Start free trial
No credit card required
Commercial Real Estate Credit —
Seattle-Tacoma-Bellevue, WA
The state of disclosed CRE credit in this market · WA
The read
$5.1B of CMBS across 154 loans. Office is both the largest book ($2.6B) and the most distressed (15.4% vs 11.1% national). Distress is rising in the filed record — 8.7% as of 2026-07. The heaviest maturity load lands in 2029 ($1.1B). 145 on-the-ground distress events in the past year (9,295 jobs).
Overview
CMBS
Submarkets
Local Banks
Jobs & Demand
On The Ground
Loading map…
The CMBS properties our stress model flags here, plus recent store closures and layoffs. Stressed is a wider net than distressed — the distressed figure above counts only loans already in special servicing or 60+ days delinquent, while these dots also include performing loans facing maturity or coverage pressure. Property dots sit at the center of their ZIP code, not the building — they show where stress concentrates, not which asset. Closures and layoffs are pinned to address where we have it. Open any dot to the building and the loan behind it →
Office carries the weight in Seattle, and the filed record is turning higher
Seattle-Tacoma-Bellevue carries $5.1 billion of CMBS across 154 loans, and the strain is concentrated where the book is heaviest. Office is both the metro's largest exposure at $2.6 billion and its most distressed, running a 15.4% distress rate against an 11.1% national mark — a gap that puts Seattle's towers on the wrong side of the comparison. Overall, the filed record has been rising, reaching 8.7% as of July 2026, so the distress here is a status servicers are declaring, not a forecast.
The maturity calendar does not offer much of a reprieve. The heaviest load lands in 2029 at $1.1 billion, the single biggest refinancing year on the page, with another $0.8 billion stacked in 2028. Against that, the median DSCR of 2.04 shows there is still coverage in the broader pool, and sectors outside office read quieter — hospitality's $0.7 billion book shows no distress on the tape.
The on-the-ground signal is the one to watch. The metro logged 145 distress events over the past year — store closures and layoff notices combined — tied to 9,295 jobs. With office-using employment down 2.0% year over year and unemployment at 4.8% as of August 2026, the local demand picture sits underneath an office book already carrying the metro's distress.
Written 2026-10-27
CMBS Distressed UPB
$455M / 8.9%
in special servicing or 60+ days delinquent
CMBS Maturing ≤ 24mo
$1.6B
Banks HQ’d Here (stressed)
0 / 10
Bank Early-Warning
1 flagged
Store Closures (1y)
69
Layoff Notices (1y)
76 / 9,295 jobs
0.52% of metro employment
CMBS Loans / UPB
154 / $5.1B
Unemployment · Aug 2026
4.8%
0.0pp yr
Office-Using Jobs · 2024
568,450
-2.0% yr
How much CRE distress is there in Seattle-Tacoma-Bellevue, WA right now?
Distress in Seattle-Tacoma-Bellevue, WA is uniformly elevated across all four independent feeds, with all 6 possible pairwise comparisons agreeing—none quiet, none disagreeing. Store closures rank 6th of 392 by count at 69 closures (3.83 per 100,000 jobs); WARN notices rank 10th of 386 at 80 notices (0.41% of employment); bank CRE over the noncurrent line sits at $1,699.0 million, 5.83% of the $29,126.2 million allocated to the metro (31st of 393 by count); and securitized loans in special servicing total 10 rows at 7.88% of the $5,065.0 million trust slice (20th of 335). No leg is blind, so no reading is unavailable. The unanimity is notable given the bank leg is allocated by branch deposits rather than confirmed by call reports, yet the signature here is broad-based—retail, employers, lenders, and CMBS all read hot simultaneously.
The figures behind this answer
| Signal | Level | Rate | Rank by count | Rank by rate | Reading |
|---|---|---|---|---|---|
| store closures store closures per 100,000 jobs |
69 closures | 3.83 per 100k jobs | 6 of 392 | 58 of 392 | elevated trailing 365 days to today
|
| layoff notices (WARN) workers on layoff notices as a share of the metro's employment |
80 notices | 0.41% | 10 of 386 | 46 of 386 | elevated trailing 365 days to today
|
| bank CRE at lenders over the noncurrent line share of the CRE lent into this metro that sits at a lender over the blended-noncurrent line |
$1.70bn | 5.83% | 31 of 393 | 131 of 393 | elevated
allocated 2026-03-31
|
| securitized loans in special servicing share of the securitized balance read here that is in special servicing |
10 loan records | 7.88% | 20 of 335 | 57 of 335 | elevated 2026-07-29
|
6 pairs compared
6 both elevated
0 both quiet
0 disagreeing
0 unreadable — a side is blind
4 of four legs elevated
one metro, four independent feeds, each on its own denominator
114 of 441 metros have at least one blind leg and 48 are blind on all four. Those 48 are the same metros the board's map cannot place and the metro resolver cannot name: one ZIP-to-CBSA crosswalk boundary, showing up three ways.
