Recent Deals methodology
How we choose deals and grade them. Rubric v1 was published before the first grade; v1.1 on Nov 1, 2026. Decision support, not a recommendation.
Public-data underwrite. The buyer's information set was larger.
Inclusion rule
A deal is added if either:
- (a) one of our 13 tracked hotel REITs is buyer or seller, at any size; or
- (b) it’s a US hotel sale with a disclosed price of $100M or more.
US means the 50 states and DC. Puerto Rico and other US territories are excluded.
Tracked REITs: AHT, APLE, BHR, CLDT, DRH, HST, INN, PEB, PK, RHP, RLJ, SHO, XHR.
- Every qualifying deal is logged, including ones we haven’t finished grading.
- Unpriced deals in non-disclosure states are listed as unpriced and never estimated. They’re graded only if they qualify under (a).
- The event date is the announcement date. A later close is a status update on the same entry.
Rubric v1.1
Each factor is scored 1 to 5. Public data only: no paid or subscription data. Spreads are stated as a plain number of basis points.
- Price per key vs replacement cost per key 30%
HVS U.S. Hotel Development Cost Survey (free): median total development cost per key, including land, blended by keys across the HVS classes of the hotels in the deal; edition and classes stated
As of: Latest edition published before the announcement; if only a price range was public, the first confirmed price, dated to its confirmation
- Implied cap rate vs 10-year Treasury spread 25%
Cap rate on net operating income (NOI) for a trailing 12 months ending at a month-end before the announcement (disclosed, disclosed NOI divided by price, or our calculation applying the company's own published NOI definition to its published hotel EBITDA, labelled as such); other bases are shown but not graded; 10-year yield from the U.S. Treasury daily par yield curve
As of: 10-year yield at the close on the announcement date
- Supply backdrop 15%
One free public source giving the total pipeline, under construction plus planned, net of closures, and the existing rooms it is measured against; a visitor bureau, city or state that republishes pipeline totals is cited as the source; for downtown hotels the visitor bureau's downtown area can be the market; never combined across sources or stages; partial-stage figures are context only; if a source lists closures without a room count, supply isn't graded for that deal
As of: Published before the announcement
- Demand trajectory 15%
Published lodging tax receipts, trailing 12 months vs the prior 12 months (fiscal year to date only if a 12-month series can't be built from public numbers); single months are context, never scored; Tuesday Trends forward curve as context for deals from October 2026
As of: Latest month published by the announcement date
- Debt and transaction context 15%
Filings, court and receiver records, public loan notices, press
As of: Public before the announcement
| Factor | Weight | Source (public only) | As-of rule |
|---|---|---|---|
| Price per key vs replacement cost per key | 30% | HVS U.S. Hotel Development Cost Survey (free): median total development cost per key, including land, blended by keys across the HVS classes of the hotels in the deal; edition and classes stated | Latest edition published before the announcement; if only a price range was public, the first confirmed price, dated to its confirmation |
| Implied cap rate vs 10-year Treasury spread | 25% | Cap rate on net operating income (NOI) for a trailing 12 months ending at a month-end before the announcement (disclosed, disclosed NOI divided by price, or our calculation applying the company's own published NOI definition to its published hotel EBITDA, labelled as such); other bases are shown but not graded; 10-year yield from the U.S. Treasury daily par yield curve | 10-year yield at the close on the announcement date |
| Supply backdrop | 15% | One free public source giving the total pipeline, under construction plus planned, net of closures, and the existing rooms it is measured against; a visitor bureau, city or state that republishes pipeline totals is cited as the source; for downtown hotels the visitor bureau's downtown area can be the market; never combined across sources or stages; partial-stage figures are context only; if a source lists closures without a room count, supply isn't graded for that deal | Published before the announcement |
| Demand trajectory | 15% | Published lodging tax receipts, trailing 12 months vs the prior 12 months (fiscal year to date only if a 12-month series can't be built from public numbers); single months are context, never scored; Tuesday Trends forward curve as context for deals from October 2026 | Latest month published by the announcement date |
| Debt and transaction context | 15% | Filings, court and receiver records, public loan notices, press | Public before the announcement |
Bands
Score 5 is the top band we set for each factor. Scores 4 to 1 step down evenly from it.
