Guy Carpenter now projects reinsurance industry return on equity falling from 18.9% in 2025 to 16.6% in 2026, 15.6% in 2027, and 15.3% in 2028, even as the sector's cost of equity climbs to 10.5% (Guy Carpenter, June 2026). The spread between return and cost of capital narrows every year through the forecast horizon, and that is a multi-year margin compression call, not a single soft-renewal headline.

The ROE Curve: From a 21.9% Peak to a Projected 15.3% Floor

The reinsurance sector's return trajectory has now moved through a full hard-to-soft arc inside five years. Guy Carpenter's own series puts industry ROE at 21.9% in 2023, the peak year of the post-Hurricane Ian repricing, before easing to 16.4% in 2024 and then rebounding to 18.9% in 2025 on a benign loss year (Guy Carpenter, June 2026). The 2026 through 2028 figures are the new part of the release: 16.6% for 2026, 15.6% for 2027, and 15.3% for 2028. Cost of equity, meanwhile, has moved in the opposite direction of what a simple mean-reversion story would predict, rising to 10.5% in 2026 from 9.7% in 2025, 10.2% in 2024, and 10.6% in 2023.

Year Reinsurer ROE Cost of Equity Implied Spread
2023 21.9% 10.6% 1,130 bps
2024 16.4% 10.2% 620 bps
2025 18.9% 9.7% 920 bps
2026E 16.6% 10.5% 610 bps
2027E 15.6% n.d. <610 bps
2028E 15.3% n.d. <610 bps

Guy Carpenter has not disclosed a specific cost-of-equity assumption for 2027 or 2028 in this release, which is itself informative: the broker is confident enough in the ROE decline to publish it two years out, but is not yet willing to commit to where the discount rate lands beyond next year. If cost of equity holds anywhere near the 2026 level of 10.5%, the 2027 and 2028 spreads compress further from the already-reduced 610 basis points logged for 2026, continuing a run that has cut the spread by more than half from the 2023 peak of 1,130 basis points in just three years.

Reading the Shrinking Spread

The spread between ROE and cost of equity is the number both cedants and reinsurers are implicitly pricing against at every renewal, because it is the economic value the sector is generating above what shareholders require to hold the capital. At 1,130 basis points in 2023, reinsurers could absorb a meaningfully worse loss year, a further round of rate cuts, or an expense creep and still clear their cost of capital with room to spare. At a projected 610 basis points for 2026, and likely tighter for 2027 and 2028 if cost of equity stays near 10.5%, that cushion has roughly halved. This is not a distress signal. Every year on Guy Carpenter's curve through 2028 still shows ROE comfortably above cost of equity, meaning the sector remains a value-creating one for its shareholders on Guy Carpenter's own numbers. But the buffer against a bad outcome, whether that is a major catastrophe year, a further leg of pricing softening, or a casualty reserve shock, is thinner than at any point since the 2023 repricing began.

The mechanism matters for actuaries pricing treaty layers or setting risk loads for 2027 renewals. A wide ROE-to-cost-of-equity spread gives a reinsurer's pricing actuary room to write business at a technical price below the theoretical fully-loaded rate and still clear the corporate hurdle, because the portfolio average is running well above target. A 610 basis-point spread removes most of that discretion. Individual treaties increasingly have to clear the hurdle on their own economics rather than riding a portfolio-level cushion, which is exactly the dynamic that shows up as selective underwriting discipline on some lines even as headline capacity stays abundant.

Two Trackers, One Capital Story

Guy Carpenter's own capital estimate puts total reinsurer capital at $663 billion for year-end 2025, up 9% over the year and split between $540 billion of traditional capital and $123 billion of alternative capital (Guy Carpenter, June 2026). Aon's midyear renewal report, using a broader scope and a later measurement date, puts global reinsurance capital at a record $790 billion as of March 31, 2026, with third-party and ILS capital rising $5 billion to a new high of $141 billion within that total (Aon, July 2026). The two figures are not directly comparable, since they draw on different capital definitions and measurement windows three months apart, but the direction is identical: both trackers show capital growing faster than the industry can profitably deploy it, and both show alternative capital as the fastest-growing component.

