Minibar and Loan Market Update from Bridge, September 2026

Welcome back to the Minibar + Loan Market Update, your monthly shot of what’s moving hotel financing. This is our August and September edition.

My name is Rohit Mathur (still typing these myself). I’m the co-founder and CEO of Bridge, a direct lender and marketplace for hotel loans. Before founding Bridge, I spent 10+ years at Citi in capital markets.

Quick programming note: we’ve moved our website to bridge.co (from bridgemarketplace.com, check it out and give us your feedback!)

An overarching thought the Bridge team has been debating for the past six months: how is AI actually helping the hotel owner? I’m sure you’ve seen the flood of AI generated images on LinkedIn and the explosion in word count (Claude and ChatGPT can be very wordy). But let’s be honest, we’re in this industry to make profits. Have your profits gone up because of AI? Or has it just added cost?

Here is where we’re coming out. AI is only worth anything to a hotel owner if it does one of three things: reduces the cost to build, buy or refinance a hotel, increases RevPAR, or reduce operating costs. Sounds basic. So where is that happening today?

Next week Bridge is launching a very practical AI benefit for all our clients. You’ll see a press release, and I’ll email you about it. We’re taking something that hasn’t changed in 100 years, and we’re going to change it.

Rates, Rates, Rates

I told you in June that the dot plot had flipped from a cut to a hike. On September 16 the FED meeting announced the move up.

The Fed raised 25 basis points to a target range of 3.75% to 4.00%. Unanimous, 12 to 0. First increase since July 2023. Fed Chairman Warsh’s line at the press conference was that inflation has been too high for too long, and he added that he’d be hard-pressed to call financial conditions restrictive. Translate that: he is expecting another hike will be needed.

The dot plot backs him up. Sixteen of the eighteen participants who submitted a dot expect at least one more hike before year end. Four of those see two more. That lands the year-end median at 4.00% to 4.25%. Eight members see another hike in 2027 and only four see a cut. Nobody is showing a path back to cheaper capital.

The speed of the change in direction matters more than the 25 bps. In March the Fed’s own median said one cut in 2026. In June it said one hike. In September they delivered one hike and are now saying another is coming. That’s roughly a 100 basis point swing in the projected path, in six months, with no change in the underlying economy other than energy prices.

Here’s where the rates that matter to your deals sit today, against a month ago and a year ago:

Sources: Federal Reserve H.15, NY Fed SOFR, FOMC statement September 16, 2026.

What it actually does to your deals

If you have a floating-rate construction loan, you got a rate increase on September 17. Prime moved to 7.00% and SOFR steps up with it.

If you’re pricing fixed-rate takeout debt, the damage is smaller than you’d think. The 5- and 10-year already moved, because the bond market priced this hike months ago. The Fed hiking the short end doesn’t automatically push the long end higher. Sometimes the opposite, because a Fed that’s credibly fighting inflation gives the long end permission to settle down.

My two cents: if you’re sitting on floating-rate construction debt, this is the quarter to stress your interest reserve and start the takeout conversation, not next spring. And stop waiting. The people who told me in March they’d refinance when rates came down have now watched the Fed move 25 bps the wrong way and signal another. Every month of waiting since then has cost them, read the next section to see how much prices have gone up in the same time. The deals getting done are the ones that pencil at today’s number.

Introducing the Bridge Hotel Construction Index

We built our own cost index because nobody publishes one for hotels

Every ground-up conversation I have starts the same way. Someone asks what construction costs have done since they last ran the numbers. And until now, the honest answer was a shrug and a figure lifted from an index built around an office building or a warehouse.

So we built our own.

The Bridge Hotel Construction Index tracks what it costs to build a hotel in the US, month to month, holding scope constant. Same hotel, same spec, what’s the number today versus last month. January 2019 = 100. We publish three series: hard cost for GC contract scope, FF&E for the furniture and operating supply package, and a blended total weighted toward hard cost.

How we built it. The weights come from real GC schedules of value on Bridge deals. Not a model building, not a public template. We take line-item budgets from ground-up hotels we have actually underwritten, map every line to a cost category, then split each line into its material, labor and equipment components. Those categories get priced month to month against BLS producer price and construction wage data going back to 1994. As new project budgets come through the platform we re-cut the basket, at most once a year so the history holds.

Public indices price a generic building and bury contractor margin swings inside the number. Ours is weighted to how a 120 to 175 key hotel actually gets built, which is why it moves differently from a broad nonresidential index when wages or fuel run.