Elevation is computed on the COUNT, over a flat threshold, which is a size filter. New York is 2nd of 392 by store closures and 264th of the same 392 by closures per job — both true, and which one you lead with decides what the reader believes.
Where two legs disagree, that is usually a fact about WHICH mechanism is running, not an error in either feed and not a worse metro.
A leg with no denominator, or whose numerator cannot be measured, reads LOW and means nothing. Each leg publishes a measurement state (allocated, floor, measured, no denominator, no rate, not measurable) for that reason.
Store closures: divided by 1,801,056 jobs
— all metro jobs, QCEW annual average — a SIZE normaliser, not a matched population: the feed spans retail, food service and pharmacy and no single QCEW sector covers it. The MATCHED read sits beside it as this leg's second reading (trailing 365 days to today).
Layoff notices (WARN): divided by 2,172,819 employed
— employed persons, BLS LAUS, not seasonally adjusted (trailing 365 days to today).
Bank CRE at lenders over the noncurrent line: divided by $29,126.2 mm
— bank CRE allocated to this metro by branch deposits (2026-03-31).
Securitized loans in special servicing: divided by $5,065.0 mm
— current balance on the non-pari-passu rows of this tape — the trust's slice, not the whole loan, on both sides of the ratio (2026-07-29).
Written from the figures above · CBSA 42660 · stress_metro · last changed 28 Aug 2026
· Ask your own question →
Which distress signals are elevated in Seattle-Tacoma-Bellevue, WA, and which cannot be read?
In Seattle-Tacoma-Bellevue, WA, credit distress is confirmed and material, with "hundreds of millions of distressed CRE exposure" behind it. The key elevated signals include bank distressed CRE at $1.70bn (with a 8.9% bank allocation share and 34.6% of bank CRE at risk at the 90th percentile), CMBS special servicing (at $399.0mm, representing 7.9% of metro UPB), and real-economy distress with 69 store closures and 80 WARN notices over the past year. The phase is "peak" and "confirmed," as leading and realized signals fire together. However, bank-level distress cannot be read: the number of distressed banks is 0 per the count, and distressed bank assets are 0 — so that signal is not elevated.
The figures behind this answer
- CMBS in special servicing
- $399.0mm
- … as a share of this metro's CMBS balance
- 7.9%
- Bank CRE lent into this metro
- $29.13bn
- … at risk at the 90th percentile
- 34.6%
- CRE at lenders over the noncurrent line
- $1.70bn
- Assets at those lenders
- $0.00bn
- … share needing no branch-deposit allocation
- 8.9%
- Signals reading elevated
- bank CRE over the noncurrent line, CMBS in special servicing, store closures, WARN layoff notices
- Legs agreeing
- 4
- Phase
- peak
- CMBS loans in special servicing
- 10
- Distressed banks
- 0
- Store closures (past year)
- 69
- WARN notices (past year)
- 80
leading (real-economy) and realized (credit) signals are firing together
hundreds of millions of distressed CRE exposure
credit distress AND ground-level distress, with real dollars behind the credit side
Written from the figures above · CBSA 42660 · geo_metro_signals · last changed 27 Aug 2026
· Ask your own question →
Jobs & Demand — the labor market under the collateralWatch it change →
Every other zone on this page reads the credit — what the loan is doing.
This one reads the demand underneath it. Office jobs pay office rent;
office rent services the office loan. When the jobs go, the DSCR follows — but not for another year or two.
That lag is the whole point of looking here.
Unemployment · Aug 2026
4.8%
0.0pp yr
Last 24 months
3.8%5.9%
Monthly metro unemployment (BLS LAUS) — the freshest labor signal there is,
about six weeks behind. Read the direction, not the level: a 4% metro that is rising and a 6% metro that is
falling are not telling you the same thing.
