---
title: the loss moved down the tower
canonical_url: https://ensurance.app/guide/the-loss-moved-down-the-tower
markdown_url: https://ensurance.app/guide/the-loss-moved-down-the-tower.md
subtitle: "what a reinsurance attachment point is, why it rose in 2023, and the layer below every deductible that the tower does not fund"
category: nature-finance
---

# the loss moved down the tower

*what a reinsurance attachment point is, why it rose in 2023, and the layer below every deductible that the tower does not fund*

A **reinsurance attachment point** is the size of loss at which a reinsurance layer starts to pay. Below it, the insurer that bought the cover (the **cedent**) pays. Above it, the reinsurer pays, up to the layer's limit. At the 1/1/2023 renewals reinsurers moved those points up. S&P found that primary insurers took most of the severe convective storm losses that followed, and loss that cedents keep can travel further. Underneath all of it sits ground the tower does not pay for: the fuel in a forest above a town, the floodplain left open or filled, the marsh in front of a street. For wildfire and flood, they help set how often the bottom of the tower gets hit.

## what a reinsurance attachment point is

A catastrophe excess-of-loss program is drawn as a **tower**: stacked **layers**, each written as "limit xs attachment." A layer of $100m xs $100m pays losses from a single event above $100m, up to $200m. The bottom attachment of the program is the cedent's **retention**: the part of each event it keeps on its own balance sheet.

**Exhaustion** is the point where a layer, or the whole tower, has paid its full limit. Loss above it comes back to the cedent, along with loss below the retention. **Rate on line** (ROL) is the premium for a layer divided by its limit: $10m of premium for $100m of limit is a 10% rate on line. An **aggregate cover** responds to the sum of losses across a year, above an annual aggregate retention, rather than to one event. It is the cover a cedent buys for frequency: many mid-sized storms, hail events and fires, none of which reach the per-event tower alone.

A hypothetical regional carrier, in round numbers: These figures are illustrative, not drawn from any company or source. The size of the move varied by cedent; the sources give no market-wide figure.

| layer (hypothetical cedent) | before the reset | after the reset |
|---|---|---|
| Retention (the cedent keeps) | first $50m of each event | first $100m of each event |
| Layer 1 | $50m xs $50m | absorbed into the retention |
| Layer 2 | $100m xs $100m | $100m xs $100m |
| Layer 3 | $200m xs $200m | $200m xs $200m |
| Aggregate cover | $100m xs $150m annual (on retained loss) | not renewed |

After the reset, a $90m hailstorm costs the cedent all $90m. Before it, the cedent paid $50m. A year of four $60m events left the cedent $50m net each, $200m in all, which reached into the aggregate cover. Now it is $240m of net loss with nothing behind it.

:::johnson
**part of the bottom of the tower is set on the ground.** Reinsurance decides where a loss attaches. For wildfire and flood, the forest, floodplain and marsh under the exposure help decide how often it gets there.

[See what an insurer can fund →](/solutions/insurers?from=guide&topic=reinsurance)
:::

## the 1/1/2023 reset

At 1/1/2023, Guy Carpenter's global property catastrophe rate-on-line index rose 27.5%, the largest increase since 2006, and its US index reached an all-time high (Guy Carpenter, 2023, via Reinsurance News). The same report notes that the index understates the change, because retentions rose at the same time. A rate index compares price per unit of limit. It does not count the loss a cedent took back by attaching higher.

Aggregate covers were among the covers cut. Zurich did not renew its global catastrophe aggregate for 2023. IAG's aggregate was cut from USD 350m to USD 250m and attached higher (Artemis, 2023). S&P Global described the pattern across the market: reinsurers cut aggregate covers, moved up towers and cut their frequency exposure, leaving primaries most of the severe convective storm losses from 2023 through mid-2024 (S&P Global, 2024).

Moving up the tower is a sound response when the bottom of it is hit more often than it is priced for. The question is where that loss went.