What it doesn’t do. No land, no financing costs, no developer fee, no franchise fee, no soft costs, no contractor overhead and fee. This is an escalation index, not a unit cost reference. It won’t tell you what a hotel costs per key. It tells you how much that cost has moved.

Does it hold up? Backtested, the hard cost series tracks the BLS nonresidential construction benchmark to within a point of a percent cumulatively, and runs materially less volatile year to year because we strip out contractor margin. It also reproduces the 2020 to 2022 run-up at the magnitude developers were actually seeing at the time.

Where it sits today. As of August 2026, blended hotel costs are up 6.86% year over year and 47.9% since January 2019. Hard cost is up 7.25% and sits at 150.4. Construction wages and equipment fuel are doing most of the work.

What we’re measuring against. Turner Construction is the largest contractor in the country by revenue and has been publishing a cost forecast for more than 80 years. The Turner Building Cost Index is a quarterly national read on nonresidential building costs, built from labor rates and productivity, material prices, and how competitive the bidding market is. It’s the number most GCs and developers reach for when they want to know which way things are heading. That makes it the right yardstick.

Turner sits at 136.7, up 5.15% year over year. BHCI hard cost sits at 150.4. That’s a 13.7 point gap since 2019, and it has been widening, not closing.

My two cents: if you’ve been escalating your budget off a generic construction index, you’ve been under-escalating. Hotels have run about 14 points hotter than the broad market since 2019 and the spread is still opening. Build that into your contingency, or your GMP will find it for you. The flip side is that the same math is keeping your competition out of the ground.

Hotel Performance: All Rate, and the Rate Is Slowing

With help from my friends at Kalibri Labs

RevPAR is up

Occupancy peaked around 70% this summer against 73.5% in 2019. ADR is doing all the work. And ADR growth peaked in mid-July.

270 basis points of ADR growth, gone in six weeks. Room nights at +0.4% is the weakest print since December.

One caveat before you extrapolate. That July 10 peak contains the World Cup data. Some of the drop is a one-time event rolling off the comparison, not demand softening, so the underlying trend is probably flatter than the chart implies. But that cuts both ways, and here’s the part that matters for the section right after this one: every hotel in a host market now carries a trailing twelve with a tournament inside it. With SBA moving to historical earnings on October 1, a distorted T12 is a real underwriting problem. Buying in a host market, ask what the numbers look like without it. Selling, know that a good lender is going to strip it out.

The higher the chain scale, the better the demand.

Economy is the only negative segment, and it isn’t leisure. It’s government at -7% and corporate at -5%. That’s structural.

Extended stay is the upside outlier, and the quality of the demand matters more than the number. Upper tier is up 6% with rack, BAR and loyalty at +12%. That’s booked direct and it flows through to the bottom line. Lower tier is up 1%, but the growth is OTA at +6% while its own direct channel is negative. Same headline, very different margin.

Secondary markets are quietly working. Small city and town demand is up 3%, corporate up 4%, which is the segment that fills midweek in select-service.

How we’re underwriting it

• Forward RevPAR at or below 3%.
• Extended stay stays the preferred format, earned through direct channels rather than OTA.
• Re-test economy and government-dependent assets. A -7% government print is not noise.

Source: Kalibri Labs, State of the Industry, September 2026. Data through August 21, 2026.

My two cents: we spent two years celebrating RevPAR growth without asking where it came from. It came from rate, and rate is fading. If your pro forma has occupancy climbing back to 2019 and ADR compounding on top of it, you’re counting the same dollar twice. Pick one.

The October 1 SBA Change Nobody Is Talking About

Your projections stop counting for 7a

Everyone in my inbox thinks the SBA story this year is the down payment. It isn’t. That fight already happened and owners lost it in 2025.

The October 1 change is bigger, and almost nobody I talk to has priced it in.

On most 7(a) hotel acquisitions, SBA is taking your projections out of the qualifying DSCR calculation.

A huge share of hotel acquisitions are improvement stories. You buy an underperforming asset, run the PIP, reflag it, replace management, push ADR, stabilize at a higher NOI. That’s not an exotic thesis. That’s the business.

Starting October 1, that story gets much harder to finance with SBA.

SOP 50 10 8.1 was issued August 14 and takes effect October 1, 2026. The dividing line is when your loan gets its SBA loan number, not when you applied. A file submitted September 25 that gets numbered October 2 is underwritten under the new rules.