Jobs by CRE-relevant sector · 2024
Annual employment by sector (BLS QCEW, 2024; 1,801,056 jobs in the metro).
QCEW lags — it is the structural read, not the current one, and we pair it with the monthly unemployment above on purpose.
Office-using sums five supersectors; where BLS withholds any one of them for
disclosure, we show nothing rather than a number we know is too low — so a missing sector here means
“suppressed at the source,” not “zero.”
Sector Heat — CMBS distress rate by property type (metro vs national)See the loans behind the bar →
This metro's CMBS runs 8.9% distressed where its own property mix predicts 7.8% — $59M more than this book would carry at the rate each property type runs elsewhere in the country, dollar-weighted and excluding this metro. The gap is measured; the cause is not. 1 of 3 property types here run a higher distress rate than the national average for that type (the tick on each bar).
Elimination — does the gap survive a control?
Each row re-prices this metro's book on a finer cell — its own property types, then those types split by loan size, then by vintage — always at the rate that cell runs elsewhere in the country, dollar-weighted. A gap that shrinks toward zero is explained by the control; one that stays has ruled it out. What is measured is what survives, not a cause.
property mix alone
+1.1pp
… and loan size held fixed
+0.8pp
the gap is still there with loan size held fixed too
… and vintage held fixed
-0.7pp
the gap is still there with vintage held fixed too
The largest single contributor is Office: 42 loans, $2.6B, running 15.4% where the same type runs 10.9% elsewhere — worth 2.2pp of this metro's own rate.
Maturity Wall — CMBS coming due by year, distress within eachSee what’s maturing →
$1.6B — 32% of the metro's balance — matures within two years; the red slice of each bar is already distressed.
Distress Trend — is this market turning?Track it quarter by quarter →
Realized distress is RISING — 8.7% now (2026-07), +2.3pp over the year.
Same definition as the cards above — distressed means in special servicing or 60+ days delinquent, dollar-weighted — on a different clock. The cards are today’s tape; this is the disclosed history panel, quarter by quarter, so you can read direction rather than level. Trusts file monthly, so a quarter is usually well covered within weeks of opening and the newest one shown is normally the current one; it is withheld only when its filed book falls below 60% of recent quarters.
Share of the metro’s CMBS in special servicing or 60+ days delinquent, by quarter — a firmer, backward-looking read. A loan under 1.0x DSCR that is still paying is not counted here.
Submarket Heat — where in the metro the distress sitsOpen the submarket →
The top three submarkets hold $3.2B of the metro's $5.1B; each bar's colored share is its distress rate.
Bank Door — what the local bank sector could absorbSee it on every loan’s exit read →
$1.6B of CMBS matures here within two years.
The 19 regional and local banks that gather deposits here could write roughly
$1.1B more CRE before the 300% supervisory line, so the maturing balance is
1.47× that room.
The median metro sits at 0.13×.
Regional Bank Room
$1.1B
After Committed Draws
$534M / −52%
Maturing ÷ Room
1.47×
Banks In Footprint
19 / 6 at the line
The room above is already part-sold. Construction lending is a promise drawn
down over two or three years, and the undrawn part is an obligation the bank cannot decline while the borrower performs
— every one of those dollars lands in the same CRE book the 300% line governs. These banks
have $932M of construction committed and not yet
advanced, of which $576M
comes out of the room above, leaving $534M,
with 11 banks
whose entire remaining room is spoken for.
A bank with no room contributes zero here, never a negative one — capacity does not net
across balance sheets — so $356M of what has
been promised is not deducted at all, because there is nothing left to deduct it from. On today’s
capital that is funding already contracted which would cross the line as it draws.
SR 06-26 draws a second line at 100% of capital for construction and land:
2 of 19
are past it on drawn balances alone, and
5 more cross it
once their own commitments fund.
