The ADU appraisal came in $73k below the build cost, and the owner still made the right call

The GRM assumption is doing the work, but the rent figure matters just as much.

The lender closing before the completion appraisal landed is the part most people will skip over, and that is the whole game right there.

I have been running numbers on Garfield Heights duplexes for months and the comp problem is exactly what kills the income approach confidence. There are maybe four or five ADU sales in a 10-mile radius with reliable data, so an appraiser is essentially building the value from rent backwards using a GRM they pulled from whatever multifamily sales were available, which may have nothing to do with your specific submarket or unit type.

What was the lender's projected as-improved value based on, and did they order their own appraisal before closing or just underwrite off a pro forma the owner handed them?

The timing structure in this post deserves more attention than the yield math. The lender used a projected as-improved value to close the cash-out refinance before a completion appraisal existed, which means the underwriting relied on an income approach the lender ran internally, not one a fee appraiser produced at stabilization. That internal projection apparently used a different rent multiple than the market appraiser applied later, and the $73,000 gap is almost entirely that difference in multiples rather than anything the owner controlled. The assumption doing the most work in the original decision was that the lender's projected value and the eventual appraised value would land close enough that the equity position would hold. They did not, and the owner got lucky that the financing had already closed rather than being contingent on it.

What this structure actually protected against was a completion-contingent draw, where the lender holds back proceeds until a finished appraisal confirms value. If the loan had that feature, the owner would have faced an equity shortfall mid-construction with a partially built cottage on a lot that is hard to sell in that state. The question worth sitting with before any ADU financing is whether your loan documents include a completion appraisal condition on disbursement, because that one clause changes the risk profile entirely. Have you seen the actual loan documents on this deal, or are you working from what the owner described?

The $137k appraisal isn't the problem. Most people building ADUs never ask their appraiser what GRM they'd apply before they pull permits. Run that number first and you know exactly what rent you need to hit a target appraised value, then you decide if the build pencils, not after you've already spent $210k.

What market is this in? GRMs in the low 70s tell me rents are relatively high against prices, which actually suggests the income approach is working as intended, not failing.

What nobody pulls out of their pocket before the permit is the rent ceiling. In a lot of rural-adjacent or secondary markets, $1,850 is already near the top of what the market will absorb for a detached cottage, so even if you build cheaper, the GRM math does not move much and you're still landing in the same appraised value neighborhood. I had a parcel in Iron County where the income approach would have capped a hypothetical ADU around $110k no matter what I spent, because rents in that zip simply don't comp out higher.

What's the market rent ceiling in your owner's area, and is there any room above $1,850?

The piece of this that changes the picture is what happens when that owner eventually wants to pull equity back out, because the income approach that a forward-looking lender used to close the cash-out refi is not the same methodology a new lender will use on a seasoned loan. If rent has grown and the market has added a handful of ADU sales that an appraiser can now work from, the sales comparison approach starts to compete with the income approach, and in some suburban markets those two converge closer to cost over five to seven years as comps accumulate. The owner who looks trapped at $137,000 today may find a reappraisal at year six lands closer to $175,000 simply because the comparable pool thickened. The $73,000 gap is not fixed; it is a function of market data scarcity, and that scarcity resolves with time in most growing ADU markets.

The related strategy that shifts the calculus is the decision about how to title and finance the ADU from the start. An owner who structures the main house and the ADU under separate financing instruments, or who works with a lender willing to underwrite on a blended DSCR for the full parcel, has more refinance optionality later than someone whose only path back to equity is a whole-property cash-out. The blended DSCR approach treats the $1,850 as part of the total income stream against the full property debt, which means the compression in ADU appraised value matters less in isolation. What the owner in this post probably needs to know before the next refinance window opens is whether their current lender's note covers the full parcel or just the primary structure, because that document is the one that either gives them flexibility or forces them into a corner.