The example that matters

You’re buying a hotel producing $650,000 of adjusted EBITDA. Post-closing debt service is $600,000.

Historical DSCR: $650,000 ÷ $600,000 = 1.08x

You’re putting $1 million into the property, replacing management, running the PIP. Your feasibility study says stabilized EBITDA hits $900,000.

Projected DSCR: $900,000 ÷ $600,000 = 1.50x

Under the outgoing rules, a well-supported projection gave the lender real room to establish repayment ability. Under 8.1, that 1.50x does not fix the 1.08x. The deal has to clear 1.25x on history before your upside case is even relevant.

And remember the amortization change, because it compounds. If part of that $600,000 of debt service was built on a 25-year schedule for the whole loan, the business portion now amortizes over 10. Your debt service goes up and your historical coverage goes down, on the same purchase price.

“It’s underperforming today, but after we renovate it and operate it correctly it works” is still a perfectly good investment thesis. It may no longer be an SBA 7(a) acquisition underwriting thesis.

504 is a different conversation

Hotels are special-purpose under 504, so borrower contribution is generally at least 15% rather than 10%. If you’re developing new and the operating business is also new, it can go to 20%. Those rules carry forward into 8.1 unchanged. So for ground-up, the feasibility study, budget and equity requirement are still your issue, not the October acquisition rules. And don’t casually apply the acquisition DSCR change to a refinance just because it’s an SBA loan. Analyze the refinance rules separately.

The LIONs Act is still worth watching

It would raise the maximum gross 7(a) loan from $5M to $10M and extend the 75% guarantee to that level. H.R. 1893 is pending in the House Small Business Committee, and there’s a Senate companion, S. 901, from Sen. Thom Tillis, sitting in Senate Small Business and Entrepreneurship. Neither has moved. Do not size a hotel acquisition around a future $10M 7(a).

My two cents: the October change isn’t “SBA wants more money down.” It’s “SBA wants the hotel to support the debt on what it has already earned, not what you believe it will earn after you fix it.” That is a much bigger deal, and it lands squarely on the most common hotel acquisition in the country. If you have an SBA deal live right now, run it under both rulebooks this week before you assume it still works.

Deal of the Month: $29.0 Million, Non-Recourse, 98 Keys

Bridge closed a $29.0 million non-recourse, interest-only loan for the acquisition and conversion of a 98-key hotel to a Hilton Tapestry Collection.

From our client, and I’m printing this one in full because of the last line:

“Many people told us this deal couldn’t be done. The Bridge team showed their deep connections and worked with us through the entire process including coordinating with the lender, appraiser, legal and other 3rd parties. We’ve been in hospitality for 20+ years with over 30 assets, we keep hearing about technology and AI, this is the first time we’ve seen AI used in getting a new loan for us.”

That’s the test I laid out at the top of this issue. Does the AI actually reduce what it costs to buy, build or refinance a hotel? Here’s a sponsor with 30 assets and two decades in the business saying yes, for the first time.

Conversions are where the market is right now. Tapestry crossed 200 open hotels globally this summer and the growth is almost entirely independents joining the Hilton system. This is that trade, financed non-recourse.

Why we could do it. Many on our team spent 10+ years at Citi and know how banks size and underwrite CMBS. We built underwriting models specifically for CMBS, so we know how each desk sizes a hotel and we pre-underwrite before the deal ever reaches the bank. We also hold direct relationships with all 20 CMBS lenders in the United States. When a deal comes in, we don’t guess which desk will like it.

And the market is open. Spreads are competitive, lenders want hospitality paper, and we have several more deals moving through these same shops right now.

My two cents: if you have a maturity in the next 12 months, you’re thinking about cash-out, or you just want to know what CMBS proceeds look like on your property, book a call with us. Worst case, you find out your asset supports more than you thought.

About Bridge

We are the first fintech focused on Hospitality loans and partner with Hilton, Hyatt, AAHOA, Choice, Red Roof, Wyndham, Kalibri Labs, M3 and others.

We closed and funded $500.0 million+ in loans in 2025 (our first full year of Hotel Lending) and we only have one goal:

Simplifying access to competitive financing for your hotel loan.

Check out our new Pro Forma Tool.

Bridge Minibar Update

We’re ready for our end of year sprint here with roughly 100 days to close deals before year end. Our office and minibar are fully stocked up, come visit us in NYC!