Three Hotel Deals We Passed On This Month

Three decline calls went out from our desk this month. The sponsors had answers for the obvious questions. The missing answer was the same in all three files: who pays when the business plan takes longer than the model says it will.

This is not a view that hotel credit has stopped working. Trepp reported that overall CMBS delinquency fell 20 basis points to 7.35% in June, helped by a 79-basis-point improvement in lodging (Trepp). A cleaner monthly reading does not make a thin reserve sufficient. It makes the distinction easier to miss.

We passed because the structure needed future good news to survive. Here are the three reasons, without the softened version.

The PIP arrived before the cash did

The first file was a branded conversion in a secondary Midwest market. The sponsor had a property improvement plan, a contractor, and a schedule. The proposed loan included an interest reserve sized to the end of the work program and no further. The brand required the rooms, public areas, and exterior work to be finished before the property could be reflagged, which is also when the rate lift in the pro forma began.

That sequence is the problem. A PIP is not just a capital-expenditure line. It is a calendar. During the work, a hotel can lose rooms, disrupt group business, and take its revenue hit before its repositioning benefit shows up. The reserve was built to completion. The revenue forecast began to improve at completion. The cash flow did neither.

We asked the sponsor to fund interest through stabilization rather than through the last construction draw. Our underwritten stabilization date sat more than a year after completion, so that request added materially to the reserve. It was not a negotiating point. It was the period in which the property would be carrying new debt while still proving the renovated product to the market.

The sponsor preferred to treat that period as upside. We cannot. In a conversion, the reserve must cover the overlap of renovation disruption, reopening, and the first months of an operating ramp. If the model has a 10-month PIP and a 12-month income ramp, a 10-month reserve is short by definition.

The ADR ramp had no reference point

The second decline was a select-service hotel in a secondary Southeast market converting from an independent flag to a national upper-midscale brand, with a forecast that got to market-leading ADR in month 12. The forecast assumed a double-digit rate increase and a multi-point occupancy gain in the first full year, and it put the new ADR above the competitive-set leader.

I did not believe the ramp.

A new flag can improve distribution. A completed PIP can support rate. Neither changes demand on day one. A lender should be able to draw a straight line from the brand, the room product, the competitive set, and the demand generators to each rate change. “New brand” is not that line.

The market data we had did support hotel revenue growth, but it was modest enough to matter. CoStar and Tourism Economics had raised the 2026 U.S. RevPAR forecast to 2.8%, with year-to-date RevPAR through April up 4.0% (CoStar). A rate jump of that size is therefore not a market assumption. It is an asset-specific claim, and it needs asset-specific evidence.

We requested monthly STR history, the current and post-conversion competitive set, a brand feasibility study, and evidence that the rate premium had worked at comparable conversions. The file had trailing market data, not conversion comparables. We would have sized the loan to a more measured ramp and revisited the upside after it appeared in the operating statements. The requested proceeds only worked at the faster case.

That is a decline, not a haircut. When the debt service survives only if the property is the rate leader in year one, the lender is financing the conclusion.

Liquidity cannot be the smallest reserve

The third file had a reserve. The sponsor did not have much left after funding it. It was a limited-service acquisition in a tertiary Southwest market. After equity, closing costs, and the required reserve, the sponsor’s post-closing liquidity came in below the interest reserve itself. The asset also carried deferred maintenance that sat outside the seller credit.

We passed for a simple reason: the interest reserve belongs to the asset. It is not the sponsor’s emergency fund. It pays contractual debt service while the business plan runs. It cannot also absorb a franchise-required repair, a tax reassessment, a delayed insurance settlement, or three weak months in a seasonal shoulder period.

Lodging credit has improved from the 23.2% June 2020 delinquency peak; KBRA put 30-plus-day lodging CMBS delinquency at 6.4% as of March 2026 (KBRA). That is useful context, not permission to treat liquidity as cosmetic. A sponsor with less unrestricted cash than the interest reserve itself has one problem, not a cushion. On this file, the model had already identified the ordinary carry before it identified an ordinary surprise.

We asked for a post-closing liquidity covenant above the interest reserve and for the deferred work to be funded in a repair escrow at closing. The sponsor did not have it. Another lender may find a different answer. Our answer was no.

The three files differed in flag, geography, and proceeds. The credit issue was identical: the proposed capital stack had no margin between the known cash need and the first bad month. We will finance a transition. We will not finance a plan that uses the last dollar of liquidity before the transition starts.

The box this week

• Loan size: $5M floor, with a $20M–$100M sweet spot
• Asset focus: hotels, resorts, and mixed-use hospitality; bridge and construction
• Leverage: to 65% of cost on transitional hospitality, sized to sustainable cash flow
• Reserve: interest funded through stabilization, not merely completion
• Liquidity: sponsor cash after closing must exceed the reserve and identified repair needs

Send the PIP, monthly operating history, reserve schedule, and post-closing liquidity schedule with the file. We can identify this issue before a term sheet turns it into a closing problem.