Refi at stabilization came in $1.8M short, so instead of cash out I wrote another check

Ran your numbers because something in the framing is off.

NOI $1.18M, less a 4% FF&E reserve. If revenue is roughly $3.1M at 38% GOP, the reserve is about $124k, so the lender's NOI is call it $1.06M. At an 8.4% cap that's $12.6M of value, and 65% is $8.2M. You said $8.9M, so either they used a slightly looser LTV or they didn't deduct the reserve twice. Either way, that's the LTV test.

Now the coverage test. At 1.45x on a 25-year amortizing loan at say 7.25%, the constant is roughly 0.0867. Allowed annual debt service is $1.06M divided by 1.45, about $731k, which supports around $8.4M. So the two tests landed within a few hundred thousand of each other, and both are miles from your $12.8M.

The binding variable wasn't the sizing test. It was the cap rate. You underwrote 7.2% on a limited service extended stay and got 8.4%, and 120bps of cap on $1.06M is $1.8M of value all by itself. Your gap is almost entirely that. The margin miss cost you real money too, but if the appraiser had come in at your cap on the same $1.06M you'd have had $9.7M of value and been about $700k short instead of $1.8M.

So the lesson I'd take is narrower than yours. You need an exit cap you can defend from actual trades in that market for that segment, and you need it stressed. The FF&E deduction is a detail you should fix, and it isn't what did this.

$25k a key of PIP on extended stay is a soft-goods number with maybe a little case goods. Did the brand sign off on that scope as the full cycle, or was there a second phase due at year five that the appraiser knew about?

Appraisers on hospitality regularly deduct remaining brand-required capital from value, and if there was a deferred piece of the PIP sitting out there it comes straight off the number. Would also explain part of the cap rate. Buyers price a property with work still owed differently than one that's clean for eight years.

The mechanism worth naming here is that hotel refinance proceeds are set by two things you don't control and one you do.

The one you influence is trailing income. Hotels reprice every night and function as operating businesses, so a refinance test is almost always run on trailing twelve months of actual performance, adjusted by the lender. Adjustments commonly include a reserve for furniture, fixtures and equipment taken off net income before sizing, and a management fee normalized to market whether or not the current operator charges it. Both reduce the number the loan is measured against.

The two you don't control are the capitalization rate the appraiser supports and the lender's sizing constraints at the moment you ask. ledger's arithmetic shows why the cap rate usually dominates: a change of 100 to 150 basis points moves value by 12 to 18 percent on the same income, and the loan is a fraction of that value. Coverage sizing and loan-to-value sizing tend to converge for hospitality because lenders set both conservatively for the property type.

Which assumption the whole thing rests on: whether stabilized operating margin holds. Hotels are the most economically sensitive and operationally intensive of the major property types, and a margin assumption is a bet on wages, insurance renewals, and travel demand at the same time. Underwriting that treats revenue per available room as the sensitivity and margin as a constant hides the exposure rather than measuring it.

On the practical side, refinance terms, sizing tests, and required reserves vary by lender and change with the capital-markets environment for hospitality. Anyone modeling an exit refinance should get the current sizing parameters and reserve requirements in writing from the specific lender, and treat prior-year quotes as history rather than as inputs.

The haircuts are the part borrowers keep discovering late.

On a hotel T-12 the lender typically strips out anything that isn't repeatable, normalizes the management fee to something like 3% of revenue even if the sponsor's affiliate is running it for less, deducts the FF&E reserve, and then may hold back a portion of proceeds for a coming PIP. Some also underwrite occupancy to a trailing average rather than the most recent quarter, which hurts if you're ramping and helps if you're rolling over.

windlass, was the appraisal ordered before or after your insurance renewal hit the trailing twelve? Because if the doubled premium had only been in the numbers for two or three months, the appraiser may have annualized it on the expense side while still valuing off the trailing revenue. That's the worst version of the timing.

The 12% accruing member loan is the expensive part of this story and it's getting one line in your post.

Go read where it sits. On $1.15M compounding at 12%, if it's outstanding four years that's roughly $1.8M owed off the top of a sale before any equity sees a dollar, and typically before the preferred too. Your $31k of new cash went in behind it. If the docs let the sponsor keep advancing member loans without a consent threshold, the balance can grow again at the next shortfall and every dollar of it gets in front of you.

Things I'd want in writing: whether it compounds or accrues simple, whether there's a cap on total member loans, whether repayment is mandatory from refinance proceeds before distributions, and whether it can convert to equity at a set price. That last one shows up more than people expect and it's how a 4.8% position becomes 2%.

Every hotel thread I read ends with somebody funding a shortfall. That's my whole reason for staying away, and this one hasn't changed my mind.

7.2% exit on limited service extended stay in a secondary market is the error, and it's the most common one I see in decks I get asked to review. People pull a cap rate from whatever they paid, or worse from a national average across all segments, and then apply it two and a half years forward with no cushion.

My rule is exit cap equals entry cap plus 75 to 100bps for a hold under five years, and for hotels I add another 50 on top of that if the hold crosses a period where travel demand could soften. Then I size the refinance twice, once on loan-to-value against that stressed value and once on coverage against a trailing twelve net of a management fee and an FF&E reserve, and I take the lower. If cash out only exists in the version with the tight cap, there is no cash out in the deal.

windlass, the more useful question for you now: what does the sponsor's revised hold look like? A property that can't refinance at month 30 is usually holding to month 60 or later, and your money is committed for that whole stretch whether the model says so or not.

Sitting on capital and trying to work out what of this was visible from outside. windlass, was the refinance sizing assumption actually written into the offering material, or was it just a proceeds number in the returns table?

Because if the docs only showed "refinance in year three returns 70% of capital" with no cap rate and no coverage test behind it, I'm not sure a co-investor could have caught it at all short of asking for the model. And sextant, if I did ask for the model, how often does a sponsor actually send the working file rather than a PDF of the summary?

meridian, in my experience you get the PDF. Ask for the tab with the debt sizing on it and watch what happens.

My version of this was smaller and it still stung. Bridge on a 22-unit, planned exit refinance, appraisal came in 9% under and the lender's proceeds left me $140k short of retiring the bridge. Instead of a cash call I took the extension, which cost a 1% fee plus a rate that had moved against me, roughly $61k over fourteen months, and I got out at the second try. windlass, if your bridge has two extension years left, somebody should be comparing the all-in cost of extending against the 12% member loan. Those two numbers are rarely as far apart as they sound, and one of them doesn't sit ahead of your equity forever.