That is the same story told from two sides. The ROE decline is not primarily a story about worsening loss experience. It is a story about a capital base that has grown enough to compete pricing down toward the industry's aggregate cost of capital, even with underwriting fundamentals intact. Global reinsurance demand rose more than 10% at the midyear renewals (Aon, July 2026), and capacity was still described as plentiful and more than adequate to meet that increased demand. When both supply and demand are growing but supply is growing faster, the clearing price falls, and the clearing price for reinsurance capital is exactly what the ROE-to-cost-of-equity spread measures.

George Attard, Aon's Chief Strategy Officer for Reinsurance, framed the current environment as offering cedants a rare opening: a stable, well-capitalized and competitive reinsurance market provides an opportunity for better capital alignment between primary carriers and reinsurers (Aon, July 2026). Steve Hofmann, Aon's Americas CEO for Reinsurance, put the corollary on the reinsurer side of the table: cycle management is becoming an increasingly important strategic priority for balancing pricing discipline with growth (Aon, July 2026). Those two statements describe the same 610 basis-point spread from opposite sides of the negotiating table.

What a Still-Profitable but Declining Curve Means for Capital Deployment

A reinsurer board looking at Guy Carpenter's curve has three broad levers, and the industry is visibly pulling all three at once rather than choosing one. The first is continued softening: writing more premium at lower rates to defend market share while the spread is still comfortably positive, which is the path implied by the double-digit property cat rate cuts logged through July 1. The second is capital return to shareholders, buybacks and special dividends that shrink the equity base against which ROE is measured, a lever several publicly traded reinsurers have already pulled through 2025 and into 2026 as underwriting capacity outran deployable opportunity. The third is selective line withdrawal, holding capital but declining to deploy it into segments where the risk-adjusted return no longer clears the hurdle even at the portfolio level.

Swiss Re's mid-2026 renewal is the clearest data point for the third lever. The reinsurer posted $2.6 billion in H1 2026 net income and a 23.0% return on equity, comfortably above even Guy Carpenter's peak 2023 industry figure, while its June and July P&C Re treaty renewals generated $4.5 billion in premium volume, a 5.9% decline from the business up for renewal (Swiss Re, July 2026). Swiss Re attributed the drop to continued restructuring of liability lines, not to pricing pressure or a capacity constraint. A reinsurer posting a 23.0% ROE while voluntarily cutting casualty volume is not responding to a capital shortage. It is signaling that specific casualty cohorts, particularly US general liability and umbrella books carrying loss development assumptions that have not caught up with social inflation and nuclear verdict severity, no longer clear its return threshold even inside a portfolio still running well above the industry's projected 16.6% for 2026. That is what selective withdrawal looks like inside an aggregate ROE curve that is still positive: the industry-level number stays comfortably above cost of equity because the capital that exits underpriced casualty lines is redeployed into property cat capacity that is still clearing the hurdle, at least for now.

The Cedant Side: ROL Down to Minus 16% and the 2027 Buying Calculus

Guy Carpenter's global property catastrophe rate-on-line index fell 16% through the July 1, 2026 renewal, deepening from a 12% decline logged at January 1, 2026 (Guy Carpenter, July 2026). That acceleration through the year, rather than a leveling off, is the renewal-market mirror of the ROE curve: capacity kept arriving faster than demand could absorb it, and pricing kept giving ground at every subsequent renewal date rather than stabilizing after the January reset. Total insured catastrophe losses for H1 2026 are running around $35 billion, below the five-year inflation-adjusted average (Guy Carpenter, June 2026), which removed the loss-experience argument reinsurers might otherwise have used to hold the line on rate at the midyear dates.

