Q3 Concentration: Residual Top-1 Eases to 38% — WBG Holds 34% of the $35B MDB Tip

Our 2026 concentration print answered the first distribution question: residual Top-1 (uninsured households & SMEs) ~40%, Top-3 ~87%, protection-gap Top-3 regions ~53% of the $424B stock, and OECD adaptation donors Top-3 ~46% of a $34.7B tip. The Q3 OECD vintage and August MDB + H1 refresh then moved the levels: MDB LMIC adaptation to $35B in 2025 (+31% YoY), H1 2026 insured nat-cat $42B with a 42% insurance ratio. This Q3 concentration lens asks the desk follow-up: did those prints change who sits at the top of the residual and finance ladders — or mostly rearrange absolute dollars under a stuck tip?

The interactive dashboard above is built as that Q3 concentration lens. Toggle HHI scoreboard, Residual ladder, MDB & donors, Vintage slope, and Gaps & scarcity. The punchline is deliberately multi-sided. On residual incidence, Top-1 eases to about 38% (−2 pp versus the FY framing) and Top-3 to about 86% as the benign H1 insurance ratio lifts the covered sleeve — HHI near 2,684. On MDB LMIC adaptation banks, World Bank Group alone is about 34% of the $35B tip and Top-3 (WBG + ADB + IDB) clears about 68%. On protection-gap geography, Top-3 regions still hold about 52.6% of the $424B stock — unchanged geography under a newer finance print. On scarcity, AGR needs mid still runs roughly 9.6× MDB LMIC adaptation.

The Q3 concentration scoreboard

Read the table as a family of market shares, not one number. The residual tip eased slightly because insurance’s constructed share rose with H1’s 42% coverage ratio — not because households stopped paying. The MDB bank tip tightened slightly as the Joint Summary made the institutional engine larger and more legible. Gap geography and OECD donors are the stuck meters: new flow prints did not redraw who holds uninsured exposure dollars or who supplies the bilateral tip.

HHI scoreboard: which tip is actually hard

Open HHI scoreboard. Ranked HHI bars put residual incidence and public instruments in the high / extreme bands, MDB banks in the moderate-to-high band (~1,986), gap geography moderate (~1,518), and OECD donors still plural (~1,124). The paired Top-1 / Top-3 bars make the same point in share space: instruments are loan-dominated at 67% Top-1; residual Top-3 clears 86%; donor Top-3 only 46%.

That ordering matters for narrative risk. Press coverage that celebrates “MDBs hit a record” is true on the August update’s levels — and still describes a tip that is more concentrated inside the MDB club than the OECD bilateral club is. Analysts who quote only donor pluralism miss that the residual bearer ladder remains the hardest concentration meter in the theme.

Residual ladder: −2 pp is not a regime change

Filter Residual ladder. The Lorenz curve still rises far above the equal-split diagonal: 38% at households, 64% once sovereigns stack, 86% once insurance’s covered share is included. Toggle Share %, Cumulative %, and Δ vs FY pp. The delta panel is the Q3-specific control: households −2 pp, sovereigns −1 pp, insurance +2 pp, MDB adapt +1 pp. The tip moved toward the covered sleeve because H1 2026 was a benign half — Swiss Re’s insured $42B against economic ~$100B — not because the $424B protection-gap stock closed.

Pair that with the prior concentration lens. FY framing put households at 40% and insurance at 20%. Q3’s H1-adjusted panel is a vintage sensitivity, labeled constructed: useful for desks that mark residual risk to the newest damage print, dangerous if mistaken for a structural deconcentration of who pays when the next severe season arrives. Insurance remains the third rung, not Top-1.

MDB banks and donors: institutional tip vs bilateral tip

Switch to MDB & donors. On the estimated bank shares of MDB LMIC adaptation $35B (2025), World Bank Group leads near 34% (~$11.9B), ADB 20%, IDB 14%, AfDB 11%, EIB LMIC sleeve 9%, other MDBs 12%. Top-3 clears about 68%. These bank shares are estimated from published MDB climate-finance patterns — not an official adaptation-only extract — and are labeled as such in the dashboard source note.

The OECD donor bars beside them stay the milder tip: Germany ~18%, Japan 16%, France 12%, Top-3 46% of $34.7B. The instrument donut still shows 67% loans / 29% grants. So the Q3 story is not “finance democratised.” It is “the largest institutional engine got larger (+31% YoY) while remaining bank-concentrated, and the bilateral tip remained plural-but-thin under a loan-heavy instrument mix.” The August update tracks the level bounce; this lens tracks who inside the club holds the tip.

Vintage slope: residual eases, gap Top-3 stuck

Open Vintage slope. The multi-vintage line walks Research → Concentration 2026 → Q3/Aug. Residual Top-1 drifts 42% → 40% → 38%; residual HHI ~3,010 → ~2,826 → ~2,684; gap Top-3 stays near 53%. Toggle slope metric among residual Top-1 / Top-3 / HHI / gap Top-3, and optionally overlay MDB Top-1 (null on the research vintage, then 32% → 34%).

