The agent who pulled comps for my refinance refi'd the same property twice in the same week

The comp selection problem you're describing is real, but the sharper issue is that appraisers and buy-side agents are solving different problems. An appraiser is defending a number to a lender. A buy-side agent who worked that corridor is reading absorption rate, days on market, and whether a sale represents a market or an anomaly, and those two skill sets rarely sit in the same room during a refi. The 22k condition adjustment is the number I'd push on hardest, because paired with two spike sales it compounds rather than offsets. A single outlier comp is adjustable; two outliers plus a generous condition bump and you've built a ceiling and called it a floor.

What changes the picture here is whether the original agent held a license in a state that permits broker price opinions for lending purposes, because in some states a formal BPO from the buy-side agent could have been submitted alongside the appraisal as a rebuttal document before the loan closed, not after the exit cap disappointed. That path is underused and the window for it is narrow, usually before the appraisal is certified. The owner in your example had the right person and the right data, just in an email from eighteen months prior instead of a dated, signed document in the lender's file.

What was the exit cap rate when the owner actually sold, versus what the model assumed at refi?

This is a good scenario to sit with if you are new and still figuring out what all these words mean, so let me define the key ones before the analysis.

A fourplex is a building with four separate rental units. A cash-out refinance means the owner borrows against the increased value of the property to pull out cash, usually to fund the next purchase. Comps (comparable sales) are recent nearby sales an appraiser uses to estimate what your property is worth today. A cap rate (capitalization rate) is a ratio that tells you how much income a property produces relative to its price.

What this story shows is that comp selection is a judgment call, and two of the three comps here reflected a six-week pricing spike caused by one unusually motivated buyer. That is not a reliable benchmark. The agent who had worked that street already knew it, and said so, in writing, before the refinance happened. The note just never made it back into the conversation.

The part a beginner might not think to ask: appraisals are not neutral. They are opinions built on choices, and the person with the most useful context for challenging those choices is often the agent who worked the deal originally, because they tracked that corridor when the numbers mattered to them directly. Bringing that agent back into a refinance conversation costs very little and can flag exactly the kind of outlier problem described here.

I am not able to tell you whether any specific appraisal or loan decision is sound; a lender, a licensed appraiser, and a CPA should all be part of that process.

One genuine question for you: when you picture yourself in real estate, are you drawn more to owning rental property like this fourplex, or to the agent side of the work? That changes which part of this story is most useful to study.

That email sitting unread is the actual failure here, not the appraiser doing what appraisers do. My out-of-state rental taught me the hard way that nobody volunteers information you did not specifically ask for, and the buy-side agent assumes the transaction is over once you close. I would want to know whether the owner had any real relationship with that agent after closing, or whether it was a one-and-done where nobody thought to pick the phone back up eighteen months later. The 22k condition adjustment is the number I keep coming back to, not egregious alone but stacked on top of two spike comps it compounds fast and nobody is doing that math out loud. What was the actual exit price two years later relative to the 390k?