Q3 Concentration: Top-1 Bank Still ~12.8% of Deposits — Top-3 33%; CMBS +51 bp, SLOOS CRE Flips to Ease

Our prior concentration print answered the distribution question with a multi-meter map: deposit top-1 / top-3 near 12.8% / 33.4%, $1–10B banks at 311% CRE/capital versus a 200% industry median, mega banks ~48% of CRE PDNA dollars, cards ~48% of charge-off dollars, and office ~42% of CMBS delinquent balances. The Q3 theme update then refreshed the process tape — July SLOOS CRE NFNR at −11.3 net easing and Trepp July CMBS overall at 7.86%. This Q3 concentration cut asks the desk follow-up: did those prints rewrite the top-of-distribution shares, or only refresh the meters that sit beside a still-sticky architecture?

The interactive dashboard above is built as a Q3 concentration lens. Toggle Scoreboard, Bank shares, CRE cohorts, Stress + HHI, and Supply + CMBS. Use deposit/asset metrics, CRE sort keys, charge-off versus delinquency dollar shares, SLOOS July-net versus Δ-vs-prior, and CMBS pie modes (delinquent $ / rate / MoM Δ). The punchline is deliberately flat on firm shares and sharp on process. Top-1 deposits remain ~12.8% and top-3 ~33.4%. Peak CRE/capital remains 311%. What Q3 adds is a disclosed CMBS MoM path (+51 bp overall; multifamily +1.05 pp), a SLOOS supply flip into net easing on CRE, and analytical HHIs that show stress-dollar tips (charge-offs ~3,420, CRE PDNA ~3,180) sit far above deposit-firm HHI (~680).

The headline ladder: Q3 vs prior restatement

Read the table as a stability-plus-process scoreboard. Q3 did not re-rank the deposit tip or the CRE capital cohort ladder. It raised several CMBS rates, flipped CRE credit-supply nets into easing, and left stress-dollar shares sticky even as rates moved. Desks that treat “SLOOS eased” as “concentration eased” are reading the wrong column.

Scoreboard: carried architecture, Q3 process flags

Open Scoreboard. The prior → Q3 restatement bars keep deposit Top-1 / Top-3, peak CRE/capital, and office delinquent-$ share flat. CMBS overall, office, and multifamily rates step up; SLOOS CRE NFNR flips from a prior estimated tighten into −11.3 net easing. The Top-1 vs Top-3 scatter (bubble size ∝ HHI) still puts charge-off and CRE PDNA stress lenses in the extreme corner, deposit/asset firm shares mid-left, and office CMBS stress dollars upper-right.

That is the first Q3 hinge. Supply easing and share concentration are different objects. A −11.3 SLOOS net print does not move JPMorgan’s ~12.8% deposit weight. A +51 bp CMBS MoM does not automatically shrink office’s ~42% of delinquent balances — rates can rise inside a sticky share geometry. Pair this with the Q3 vintage update: that post owns the level tape. This lens owns the distribution question.

Bank shares: Lorenz confirms a thick tip, not a monopoly

Filter to Bank shares. Ranked deposit bars still put JPMorgan first near 12.8%, Bank of America near 11.4%, Wells Fargo near 9.2%, and Citigroup near 7.2% — top-4 around 41%, long tail still above half of domestic deposits. Switch to assets and the order is familiar. The Q3 addition is the deposit Lorenz against a 45° equal-share diagonal: cumulative deposit share stays below the diagonal through the named tip, then the residual “Rest” bucket closes the curve — a visual that firm concentration is moderate, not winner-take-all.

The cumulative top-N ladder still clears ~33% by top-3, ~41% by top-4, and ~54% by top-10. Deposit-firm HHI near 680 is the analytical twin of that geometry — far below stress HHIs on the Stress + HHI panel. Treat firm ranks as estimated order-of-magnitude from FDIC QBP / Summary of Deposits framing, rounded for visualization.

CRE cohorts: capital intensity and stress dollars still disagree

Switch to CRE cohorts. Median CRE / (Tier 1 + ACL) still peaks at 311% in the $1–10B band, 289% in $10–100B, industry ~200%, mega banks near 95%. PDNA rates still run the other way: mega banks near 1.67% versus mid/community under 1%. Stock-versus-stress bars keep mega banks at ~38% of CRE stock and ~48% of CRE PDNA dollars.

Q3 does not invent a new size census — the FDIC Risk Review framing and the prior concentration print still own that architecture. The Q3 contribution is juxtaposition: CRE supply is easing on SLOOS while CRE capital concentration and PDNA dollar concentration remain the same two disagreeing maps. Easing standards do not automatically rebalance who holds the past-due dollars.

Stress + HHI: cards own losses; analytical indexes confirm the tip

Open Stress + HHI. Sorting by charge-off dollar share still puts cards near 48%, then C&I near 26%, with CRE only ~8% of realized loss dollars. Delinquency dollar share lifts residential and CRE — past-due stock without matching charge-off velocity. Fed SA still closes on 2026Q1: cards 3.84% charge-off / 2.92% delinquency; CRE 1.56% delinquency / 0.17% charge-off (~9× multiple).

