There's a triplex in Farmington Hills that's been listed three times in fourteen months. The seller has owned it since 1999. Every time it comes back on the market, I run the numbers again, because in theory, it's the kind of deal that could meaningfully move my net worth.

Let me show you how I underwrote it, and why I decided to walk based on a single line item.

Three-unit building, $4,300/month gross rent ($51,600/year), asking $499,000. Insurance runs $2,000/year. I underwrite maintenance and capex reserves at 8% of gross rent each (this is an 85-year-old building, so I don't cut corners there). Financing is a standard DSCR loan: 25% down, 30-year amortization, 7.0% rate.

The one number I hadn't nailed down yet was property tax. The listing showed the seller's current bill: $2,000/year.

I ran the deal on that number first, because that's the number every casual buyer runs.

That's a real deal. A 7.9% cap rate on a cash-flowing triplex in a good school district, with debt coverage well above the 1.20–1.25x floor most lenders want. If I stopped here, I'd be writing you a very different piece.

The seller keeps listing, delisting, and relisting, each time with the price lower. On this most recent relisting, all their tenants are gone and they held an open house to let potential buyers properly size up the place.

The result three weeks after the open house was a whopping ZERO interest.

I figured it wouldn’t be insane to offer a noticeably lower price, so I ran another underwriting based on a price of $425K (using the same tax assumption as before).

At $425K, this isn't just a good deal — it's the kind of number that changes a portfolio. Double-digit cash-on-cash, DSCR with real cushion, a cap rate most operators would fight over.

This is the version of the story everyone wants to tell. It's also the version that's wrong.

Michigan runs on a system called Proposition A. It caps how much a property's taxable value can rise each year while the same owner holds it — which is exactly why a seller who's owned since 1999 can be sitting on a $2,000 tax bill on a property that would sell today for half a million dollars.

The catch: that cap doesn't transfer.

The moment the property sells, the taxable value uncaps and resets to the State Equalized Value (SEV) — roughly the assessor's estimate of half the market value. I didn't estimate this. I pulled the actual SEV from the Oakland County assessor: $220,000, taxed at Farmington Hills' non-homestead effective rate of 56.3 mills.

That's a confirmed post-sale tax bill of $12,393 a year — not $2,000. Every underwriting model I ran above was quietly built on a number I was never going to pay.

DSCR under 1.0 means the property doesn't even generate enough income to cover its own debt payment — before you take a dollar of cash flow home. No DSCR lender closes this loan at this price.

Better — but still below the 1.20–1.25x DSCR floor most lenders require, and a cash-on-cash return that barely beats a savings account for an illiquid, management-intensive, 85-year-old asset.

And to also note: the cap rate was now lower than the borrowing costs, which means using leverage on this deal would actually worsen the returns.

The "exciting" counteroffer from Underwriting #2 evaporates entirely once the real tax bill shows up. Same building. Same rent. Same $425K price. The only thing that changed is which year's tax bill you used.

Solving backward on the uncapped tax number:

That's the zone where this triplex is actually a good investment — not $499K, not even $425K.

Here's the part that makes this a pass rather than a negotiation: this property has been listed three separate times in fourteen months, and in that entire span the price has moved 5% — from $525K to $499K.

The seller has owned the property since 1999 on what's effectively a frozen tax basis, has no urgency, and has shown zero willingness to approach the $335K–$387K range where the deal clears for a buyer who has to pay the real, uncapped tax bill.

The range where this deal works for me and the range where this seller will sell simply don't overlap. They may never overlap, because the seller isn't pricing the building — they're pricing their own decades of accumulated tax savings, which aren't for sale.

If you're underwriting real estate in a state with an uncapping mechanic — Michigan's Prop A is the one I know best, but versions of this exist elsewhere — the seller's current tax bill is not data. It's a mirage that belongs to them and dies the moment they sign.

Always underwrite the sale on the post-sale reassessed value, confirmed against the actual SEV, not the number printed on last year's listing sheet.

The net-worth upside on this deal was real. The price required to capture it was one nobody currently at the table was ever going to agree to.