## the softening, and what the price figures do not show

Prices have since come down. At 1/1/2025, Guy Carpenter's global rate-on-line index fell 6.6%, the first fall since 2017 (Guy Carpenter, 2025). At 1/1/2026, Howden Re put risk-adjusted property catastrophe rates down 14.7%, the largest fall since 2014, and retrocession (reinsurance bought by reinsurers) down 16.5%; Guy Carpenter's global index was down 12% (Howden Re and Guy Carpenter, 2026, via Artemis). At the 6/1/2026 Florida renewals, Howden reported risk-adjusted declines of up to 25%, and warned that a repeat could push large segments of the market below the cost of capital by 2027 (Reinsurance News, 2026). Fitch kept a "deteriorating" outlook for global reinsurance in 2026, and S&P said pricing has peaked (Insurance Business, 2026).

The sources above report price, not attachments or aggregate covers, so we make no claim that either returned to 2022 levels. Swiss Re sigma 1/2026 puts the trend in insured natural catastrophe losses at 5% to 7% a year in real terms (2025 came in at USD 107bn against a trend-implied USD 145bn or so), with a record 92% of 2025's insured losses from secondary perils (Swiss Re, 2026). Cheaper reinsurance on top of a rising loss trend is a cycle, not a fix. Swiss Re attributes most of loss growth to exposure (values, urbanization, rebuild cost), with climate amplifying some perils (Swiss Re, 2025, via Risk & Insurance).

## where the retained loss lands

When the reinsurer moved up the tower, the loss did not move with it. It moved down — to the cedent, the homeowner, and the condition of the ground.

| who | how retained loss can reach them |
|---|---|
| The cedent's net | Higher retentions and fewer aggregates put frequency losses on the primary balance sheet (S&P Global, 2024) |
| The deductible | A carrier that keeps more of each event may pass more of each claim to the policyholder. We have not found a figure for how much |
| Policyholders | One example of a primary insurer pulling back (the sources do not tie it to attachment points): State Farm stopped taking new California property applications in May 2023 and non-renewed about 72,000 policies in 2024 (United Policyholders, 2023–24) |
| The state | A proposal for who could carry it: the EIOPA–ECB joint paper proposes an EU public-private reinsurance scheme and an EU disaster-financing fund (EIOPA, 2024) |
| No one | In 2024, 57% of USD 318bn in economic natural catastrophe losses were uninsured (Swiss Re sigma 1/2025; see [the protection gap is a condition gap](/guide/the-protection-gap-is-a-condition-gap?from=guide)) |

Each row is a way of carrying the same loss. None of them changes it. A cedent can keep a policyholder before it cuts the price ([keep the policy before you cut the price](/guide/keep-the-policy-before-you-cut-the-price?from=guide)). Keeping it does not touch the hazard that produced the claim, which is why moving risk around is not the same as lowering it ([there is no risk transfer](/guide/there-is-no-risk-transfer?from=guide)).

## the layer below the deductible

Below the reinsurance, below the retention, below the homeowner's deductible, there is a layer the tower does not pay for: the condition of the ground. Public agencies, landowners and conservation groups already work there; the tower does not price or fund it.

A forest's fuel load sets how a fire runs toward the houses below it. A floodplain left open holds water that a filled one sends into the street. A salt marsh in front of a neighborhood slows waves before they reach the doors; in Ocean County, New Jersey, salt marshes cut modeled average annual flood losses 16% (Narayan et al., 2017). Those living systems do that work whether anyone pays for them or not. A marsh is not worth only the claims it avoids. The avoided claim is how a balance sheet learns to see it. When they degrade, the loss they were holding back can show up as more frequent mid-size loss: more claims in the retention, more hits on the bottom layer, more years where an aggregate would have mattered.

The Swiss Re Institute found that in Florida, from 2009 to 2022, areas with the best-protected coastal habitat had about half the National Flood Insurance Program claim frequency in frequent storms (Swiss Re Institute, 2025). The Nature Conservancy and WTW found that ecological forestry in the Northern Sierra lowered modeled average annual loss by 40% to 60% (TNC and WTW, 2021): a modeled result for treated landscapes, not an observed change to a reinsured layer. Those are the field's results, measured and modeled by others. They are not ensurance outcomes.

