How Rate Increases Hid The Cost Problem For Three Years
Marcin Nowak, board member at Decerto, has 20+ years in insurance, focusing on automation, technology impact and software solutions.
gettyFor most of the past decade, a commercial carrier could carry an inefficient operating model because the top line grew faster than the drag. That arithmetic broke this year. According to The Council of Insurance Agents & Brokers’ “Q2 2026 P&C Market Survey,” average premiums fell 2% across all account sizes, with commercial property down 6.3%. It was the second consecutive quarterly decline, following a Q1 drop that was the first across all account sizes since 2017.
Rate is no longer doing the work. What remains is the expense ratio, and in most carriers, that number is a product of architecture, not headcount.
Three years of compounding rate increases did something subtle to carrier operations. They made cost discipline optional. When written premium grows 8% and the expense base grows 4%, the ratio improves without anyone touching a process. Executives often read that as improving efficiency. Most of the time, it was arithmetic.
AM Best now “expects lower net premiums written growth in 2026 and tighter margins across the P/C industry in 2026,” according to a press release from February 2026. This is driven by rising repair and materials costs alongside declining rate levels across several commercial lines. Both halves of the combined ratio are moving the wrong way at the same time.
That leaves one line item a carrier still controls. Loss cost is set by inflation, litigation and weather. Rate is set by the market and the regulator. The expense ratio is set by decisions the carrier made about its own systems, some of them 20 years ago.
In my experience, the carriers caught out here are not the ones with the worst technology. They are the ones that never had to look because rate covered for them.
The standard reflex many business leaders default to in a softening market is a hiring freeze and a 10% expense reduction target. I disagree with that playbook because it tends to fail in a predictable way.
The cost inside a mid-tier carrier’s operation is rarely the people. It is what the people are required to do: rekeying submission data that already arrived in ACORD format, reconciling reinsurance cessions in spreadsheets because the policy system and the ceded system disagree and hand-checking state rule variants before a filing goes out. Remove 10% of the staff, and none of that work disappears. Cycle times stretch, service metrics slip and the expense ratio improves for two quarters before rework pushes it back.
I’d recommend a different sequence. Before setting an expense target, measure how many times a single mid-term endorsement is touched by a human between agent submission and general ledger. The answer is, in my experience, usually somewhere between four and seven. Each of those touches is a salary line that a cost program will try to cut and an architecture decision that put it there in the first place.
I worked with a regional commercial carrier whose endorsement processing ran at 11 days and three internal handoffs. Nothing about the work was complex. The policy system could not accept a mid-term change without a manual rerate, so every endorsement queued behind an underwriting assistant. Removing the rerate step, not the assistant, brought the cycle under two days.
Boston Consulting Group’s “2026 Insurance Value Creators Report” explains why this matters now: “From 2021 through 2025, top-quartile P&C companies generated an estimated underwriting [return on tangible equity] contribution of approximately 11 percentage points, while bottom-quartile reported negative underwriting contributions.” In a hard market, that spread is a footnote. In a softening one, it is the whole story.
I don’t have visibility into every carrier that has tried this and failed. What is consistent is that the fix is bounded. Replacing a policy administration system takes 24 to 36 months. Removing the three worst handoffs in a single line of business takes two or three quarters, and it shows up in the same fiscal year.
There are four things I would urge business leaders to address now, and they’re all achievable inside two quarters.
1. Instrument the process before cutting it. Cycle time and touch count per transaction type should be measured rather than estimated. Most carriers cannot produce those numbers today.
2. Pick one line of business, not the enterprise. Enterprise-wide efficiency programs produce steering committees. A single line produces a result.
3. Externalize what changes most often. Rating factors, state variants and endorsement rules belong in configuration a business analyst can edit, not in code sitting behind a release window.
4. Fix the handoff, not the headcount. If a step exists because two systems cannot agree on a record, the integration is the cost, and the person reconciling it is only the symptom.
Deloitte’s “2026 Global Insurance Outlook“ argues that insurers who “act decisively to reconsider their current business models, products, tools, and stakeholder interactions” are the ones positioned to do well. In a soft market, acting decisively means starting before the pressure shows up in quarterly results.
Soft markets end, and this one will, too. But it will outlast the 18 to 24 months it takes to pull structural cost out of an operating model. Carriers that wait until the damage is visible in their numbers will finish the work sometime in the next hard market. But by that time, nobody will care.
Here’s what you need to start asking: If rate went flat across your entire book tomorrow, what is the first cost you would take out, and how long would it take? If the honest answer is people and the honest timeline is next quarter, that cost will come back. The expense ratio a carrier can defend through a soft market is the one that was designed, not the one that was trimmed.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