The HHI panel is the Q3 analytical addition. Deposit and asset firm HHIs sit near 680–705. CRE PDNA dollar HHI ~3,180, CMBS delinquent-$ HHI ~2,680, and charge-off dollar HHI ~3,420. Those indexes are constructed on stated share buckets (0–10,000 scale) — not a regulator filing — but they make the geometric claim falsifiable: stress dollars concentrate harder than deposit firms. Our bank loan charge-offs post tracks the loss tape; this lens tracks who owns the dollars inside that tape and how tip-heavy the analytical index sits.

Supply + CMBS: July easing meets sticky office share

Toggle Supply + CMBS. July SLOOS nets print CRE NFNR −11.3, multifamily −5.7, construction & land −3.7, C&I large 0.0, C&I small +1.8, and cards still +6.7 net tightening. Flip to Δ vs prior estimated tighten and CRE NFNR’s −23 pp swing is the visual hinge — the largest supply re-pricing in the theme’s Q3 window. Cards remain the product that tightens while CRE eases: supply concentration is asymmetric across the same loan book that already concentrates losses in cards.

The CMBS prior-versus-July bars show multifamily as the fastest MoM (+1.05 pp to 7.69%), office still the highest rate (11.91%, +0.6 pp), overall 7.86% (+51 bp). The pie — whether delinquent-$ share, rate, or MoM Δ — keeps office near 42% of delinquent balances. That is the second Q3 hinge. Rate heat and share architecture can move on different clocks. Multifamily’s MoM spike does not yet dethrone office’s dollar tip; industrial remains the calm residual.

Who is exposed — and what would change the story

Exposed: desks that translate SLOOS CRE easing into thinner CRE PDNA dollar concentration while mega banks still hold ~48%; CMBS books that treat multifamily’s +1.05 pp as a finished re-ranking of delinquent-$ shares while office stays ~42%; consumer lenders whose charge-off share remains near half of industry loss dollars while cards are the one major product still net-tightening on SLOOS; $1–100B banks still printing 290–310% CRE/capital even as standards ease; policymakers who quote only deposit top-3 ~33% as the concentration meter for commercial credit.

Relative winners under current rules: diversified mega banks with lower CRE/capital ratios and capital-markets access; industrial/logistics CMBS sleeves where delinquency stays low-single-digit; regional books tilted away from office; desks that keep the August maturity / special-servicing tape and this Q3 share map in separate columns instead of averaging “easier standards = less concentration.”

What would change the story: a sustained collapse of office CMBS delinquency that breaks office’s ~42% delinquent-dollar share; FDIC prints that bring $1–10B CRE/capital medians back toward 200%; Fed SA where cards cease to dominate charge-off dollars; or a multi-survey SLOOS path that eases cards as hard as CRE while charge-off shares rotate. None of those clear this July / 2026Q1 window.

Caveats and methodology

• Bank asset and deposit shares are rounded FDIC QBP / SoD–style ranksorder-of-magnitude concentration, not a live league table after every acquisition.
• CRE/capital medians and CRE PDNA by size follow FDIC 2026 Risk Review disclosed framing where labeled; some small-bank cells are estimated.
• CRE stock / PDNA dollar shares and loan-book stress-dollar shares are estimated reallocations that sum to 100% for visualization.
• HHI values are analytical indexes on stated bucket shares (0–10,000), not a published FDIC concentration filing.
• SLOOS net % = share of banks tightening minus easing (July 2026 survey). Prior estimates used for Δ are theme-carried, not a second official vintage.
• CMBS property rates follow Trepp July 2026; delinquent-dollar shares are estimated from rates × approximate stock mixes.
• Fed SA closes on 2026Q1 (May 19, 2026) while CMBS/SLOOS are July 2026do not force a single as-of date.
• Concentration ≠ causation. Easier CRE standards do not automatically shrink PDNA dollar tips; higher CMBS rates do not automatically rewrite firm deposit shares.

The shareable takeaway

In the Q3 2026 concentration vintage, deposit Top-1 / Top-3 still sit near 12.8% / 33.4%. Peak CRE/capital remains 311% in the $1–10B band. Mega banks still hold ~48% of CRE PDNA dollars; cards ~48% of charge-off dollars; office ~42% of CMBS delinquent balances. What Q3 changed is the process overlay: CMBS overall 7.86% (+51 bp MoM), multifamily +1.05 pp, and July SLOOS CRE NFNR at −11.3 net easing while cards stay net-tight. Firm tip sticky. Stress tip skewed. Supply easing concentrated in CRE — not in the products that already own realized losses.

Related reading: Prior concentration print · Q3 theme update · August vintage · Research roll-up · CRE size split · Bank loan charge-offs.