Apartment Insurance Just Split Into Two Markets. Your Underwriting Model Doesn't Know Yet.

The Renewal That Moved In Two Directions At Once

The renewal quote came back lower than last year.

Not by a little. Not by a rounding error. The property side of the policy was down roughly 10% on the same building, in the same market, with the same carrier the operator had been working with for three years.

Then the broker turned to the second page.

The liability side was up more than 30%. The retention had been doubled. Two of the carriers on the previous renewal had quietly exited the apartment line altogether between the last renewal and this one. The broker had spent six weeks finding a market that would write the coverage at all.

Net of both lines, the renewal was up. But "up" is the wrong word for what actually happened. The two sides of the same policy moved in opposite directions, and the operator's pro forma had treated them as one number.

That is the single most expensive assumption error I see in multifamily underwriting right now.

TL;DR

Apartment insurance is not one market anymore. Property rates are softening in 2026 — Lockton's February 2026 market update has nonhabitational commercial real estate down 5–10% at renewal — while liability rates are hardening, retail carriers have exited the space, and operators with a single-line insurance assumption in their pro forma are underwriting the wrong renewal. This piece walks through the split, what changed, and what operators need to rebuild in their underwriting models before the summer renewal cycle.

I. The Narrative That's Now Half Wrong

Since 2022, the operator-facing story on insurance has been one sentence: insurance is killing multifamily NOI.

That sentence was true. It is no longer the whole truth.

Lockton's February 2026 market update has the cleanest read on the bifurcation. Property rates for nonhabitational commercial real estate are down 5–10% at renewal on single-insurer placements, with larger decreases on shared and layered structures — and Lockton expects further softening through the summer. The relief is concentrated in non-catastrophe markets and well-maintained portfolios, where well-documented multifamily can run toward the larger end of the range. The hard market on the property side has rolled over. Capacity is returning. Renewals are clearing at lower numbers than they did 18 months ago.

That part of the story is the good news.

The other part is the story the trade press is still figuring out how to cover, because it doesn't fit the headline.

Liability coverage on apartment buildings is hardening so sharply that many retail insurers have exited the multifamily space entirely, pushing coverage into the excess & surplus channel. The carriers that remain are pricing the risk at a different tier. Risk & Insurance, reporting on the same Lockton data, describes a bifurcated market where favorable property pricing contrasts sharply with mounting challenges in casualty coverage — and the pattern is consistent across the major brokers reporting from this renewal cycle.

Two markets. Same policy. Moving in opposite directions.

II. Why The Property Side Is Softening

The mechanic on the property side is straightforward.

Reinsurance pricing softened into the January 2026 renewals after two consecutive years of capacity tightening. Guy Carpenter's renewal report describes a market shaped by capital growth, one of the lowest reinsured catastrophe-loss years in a decade, and strong reinsurer returns — a 17.6% return on equity in 2025, with dedicated reinsurance capital up another ~9%. The catastrophe load on the books of the primary carriers came in below modeled expectations. Capital that had exited the property line during the 2022–2024 hard market started flowing back in.

For an apartment operator, the practical implication is that the property renewal that came in 20% higher last year is now coming in 5–10% lower, sometimes more on portfolios in stable markets without recent claims activity.

Three factors widen the relief inside that range.

Roof age and condition. Carriers are still actuarially conservative on aged roofs. Well-documented roof maintenance and replacement records are pulling renewal quotes toward the softer end of the range. Operators who have not invested in roof documentation are pulling toward the harder end.

Water mitigation systems. Leak-detection technology, shutoff systems, and documented loss-control programs are increasingly priced into the renewal. The differential is real — two otherwise identical buildings can be meaningfully apart on the property line based on water mitigation alone.

Claims history. Three years of clean claims data is pulling renewals significantly softer. Three years of any meaningful claim activity is pulling them harder, even where the broader market is softening.

The relief is also uneven by geography and asset quality. As PwC's Veronika Torarp told Multi-Housing News, even in California's high-risk environment some large insurers approved to raise rates ~7% are actually cutting them ~9% across the board, with Southern California counties seeing cuts near 15% — but the differentiator is construction type and whether capex has hardened the building, not just location. For operators who have the documentation, the property-side relief is real money. For operators sitting on a deferred-maintenance asset with a thin claims file, the softening is going to feel uneven.

III. Why The Liability Side Is Hardening

The liability story is the opposite. Three things are pushing apartment liability coverage into a structural hard market.

The exit of retail carriers from the multifamily line. This is the most operator-relevant of the three. Carriers that historically wrote apartment liability are pulling out of the segment — either because their loss ratios in the line have deteriorated or because their reinsurance partners have de-prioritized it — and much of the coverage has migrated into the excess & surplus channel. The pool of carriers willing to quote a multifamily liability renewal in 2026 is materially smaller than it was a few years ago.