Price per key minus replacement cost per key, divided by replacement cost per key, times 100. Negative means below replacement cost. Replacement cost per key is the HVS U.S. Hotel Development Cost Survey median total development cost per key, including land, from the latest edition published before the announcement date, blended by keys across the HVS classes of the hotels in the deal (each hotel's class is stated on the entry). If only a price range was public at the announcement, price per key uses the first confirmed price, and that factor is dated to the day the price was confirmed (stated on the entry).
- 540% or more below replacement cost · top band
- 420% to 40% below · stepped down
- 3within 20% either way · stepped down
- 220% to 40% above · stepped down
- 1more than 40% above · stepped down
Cap rate minus the 10-year Treasury par yield, in basis points. Cap rate basis: net operating income (NOI) for a trailing 12 months ending at a month-end before the announcement (for example, the last completed quarter), divided by the price, or a disclosed cap rate on that same basis. NOI as the company defines and discloses it (for example, Adjusted EBITDAre less FF&E reserve) counts. Where the company publishes hotel-level EBITDA for the trailing 12 months and, separately, its own definition of NOI, we apply that definition to the hotel EBITDA and label the result "our calculation using <company>'s NOI definition", with both sources cited. EBITDA alone is not NOI. Cap rates on other bases (forward or pro forma NOI) are recorded but not graded. 10-year: the Treasury daily par yield curve close on the announcement date (or the prior trading day if the announcement date isn't a trading day). If no NOI or cap rate on a qualifying trailing-12-month basis was public, the factor is not graded. Stated as a plain number of basis points.
- 5400bp or more · top band
- 4300bp to under 400bp · stepped down
- 3200bp to under 300bp · stepped down
- 2100bp to under 200bp · stepped down
- 1under 100bp · stepped down
Total market hotel pipeline, under construction plus planned, net of announced closures, as a percent of the market's existing rooms, taken from a single public source that publishes that total share (or both its numerator and denominator) and was published before the announcement date. The market is the metro area; for a downtown hotel, a downtown area defined by the city's convention and visitors bureau counts as the market. Pipeline totals compiled by a data vendor count when a public body (a convention and visitors bureau, city or state) republishes them for free; they are cited to that public body. Figures are never stitched together across publishers or pipeline stages. If no single public source gives the total, the factor is not graded, with a reason. Partial-stage figures may appear only as cited, ungraded context. Supply: pipeline net of closures. If a source lists closures without a room count, supply isn't graded for that deal.
- 5under 2% of existing rooms · top band
- 42% to under 4% · stepped down
- 34% to under 6% · stepped down
- 26% to under 8% · stepped down
- 18% or more · stepped down
Year-over-year change in published lodging (tourist development) tax receipts for the market: the latest 12 months published by the announcement date vs the 12 months before that (sums of the monthly collections). Fallback, only if a trailing-12-month series can't be built from public monthly numbers: fiscal year to date vs the same period of the prior fiscal year, through the same latest published month. Single-month changes are shown as context only and never scored. The measure used is stated on the entry. The Tuesday Trends forward curve is context only, for deals from Oct 2026 on.
- 5up 5% or more · top band
- 4flat to up under 5% · stepped down
- 3down under 5% · stepped down
- 2down 5% to 10% · stepped down
- 1down more than 10% · stepped down
Categorical. Assigned from filings, court and receiver records, public loan notices and press public before the announcement date. When more than one category applies, the lowest applicable score is used.
