lender in writing what their actual coverage requirement is for the property type and loan size in question, because a mismatch there creates a second point of failure after the ARV risk.

Your backward-from-refi approach is structurally sound. Most of the BRRRR failures I see documented follow exactly the pattern you're describing, forward-modeled deals where the refinance was never stress-tested as a standalone underwrite.

When you're reviewing these, are sponsors typically bringing their own DSCR lender term sheet, or are they modeling against a rate and coverage assumption they haven't confirmed?

@lending_side what's the ARV source when Snohomish comps are thin?

The part that caught me was "no amount of optimistic comps fixes it" because I see the inverse problem from the borrower side, people who have already fallen in love with the ARV before they have ever talked to a DSCR lender about what that lender's appraiser is actually going to accept. The ARV and the appraised value are not the same number, and in a county like Snohomish where you have genuine price variance between, say, Marysville and Edmonds, a sponsor pulling comps from a half mile too far south can be off by forty or fifty thousand dollars on a refinance number, which at 75 percent LTV is thirty to thirty-seven thousand dollars of equity that never existed.

What I keep watching from my seat, and I am on the passive end so I am reading through other people's deals rather than running my own, is that sponsors underestimate the gap between their broker's opinion of ARV and what a lender-ordered appraisal produces under FNMA guidelines or whatever the DSCR lender uses. Those are different methodologies and the adjustment grids cut differently. Do you ever push back when a sponsor brings you their ARV source, or do you mostly let the appraisal settle it and just underwrite to a haircut on whatever number the sponsor gave you? Because that discount factor seems like where the real discipline lives.