Counted — 19 regional & local CRE lenders
| Bank | Deposit share | CRE / Capital* | Room contributed | Noncurrent CRE |
|---|---|---|---|---|
| Coastal Community Bank WA | 98.6% | 219% total 295%
|
🔒 | 0.28% |
| Washington Trust Bank WA | 24.7% | 216% total 310%
|
🔒 | 0.50% |
| 1st Security Bank Of Washington WA | 43.8% | 221% total 268%
|
🔒 | 0.73% |
| Banner Bank WA | 14.2% | 253% total 379%
|
🔒 | 0.22% |
| Northwest Bank ID | 30.4% | 177% total 224%
|
🔒 | 5.26% |
| Peoples Bank WA | 14.8% | 232% total 328%
|
🔒 | 0.00% |
| Commencement Bank WA | 69.8% | 246% total 391%
|
🔒 | 0.00% |
| Heritage Bank WA | 41.2% | 295% total 432%
|
🔒 | 0.20% |
| Timberland Bank WA | 25.3% | 289% total 378%
|
🔒 | 0.43% |
| Beneficial State Bank CA | 11.9% | 285% total 395%
|
🔒 | 3.44% |
| Mountain Pacific Bank WA | 79.6% | 296% total 444%
|
🔒 | 0.39% |
| Gbc International Bank CA | 8.5% | 269% total 379%
|
🔒 | 0.12% |
| American Continental Bank CA | 3.6% | 196% total 264%
|
🔒 | 0.00% |
| Unibank WA | 100.0% | 365% total 570%
|
🔒 | 0.81% |
| Portage Bank WA | 78.3% | 321% total 415%
|
🔒 | 2.53% |
| Sound Community Bank WA | 45.8% | 311% total 378%
|
🔒 | 0.98% |
| Us Metro Bank CA | 12.6% | 393% total 560%
|
🔒 | 1.97% |
| Mechanics Bank CA | 7.8% | 340% total 362%
|
🔒 | 0.30% |
| United Business Bank CA | 6.6% | 402% total 555%
|
🔒 | 0.48% |
Not counted — 22 banks with deposits here
These hold deposits in this metro but are left out of the capacity above, because for them deposit location stops indicating where they lend — or they do not lend CRE at all. Real money that could refinance here is deliberately not counted.
no CRE book — CRE under 100% of capital — not a CRE lender
Bank Of America, National Association NC · Jpmorgan Chase Bank, National Association OH · U.s. Bank National Association OH · Wells Fargo Bank, National Association SD · Keybank National Association OH · Hsbc Bank Usa, National Association VA · Bmo Bank National Association IL · Beal Bank Usa NV · First-Citizens Bank & Trust Company NC · The Northern Trust Company IL · Native American Bank, National Association CO · Bny Mellon, National Association PA
national — operates in more than 5 states, so deposits stop indicating where it lends
East West Bank CA · Zions Bancorporation, N.a. UT · Columbia Bank OR · Armed Forces Bank, National Association KS · Cathay Bank CA · Bank Of Hope CA · First Interstate Bank MT · Wafd Bank WA
booked here — books nearly all deposits to one branch — a charter address, not a footprint
* Two ratios, two perimeters.
The large figure is supervisory CRE over total risk-based capital —
construction, multifamily and non-owner-occupied nonfarm nonresidential. That is the book SR 06-26 draws its
300% line against, and it is what the red coloring and the room contributed are both measured on.
Total beneath it adds owner-occupied CRE — lending to a business on
its own premises — which the guidance deliberately leaves out. A bank can sit well above 300% on total CRE and
still be inside the line, with real room to lend; both are shown so the gap is visible rather than surprising.
How to read this. It is not whether any particular loan can refinance — a borrower also reaches
national banks, life companies, agencies and debt funds. It is how much of this metro’s maturing volume the local
bank sector could absorb if it were the only door.
Banks are placed by deposit share (FDIC Summary of Deposits, June 2026) — a proxy for
where they lend, not a measurement; call reports carry no collateral geography. Room = 3× total risk-based capital less CRE held
(SR 06-26, the capital base the rule names — not book equity), from Q2 2026 call reports, apportioned by that share.
Committed construction comes from FFIEC Schedule RC-L as filed at 2026-06-30 —
the undrawn half of loans already written, which the balance sheet above does not show.
Banks in more than 5 states, single-branch bookers and banks that do not lend CRE are excluded —
315 banks holding $263.8B of national capacity not counted here.