- 5Clean arm's-length sale with no distress · top band
- 4Arm's-length, no distress, but the structure blurs the price signal: portfolio-allocated price, loan assumption or seller financing, earn-out or contingent consideration, partial-interest or JV recapitalization · stepped down
- 3Pressured seller without formal distress (publicly flagged loan maturity within 12 months, watchlist, fund wind-down) or a related-party / non-arm's-length transaction · stepped down
- 2Formal distress: loan in special servicing, notice of default, or a lender-negotiated sale (discounted payoff, deed in lieu) · stepped down
- 1Forced or receiver sale: receivership, foreclosure, bankruptcy (Section 363) or court-ordered sale · stepped down
Missing factors and coverage
A factor without qualifying public input is not graded. Its weight is spread pro rata across the graded factors, and the entry shows coverage ("k of 5 factors graded"). With fewer than 3 of 5 factors graded there is no overall grade ("Insufficient public data").
The overall score is the weighted mean of the graded factors, using their renormalized weights, rounded to one decimal and shown as “X.X of 5”. Each entry states its coverage, e.g. “4 of 5 factors graded”.
No look-ahead
Every input must have been public on or before the deal's announcement date. Replacement cost uses the latest HVS edition published before announcement. The 10-year yield is the Treasury daily par yield curve close on the announcement date. The cap rate uses trailing-12-month NOI ending at a month-end before the announcement. Demand uses tax receipts through the latest month published by the announcement date. Supply inputs must be published before announcement. One exception: when only a price range was public at announcement, price per key uses the first confirmed price and that factor is dated to the confirmation. Inputs dated after the relevant date are rejected and the factor is left ungraded. Later data (closes, resales, RevPAR, tax receipts) goes only to the anniversary tracker, never the grade.
The output line
Every grade ends in one mechanical sentence built from the inputs, for example:
“Scored using an estimated [X]% discount (or premium) to replacement cost with a [N]bp spread; [k] of 5 factors graded.”
Each factor that isn’t graded is named with its reason before the count, for example “…; supply not graded (closures listed without room counts); 3 of 5 factors graded.”
We grade deals, never buyers. We don’t total or rank grades by company.
Version history
- Rubric v1, Oct 25, 2026: first version, approved and published before any deal was graded. Same day, before any grade: named the cap-rate basis (trailing-12-month NOI), the 10-year timing (close on the announcement date), the demand measure (trailing 12 vs prior 12 months), the replacement-cost basis (HVS median including land, blended by keys), the supply source rule (one source, total pipeline) and the geography (50 states and DC).
- Rubric v1.1, Nov 1, 2026: Same 5 factors, weights, bands and 3-of-5 cutoff; supply is net of closures, a visitor bureau's downtown area can be the market for downtown hotels, and free public-body republications of pipeline totals qualify; a company's published NOI definition can be applied to its hotel EBITDA; a trailing 12 months ending at a month-end before the announcement qualifies; when only a price range was public at announcement, price per key uses the first confirmed price, dated to its confirmation. Every deal was re-graded; the v1 grade stays on each deal page as the original.
- Rubric v1.1 clarification, Nov 1, 2026: Supply: pipeline net of closures. If a source lists closures without a room count, supply isn't graded for that deal. Affected deals were re-graded as a new revision; earlier revisions are kept.
What changed from v1
Same five factors, weights and 3-of-5 cutoff. The rules for what counts as an input changed:
- Supply: pipeline net of closures. If a source lists closures without a room count, supply isn't graded for that deal.
- Supply market: for a downtown hotel, the downtown area defined by the city’s visitors bureau counts as the market.
- Spread: a company’s own published NOI definition may be applied to its reported hotel-level EBITDA, shown as our calculation using that company’s definition.
- Spread: a trailing-12-month period ending at a month-end before the announcement qualifies.
- Replacement cost: if only a price range was public, we grade on the first confirmed price, dated to the confirmation.
- Sources: industry-data totals that a public body republishes free can be graded, cited to that public body.
A rubric change re-grades every entry and shows both versions. The original grade is never overwritten.