Nuclear verdicts. Plaintiff verdicts in habitability, assault, and slip-and-fall cases in multifamily settings have been trending upward. Single judgments in the $20–50 million range have become reported regularly enough that they price into the underwriting on every apartment liability renewal. Carriers are pricing for a tail risk that did not exist at the same magnitude five years ago.

Social inflation and discovery costs. Even cases that settle below the headline verdict numbers are settling higher than historical norms. The cost of defending the cases that don't settle is also up. The underwriting math reflects both.

The operator-side reality of this hard market is not just a higher premium number. Industry reporting on the 2026 cycle points to liability renewals landing roughly 20–40% higher, on top of higher retentions and longer broker work times. Some operators are renewing with the same carrier at materially worse terms because the alternative is non-renewal. That is the practical shape of a hard market: a smaller set of carriers to choose from, higher retentions on the policies that do get written, and tighter coverage terms inside those policies.

IV. The Underwriting Model Error

Here is what I see in pro formas I review.

Most multifamily underwriting models have a single line item called "insurance" that gets year-over-year growth assumptions of 5–8% in the optimistic case, 10–15% in the conservative case.

That assumption is wrong on both ends.

It is wrong on the property side because the property line is softening, not growing. The operator who builds in 10% property growth is overstating expense in years one through three of the hold, which overstates exit value through suppressed in-place NOI assumptions. The error is small but real, and it is consistent.

It is wrong on the liability side because the liability line is hardening at a rate that 5–15% growth doesn't capture. The operator who builds in 8% liability growth in a market where renewals are coming in 25% higher is understating expense, overstating NOI, and overstating exit value. The error here is larger, and it compounds over the hold.

The two errors don't offset. They move in the same direction structurally — the property error overstates NOI and the liability error overstates NOI.

The fix is not hard. Split the insurance line into two. Run a different rate assumption on each. The property side gets a softening assumption through 2027, with hardening risk priced back in around lease three. The liability side gets a 15–25% growth assumption through the underwriting horizon, with retention and coverage-term degradation priced in separately.

That single change closes the assumption gap most operators are carrying.

V. What Disciplined Operators Are Building Into Their Renewal Strategy

Operators who are reading the bifurcation correctly are doing four things going into the summer renewal cycle.

Starting the liability renewal eight to twelve weeks early. The broker timeline on liability is long, and the carriers willing to quote are working through a backlog. Starting late produces non-renewals, not better terms.

Building the loss-control file before the renewal, not during it. Maintenance documentation, water-mitigation records, security protocols, tenant-screening procedures. Every documented piece of loss control moves the property quote toward the softer end of the range and moves the liability quote into the carrier's "consider" pile rather than the "decline" pile.

Stress-testing retention levels. The retentions on the policies getting written in 2026 are higher than the ones written a few years ago. Operators are sizing their reserves to absorb the new retention without breaking the asset's cash-flow profile.

Reading the regional bifurcation. Florida-specific relief is real, particularly after the state's insurance reforms that are beginning to reshape the market. Texas and California are running their own dynamics. The "average" national number hides material regional variance, and the operators sourcing capacity through brokers with deep regional carrier relationships are getting different renewals than operators on the standard national rotation.

None of this is heroic. It is the work the hard market on liability requires and the work the soft market on property rewards.

VI. The Read I Keep Coming Back To

The bifurcation in apartment insurance is not a temporary divergence. It is a structural sorting of two different risks that the underwriting market is finally pricing separately.

Property risk on apartments is a function of physical asset characteristics. It is modelable, diversifiable, and reinsurable. The market for it works the way insurance markets are supposed to work.

Liability risk on apartments is a function of legal environment, social inflation, and verdict trends. It is harder to model, less diversifiable, and is sitting at a different structural place in the cycle than property risk.

The two sides of the same policy are moving on different clocks. They will probably continue to move on different clocks for as long as the underlying drivers stay in place — which is to say for at least the next two renewal cycles.

The operator who underwrites this market with one insurance assumption is mispricing both sides at once. The operator who splits the line, runs the right growth assumption on each, builds the loss-control file, and starts the liability renewal early is going to find that the 2026 insurance environment is a much more manageable expense story than the 2022–2024 environment was.

The narrative that insurance is killing multifamily NOI is not the operative narrative anymore. The operative narrative is that apartment insurance is splitting in half, and the operators reading the split correctly are quietly recovering some of the NOI the last cycle took from them. The ones still underwriting on the 2024 assumption are leaving that recovery on the table.

If you're heading into a summer renewal cycle and want to talk through how to rebuild the insurance assumption in your underwriting model, I keep 30 minutes a week open for conversations like that. No pitch. Just the read. Book a meeting with me here.

— Andrew