Local Lenders — banks headquartered here, worst CRE noncurrent firstUnlock each bank’s early-warning flag →
| Bank | Total CRE | CRE / Capital* | Noncurrent CRE | Early Warning |
|---|---|---|---|---|
| Mountain Pacific Bank | $502M | 296% total 444%
|
0.39% | 🔒 |
| Sound Community Bank | $442M | 311% total 378%
|
0.98% | 🔒 |
| Unibank | $181M | 365% total 570%
|
0.81% | 🔒 |
| 1st Security Bank Of Washington | $1.1B | 221% total 268%
|
0.73% | 🔒 |
| Pacific Crest Bank | $183M | 449% total 475%
|
0.62% | 🔒 |
| Wafd Bank | $10.2B | 332% total 373%
|
0.38% | 🔒 |
| Coastal Community Bank | $1.5B | 219% total 295%
|
0.28% | 🔒 |
| Commencement Bank | $321M | 246% total 391%
|
0.00% | 🔒 |
* Two ratios, two perimeters.
The large figure is supervisory CRE over total risk-based capital —
construction, multifamily and non-owner-occupied nonfarm nonresidential. That is the book SR 06-26 draws its
300% line against, and it is what the red coloring and the room contributed are both measured on.
Total beneath it adds owner-occupied CRE — lending to a business on
its own premises — which the guidance deliberately leaves out. A bank can sit well above 300% on total CRE and
still be inside the line, with real room to lend; both are shown so the gap is visible rather than surprising.
Banks are matched to their HQ metro (a proxy for footprint). Noncurrent CRE = CRE-balance-weighted blend of multifamily / nonresidential / construction noncurrent (distress already arrived). Early Warning is a validated leading signal — the bank’s 30-89 early-delinquency rank among all CRE lenders this quarter, which has predicted ~2.3× forward CRE distress. Per-bank flags are on a plan.
Do the banks lending in Seattle-Tacoma-Bellevue, WA have the capacity to refinance its maturing CRE?
Seattle-Tacoma-Bellevue, WA ranks 24 of 275 on strain, with $1.63bn of CMBS maturing across 47 loans against bank room after committed draws of $534.4mm — a wall-to-room ratio of 3.06, versus 1.47 before committed draws are deducted, so the refinancing gap is real but the local capacity question is only partly answerable: 19 qualifying banks is measured against 10 banks_excluded_here, and the exclusion is not flagged as the explanation here, meaning the strain reflects a thin visible lending footprint rather than a demonstrated lack of bank capacity, and the $5.10bn CMBS UPB with an 8.9% distressed share should be weighed alongside the fact that the maturity wall is CMBS only, an upper bound on what these banks would need to absorb.
The figures behind this answer
- CMBS maturing in the window
- $1.63bn
- … across this many loans
- 47
- Local bank room, before committed draws
- $1.11bn
- Committed construction draws
- $932.0mm
- Local bank room, after those draws
- $534.4mm
- Wall-to-room ratio
- 3.06
- Rank, most strained
- 24
- … out of this many metros ranked
- 275
- … before committed draws
- 1.47
- Population ranked against
- CMBS maturing within 24 months against regional/community bank room in this metro
- Qualifying local banks
- 19
- … excluded from the calculation
- 10
- Distressed share of this metro's CMBS
- 8.9%
- Total CMBS balance here
- $5.10bn
- Capacity band
- tight
Compared against a median wall-to-room ratio of
0.24 across the ranked metros.
3.06 is ABOVE the median of 0.24, so this metro is MORE strained than the typical ranked metro
24 of 275, counting from the MOST strained — a LOWER rank number means MORE strain
Room is what regional and community banks could still write before the SR 06-26 concentration line, after deducting construction draws already committed. Deposit footprint is a proxy for lending footprint, not a measurement of it.
The most extreme ratios belong to the smallest metros.
The wall is the CMBS balance maturing within 24 months — the maturing debt this platform can see, not the metro's whole maturity load. Every ratio is an upper bound on how much of it local banks would have to absorb.
A high ratio is as often an exclusion artifact as a credit event. Banks whose deposits do not indicate where they lend (nationals, card and charter banks) are excluded from the room, so a metro served mainly by those reads strained while its credit is simply invisible here.