## what ensurance does in that layer

**Ensurance** funds the condition of a named place before the season: a certificate tied to a forest, floodplain or marsh, with its condition read beside its price. (A certificate is a record of funding tied to that place; see [specific ensurance](/specific?from=guide).)

Here is what that can and cannot do for a tower:

- It can reduce how often losses reach the retention and the bottom layer, mostly flood and wildfire losses in the zone where the work is done. Wildfire is partly severity- and wind-driven, so the effect there is the least certain.
- It does little for hail and other severe convective storm losses, the peril S&P says primaries kept. Those are weather, not ground. The work reaches wildfire and flood in the zone where it is done.
- It does nothing for the top layers. A Category 5 hurricane at landfall or a large earthquake runs through every layer whatever the ground looks like, and peak-peril tail capital is untouched.
- It does not reinsure anyone. Ensurance holds no tail risk and writes no cover. The cedent's tower stays where its reinsurers put it.
- No reinsurer or cedent has booked a loss reduction from ensurance-funded work.

For a cedent that keeps more of each event than it used to, a lower expected loss in the retention layer is the part of the tower it can work on itself. The case for funding the ground is that it can lower the loss the cedent now owns, measured with and without the work ([how an intervention lowers the loss](/guide/how-an-intervention-lowers-the-loss?from=guide)). If part of that shows up as a premium saving, some of it can go back to the land that produced it ([share the premium saving with the land](/guide/share-the-premium-saving-with-the-land?from=guide)).

If you buy reinsurance for a carrier, [see how this applies to insurers](/solutions/insurers?from=guide&topic=reinsurance), or [talk to someone about a specific zone](/contact?from=guide&topic=reinsurance).

Read next: [the model learns from a past that is leaving](/guide/the-model-learns-from-a-past-that-is-leaving?from=guide).

## frequently asked questions

### what is a reinsurance attachment point?

A reinsurance attachment point is the loss amount at which a reinsurance layer begins to pay. In a layer written "$100m xs $100m," the attachment point is $100m: the reinsurer pays losses above it, up to $100m more. The lowest attachment in a program is the cedent's retention.

### did attachment points come back down when reinsurance prices fell?

Rates on line fell at 1/1/2025, 1/1/2026 and 6/1/2026 by the measures cited above. The sources we cite report prices, not attachments, so we do not claim retentions have returned to 2022 levels.

### can ensurance lower a cedent's reinsurance attachment point?

No. Attachment points are set by the cedent and its reinsurers. Funding the condition of the ground can reduce how often losses reach the retention, mostly for frequency perils like wildfire and flood. No such reduction has been booked yet.

## sources

[Guy Carpenter via Reinsurance News, global property cat rates on line up 27.5% at Jan 1, 2023](https://www.reinsurancene.ws/global-property-cat-rates-on-line-up-27-5-at-jan-1-renewals-guy-carpenter/) — 1/1/2023 index +27.5%, largest since 2006; US index at an all-time high; index understates because retentions rose

[Artemis, Zurich opts to non-renew its cat aggregate reinsurance for 2023, 2023](https://www.artemis.bm/news/zurich-opts-to-non-renew-its-cat-aggregate-reinsurance-for-2023/) — Zurich's global catastrophe aggregate not renewed

[Artemis, IAG's aggregate reinsurance shrinks, attachment rises at renewal, 2023](https://www.artemis.bm/news/iags-aggregate-reinsurance-shrinks-attachment-rises-at-renewal/) — IAG aggregate cut from USD 350m to USD 250m at a higher attachment

[S&P Global, global reinsurance highlights 2024, 2024](https://insight.spglobal.com/story/global-reinsurance-highlights-2024/page/4) — Reinsurers cut aggregates, moved up towers, cut frequency exposure; primaries took most 2023–H1 2024 SCS losses

[Guy Carpenter, January 1, 2025 renewals news release, 2025](https://guycarp.com/content/dam/guycarp-rebrand/pdf/pressreleases/News-release-Jan-1-2025-renewals.pdf) — 1/1/2025 global rate-on-line index −6.6%, first fall since 2017