For primary cedants building 2027 programs, the buying calculus this trajectory implies is straightforward but consequential. A reinsurance market pricing toward a compressing but still-positive spread, rather than toward distress, is a market where buyers can reasonably expect continued rate relief at January 2027 rather than a snap-back, but it is also a market where that relief has a floor somewhere near the point at which the ROE curve crosses cost of equity for enough of the market to trigger capacity withdrawal. Cedants layering in lower retentions to capture cheaper cat limit at today's pricing are making a bet that the 610 basis-point spread projected for 2026 does not compress to zero before their next renewal locks in the benefit. Given that Guy Carpenter's own curve shows the spread narrowing in both 2027 and 2028, that bet has a shorter runway than the still-double-digit rate cuts at July 1 might suggest.

When Does the Cycle Turn?

Prior soft-to-hard transitions offer a rough calibration for where this one stops. The 2017-2019 cycle turned only after a run of major catastrophe years, Harvey, Irma, Maria in 2017 and a heavy 2018-2019 loss period, pushed realized returns meaningfully below cost of capital for enough of the market to force capacity out; pricing did not turn on capital discipline alone while losses stayed benign. The 2023 hard market, by contrast, was triggered less by aggregate loss totals than by a retrocession and ILS capital contraction following 2022's inflation shock and Hurricane Ian, which shows that a capital-side shock can turn the cycle even without a correspondingly extreme loss year. Guy Carpenter's own forecast implies neither trigger is currently in view: H1 2026 losses are below trend and both Guy Carpenter's and Aon's capital trackers show growth, not contraction, in the base that would need to shrink to force a hardening.

What the 610 basis-point 2026 spread and the narrowing path into 2027 and 2028 suggest is that this cycle turns, if it turns without an exogenous shock, only when the spread compresses toward the low hundreds of basis points across enough of the market simultaneously, roughly the zone reached in 2024's 620 basis-point year before the 2025 rebound. The 2024-to-2025 pattern is itself a caution against assuming a straight-line decline: ROE rose from 16.4% to 18.9% in a single year on loss experience alone, even as cost of equity fell. A single elevated catastrophe half, an Atlantic hurricane season that behaves less benignly than H1 2026's convective-storm-dominated loss profile, or a casualty reserve shock large enough to force multiple reinsurers into Swiss Re-style restructuring simultaneously, could each independently interrupt the smooth 2026-to-2028 glide path Guy Carpenter has now put on record. Rating agencies and reinsurer boards building 2027 and 2028 capital plans off this curve should treat it as a base case conditioned on continued benign loss experience and continued capital growth, not as a forecast immune to either input reversing.

Why This Matters for Actuaries

The practical implication for pricing and portfolio actuaries is that the industry-level ROE curve is now explicit enough to underwrite against directly. A treaty priced to clear a 15% return in 2026 is priced with almost no margin above Guy Carpenter's own projected industry average, let alone above the tighter number implied for 2027 and 2028 if cost of equity stays near 10.5%. Reserving actuaries watching casualty lines should read Swiss Re's 5.9% volume cut alongside its 23.0% ROE as confirmation that capital is already discriminating within lines, not just across the property-casualty divide, rewarding cedants whose loss development assumptions have caught up with social inflation while withdrawing from those that have not. Capital modeling teams supporting 2027 and 2028 planning have, for the first time in this cycle, an explicit multi-year return curve from a major broker to stress-test against rather than having to extrapolate one from renewal-by-renewal rate commentary alone.

Further Reading

Sources

  1. Guy Carpenter, “Guy Carpenter Expects Reinsurance Industry to Generate ROE of 16.6% in 2026,” Reinsurance News, June 2026
  2. Guy Carpenter, “Global Property Cat Rates Down 16% as Softening Extends into July Renewals,” Reinsurance News, July 2026
  3. Guy Carpenter, “Global Property Cat ROL Down 12% at Jan 1 Reinsurance Renewals,” Reinsurance News, January 2026
  4. Aon, “Record $790bn Reinsurance Capital Underpins Softer Mid-Year Renewals,” Reinsurance News, July 2026
  5. Aon, Reinsurance Market Dynamics: Midyear 2026 Renewal Report, Aon, July 2026
  6. Swiss Re, H1 2026 Results and P&C Re Treaty Renewals, finews, July 2026