The insured-share composed panel under the same view is the damage-side twin: FY 2025 insured share ~29%, H1 2025 a severe half, H1 2026 benign at 42%. A higher insurance ratio in a quiet half-year is not a resilience-index rewrite — global resilience remains ~27% on the Swiss Re-style stock used across this theme. Desks that mark “adaptation is working” from H1 insured totals alone are sampling the covered sleeve.

Gaps, scarcity, and the FRLD micro-tip

Toggle Gaps & scarcity. Protection-gap geography is unchanged in ranking: North America ~$95B (22.4%), South Asia, LAC, SSA, Western Europe, advanced APAC, MENA. Filter by Income and the stress quadrant on the resilience×gap scatter still lights up South Asia and SSA — low resilience, large absolute gap. Scarcity ledgers keep AGR needs mid (~$338B), the implied gap versus MDB $35B, OECD $34.7B, CPI $65B, UNEP $26B, and the FRLD delivered micro-print (~$0.45B of ~$0.82B pledged). Do not sum across ledgers.

The FRLD row is the extreme thin tip: even if Top-3 pledgers concentrate ~61% of a sub-billion pledge book, delivery remains ~$449M against multi-hundred-billion needs. Loss-and-damage capitalization is a political milestone and a rounding error on the residual ladder. Our US billion-dollar weather disasters piece tracks the rich-country insured tape; this lens shows why that tape plus a sub-billion FRLD window still leaves households as Top-1.

Who is exposed — and what would change the story

Exposed: uninsured households and SMEs whose constructed share only fell 2 pp on a benign H1; LMIC sovereigns borrowing into a 67% loan-heavy public tip while contingent disaster debt still hits budgets; regional corridors (SSA, South Asia) stuck in the low-resilience / large-gap quadrant; desks that treat MDB +31% YoY or H1 insured $42B as proof the residual tip deconcentrated.

Relative winners under current rules: advanced-economy insured systems that printed a higher H1 coverage ratio; World Bank Group and the Top-3 MDB bloc inside the $35B LMIC adaptation engine; grant niches inside the 29% public-adaptation grant sleeve; investors and cities that price residual incidence by bearer and bank rather than by headline insured loss alone.

What would change the story: a severe season that pushes the insurance ratio back toward the FY ~29% framing and restores household Top-1 near 40%+; official MDB adaptation bank tables that break the estimated 34% / 68% tip; donor and grant diffusion that lifts UNEP/OECD/MDB numerators toward the $310–365B needs band; or resilience in developing regions that lifts the global index well above ~27%. None of those dominate the Q3+Aug official vintages summarised here.

Caveats and methodology

• Residual bearer shares are a constructed incidence panel with an explicit H1-adjusted sensitivity versus the FY framing in the 2026 concentration postnot a single audited global ledger of who paid last year’s losses.
• MDB bank shares of LMIC adaptation are estimated from published MDB climate-finance patterns and labeled estimatednot an official bank-by-bank adaptation extract from the Joint Summary.
• Regional protection-gap dollars are estimated allocations of the Swiss Re-style $424B stock; geography illustrates uninsured exposure, not a country extract.
• Donor country shares of OECD adaptation remain estimated from bilateral patterns.
• OECD, UNEP, MDB, CPI, and FRLD figures must not be summed; each answers a different accounting question.
• HHI values are analytical indexes on stated bucket shares (0–10,000), comparable within this dashboard.
• H1 2026 insured nat-cat $42B is a half-year damage print; it does not revise FY 2025 insured $107B or the $424B protection-gap stock.
• Needs band $310–365B/yr by 2035 is UNEP AGR 2025 (2023 prices)developing-country adaptation needs, not a global spending ceiling.
• FRLD pledge shares are a constructed micro-concentration meter on a sub-billion book; treat as illustrative tip geometry.

The shareable takeaway

Q3 adaptation-economics concentration is a tip that eased on damage accounting and tightened on institutional finance — without rewriting scarcity. Residual Top-1 (households & SMEs) prints about 38% (−2 pp vs FY), Top-3 about 86%. World Bank Group holds about 34% of the $35B MDB LMIC adaptation tip; Top-3 banks ~68%. Protection-gap Top-3 regions remain ~53% of the $424B stock. OECD donors Top-3 stay ~46% of a $34.7B tip that is still 67% loans. Needs mid still runs ~9.6× the MDB engine. Who pays before policy catches up is still a distribution problem: the residual tip is thick, the institutional tip is bank-concentrated, and the needs ceiling still sits an order of magnitude above both.

Related reading: 2026 concentration lens, August MDB + H1 update, Q3 OECD update, adaptation economics research, and US billion-dollar weather disasters.