[Howden Re and Guy Carpenter via Artemis, property cat down 14.7%, retro down 16.5% at January 2026 renewals, 2026](https://www.artemis.bm/news/property-cat-reinsurance-down-14-7-retrocession-down-16-5-at-jan-2026-renewals-howden-re) — Howden risk-adjusted −14.7% and retro −16.5%; Guy Carpenter global −12%

[Reinsurance News, property cat ROL declines accelerate at June 1 renewal, 2026](https://www.reinsurancene.ws/property-cat-rol-declines-accelerate-at-june-1-reinsurance-renewal-howden-re/) — Florida declines up to 25%; Howden cost-of-capital warning for 2027

[Insurance Business, Fitch maintains deteriorating outlook for global reinsurers in 2026, 2026](https://www.insurancebusinessmag.com/reinsurance/news/breaking-news/fitch-maintains-deteriorating-outlook-for-global-reinsurers-in-2026-561486.aspx) — Fitch "deteriorating" outlook; S&P "pricing has peaked"

[Swiss Re, sigma 1/2026, natural catastrophes in 2025, 2026](https://www.swissre.com/institute/research/sigma-research/sigma-2026-01-natcat-2025-wildfire-storm-risk.html) — Insured losses trending 5–7% a year in real terms; record 92% of 2025 insured losses from secondary perils

[Swiss Re, sigma 1/2025, natural catastrophes trend, 2025](https://www.swissre.com/institute/research/sigma-research/sigma-2025-01-natural-catastrophes-trend.html) — 2024 economic losses USD 318bn, 57% uninsured

[United Policyholders, State Farm will not renew thousands of California policies, 2023–24](https://uphelp.org/state-farm-will-not-renew-thousands-of-california-insurance-policies-over-the-next-year/) — New California property applications stopped May 2023; about 72,000 non-renewals in 2024

[EIOPA and ECB, towards a European system for natural catastrophe risk management, 2024](https://www.eiopa.europa.eu/publications/eiopa-and-ecb-joint-paper-towards-european-system-natural-catastrophe-risk-management_en) — Proposed EU public-private reinsurance scheme and disaster-financing fund

[Risk & Insurance on Swiss Re, natural catastrophe losses set to reach USD 145bn in 2025, 2025](https://riskandinsurance.com/natural-catastrophe-losses-set-to-reach-145-billion-in-2025-with-1-in-10-risk-of-300-billion-peak-year/) — Loss growth mostly exposure (values, urbanization, rebuild cost); climate amplifies some perils

[Narayan et al., Scientific Reports 7:9463, 2017](https://doi.org/10.1038/s41598-017-09269-z) — Ocean County, New Jersey salt marshes cut average annual flood losses 16%, modeled with industry catastrophe models

[Swiss Re Institute, natural habitats can reduce flood losses, 2025](https://www.swissre.com/institute/research/topics-and-risk-dialogues/climate-and-natural-catastrophe-risk/coastal-and-flood-protection-natural-habitats.html) — About half the NFIP claim frequency in best-protected Florida coastal-habitat areas, frequent storms, 2009–2022

[TNC and WTW, wildfire resilience insurance, 2021](https://wtwco.com/en-US/Insights/2021/07/wildfire-resilience-insurance-quantifying-the-risk-reduction-of-ecological-forestry-with-insurance) — Northern Sierra ecological forestry lowered modeled AAL 40–60%

## the series

1. [what reinsurance cannot cede](/guide/what-reinsurance-cannot-cede?from=guide)
2. [secondary perils are not secondary](/guide/secondary-perils-are-not-secondary?from=guide)
3. [the loss moved down the tower](/guide/the-loss-moved-down-the-tower?from=guide)
4. [the model learns from a past that is leaving](/guide/the-model-learns-from-a-past-that-is-leaving?from=guide)
5. [capital is not the shortage](/guide/capital-is-not-the-shortage?from=guide)
6. [what a reinsurer can fund before the season](/guide/what-a-reinsurer-can-fund-before-the-season?from=guide)