Written from the figures above · CBSA 42660 · capacity_metro · last changed 25 Oct 2026
· Ask your own question →
Apartment-REIT Operating Read — same-store disclosures for this metro · as of 2026-07-30Set against the loan book →
Same-Store NOI
+1.1%
Same-Store Revenue
+1.2%
Occupancy
96.4%
Rent Growth
+1.3%
| REIT | SS NOI | SS Revenue | Occupancy | Rent | As Of |
|---|---|---|---|---|---|
| AVB | — | -0.6% | 95.9% | -0.3% | 2026-07-30 |
| EQR | +0.4% | +2.0% | 96.1% | +2.4% | 2026-04-28 |
| ESS | — | +1.7% | 96.4% | — | 2026-07-30 |
| UDR | +1.9% | +1.9% | 97.3% | +1.7% | 2026-07-27 |
Same-store operating results disclosed by public apartment REITs (AVB · EQR · ESS · MAA · CPT · UDR) for this market in their quarterly supplements. Directional market context — an operating trend, not a Verstavo score.
On The Ground — recent closures, layoffs & CRE bankruptciesTie these events to loans →
145 local distress events in the past year (9,295 jobs) — store closures, layoffs, and commercial real estate bankruptcies near this metro's collateral.
2026-11-02
LAYOFF
Scheduled · 93 jobs · Seattle
2026-11-07
LAYOFF
Scheduled · 33 jobs · Auburn
2026-11-13
LAYOFF
Scheduled · 359 jobs · Seattle
2026-10-26
BANKRUPTCY
CRE-linked bankruptcy
2026-10-23
CLOSURE
Seattle
2026-10-21
BANKRUPTCY
CRE-linked bankruptcy
2026-10-14
CLOSURE
Seattle
2026-10-09
CLOSURE
Seattle
2026-09-23
CLOSURE
Seattle
2026-09-22
CLOSURE
Seattle
2026-09-21
CLOSURE
Tacoma
2026-09-10
CLOSURE
Redmond
2026-09-01
CLOSURE
Tukwila
What has actually happened on the ground in Seattle-Tacoma-Bellevue, WA recently?
Over the trailing window_days of 365, the on-the-ground reading for Seattle-Tacoma-Bellevue, WA shows 70 store_closures and 77 warn_notices affecting 9,436 jobs_affected — a figure that is a floor, since notices stating no headcount are counted but contribute nothing. CRE distress is thinner on this evidence: 1 cre_bankruptcies, though that is a STATE PROXY rather than a metro-native count, so it should not be read as one filing tied to local collateral.
The figures behind this answer
- Store closures
- 70
- WARN layoff notices
- 77
- Jobs on those notices
- 9,436
- CRE-related bankruptcies
- 1
- Window, in days
- 365
STATE PROXY — a court case counts in its filing district and an SEC 8-K in the filer's state, not where the property is, so this counts CRE-likely filings in the states this metro's collateral sits in. It is not a metro-native count. Each court case is counted once even when both PACER and CourtListener carry it. A court case's state is its filing district; an SEC 8-K or news item carries the filer's state. Coverage widened to both court feeds on 2026-09-16, so counts before and after that date are not comparable.
affected_employees is nullable — notices that state no headcount are counted but contribute 0 jobs, so this is a floor.
Closures are matched ZIP -> CBSA and only APPROVED rows count (a pending, machine-extracted row is not yet a closure). This mirrors the Metro Pulse page, not the convergence board — the board applies neither filter and counts higher. Layoffs take the same ZIP -> CBSA and review treatment, and a WARN notice counts if it affects >= 10 people or does not state a headcount (a NULL is not 'fewer than 10').
Written from the figures above · CBSA 42660 · geo_events · last changed 25 Oct 2026
· Ask your own question →
What this page can’t do
Everything above is a snapshot — one market, as it stands today, and
it’s yours to read. What a snapshot can’t tell you is when it moves: which loan slipped a DSCR band
this month, which bank’s early-warning rank crossed into the top quartile, which submarket started turning
while the headline number still looked fine. Distress doesn’t announce itself on the day you happen to check.
This page
One metro · one moment · you come looking.
Verstavo
Every metro · every month · it comes looking for you.
CMBS covers the securitized slice of CRE; bank, closure and layoff signals widen the view.
“No data” in a zone means we don’t observe it here, not that there is no stress.
Employment is BLS (QCEW annual, LAUS monthly) — public data, and the demand leg under all of it.
See the whole picture, not just the pulse.
Metro Pulse is the free, public read — one page for every market we cover. The platform goes loan‑by‑loan inside it: which buildings carry the stress, when their debt comes due, and your own portfolio benchmarked against the market you are standing in.