Two numbers that do not agree
Aon's Global Medical Trend Rates report projects India's medical trend rate at 11.5% for 2026, down from the 13% it projected for 2025. That sits against an Asia Pacific average of 11.3% and a global average of 9.8%. Trend rate is the expected year-on-year increase in the per-person cost of delivering the same medical benefit, driven by provider pricing and utilisation. It is the industry's published answer to the question every CFO asks at renewal: how much more should the same cover cost next year?
The number an Indian employer actually sees on the renewal quote is different. Truworth Wellness's 2026 analysis puts the group mediclaim premium increases Indian HR and benefits managers are facing at 20 to 35% year on year. Mercer Marsh Benefits' Health Trends 2026 describes the same pressure globally, with employers absorbing persistent double-digit medical trend increases alongside coverage gaps.
So the gap between the published benchmark and the quote in front of you is roughly 9 to 24 percentage points. That gap is not medical inflation. Medical inflation is already inside the 11.5%. The rest is a mix of three other things, and each one has a different negotiation answer.
Decompose the loading before you argue about it
A CFO cannot negotiate a single 25% number. A CFO can negotiate four numbers. Break the loading into components and force each one to be defended separately.
- Defensible trend. The industry benchmark for the coming policy year. For India in 2026, 11.5% on a like-for-like benefit and a like-for-like population. This part is not worth fighting; it is worth verifying that it has not been applied twice.
- Burning cost on your own experience. The rate implied by your own incurred claims over the expiring period, projected forward. This is where a bad year genuinely costs you, and it is the component you can influence over the next twelve months.
- Demographic and design drift. Headcount growth, an ageing or shifting employee mix, more dependants enrolled, a mid-term sum insured uplift, a maternity sub-limit relaxed at last renewal. Each of these changes the risk, and each should be priced explicitly rather than folded into a headline percentage.
- Portfolio correction. The insurer's own book repair, applied across accounts irrespective of individual performance. This is the component you should refuse to fund, and the one an insurer will almost never itemise unless asked.
The question that splits the quote
Ask the underwriter, in writing: what incurred claims ratio have you assumed for this account over the expiring period, at what date of valuation, and what loss ratio are you targeting for the renewal year? Two numbers and a date. If the answer to the first is materially different from your own claim register, you have a data dispute to settle before you have a price dispute. If the target loss ratio is far below the level at which insurers normally act, you are being asked to pay for margin recovery rather than for your own risk.
The 80 to 85% threshold and what sits on either side of it
Truworth Wellness's 2026 analysis identifies the trigger point plainly: insurers hike premiums when an enterprise's group mediclaim loss ratio crosses the 80 to 85% benchmark. That threshold is the single most useful fact in a renewal conversation, because it converts a vague argument about fairness into an arithmetic one.
If your incurred claims ratio for the expiring year is, say, 74%, you are inside the band where the account is working for the insurer. A 25% loading on that account is not being driven by your experience. It is being driven by the book, and the correct response is to test the market rather than to accept the number.
If your ratio is 96%, the conversation changes shape entirely. A loading is coming, and the useful work is not resisting it but structuring it: how much of the correction lands this year versus over two years, what plan design changes offset it, and what the insurer will commit to if the corrective measures work.
Normalise the ratio before you quote it back
Two adjustments matter and both usually move the number in your favour:
- Valuation date and IBNR. An incurred claims ratio quoted at month nine of a twelve-month policy is incomplete. Insurers project the remainder, and the projection assumption is negotiable. Ask what has been loaded for claims incurred but not reported, and on what basis.
- One-off large claims. A single catastrophic hospitalisation on a 400-life group can add fifteen points to a ratio and tells you nothing about the underlying risk. Present the ratio with and without the outlier, and argue the loading against the trend line rather than the spike.
The distinction between an account that is structurally loss-making and an account that had one bad quarter is the entire argument. If your data cannot make that distinction, fix the data before the next renewal cycle, not during this one. The mechanics of getting that data clean sit in group mediclaim administration and the CD account.
The market context the underwriter is not volunteering
The insurer's pricing posture in 2026 is not a secret, and knowing it changes how you read the quote. A Kotak Institutional Equities note reported by ANI and The Tribune on 20 August 2026 put health insurance premium growth at 26% year on year in July 2026, with retail health leading at 31% growth.
Read that against the corporate group number. Retail health is growing faster than the book overall, which tells you where insurers are winning rate most easily. Group business is the part of the health portfolio insurers have historically underpriced to hold volume, and periods of strong retail growth are exactly when they feel able to correct group pricing without fearing the loss of the account.
Two practical consequences follow. First, this is a year in which walking an account to market is more likely to produce a genuine alternative quote than a token one, because the correction is being applied unevenly across insurers and their appetites are diverging. Second, an insurer with a strong retail engine has less need to hold a loss-making group account, so a poor-performing programme should expect a harder line than in previous cycles. The broader hardening in corporate group health is the backdrop against which every individual quote is being written.
Four plan-design levers that move the number
The levers that reduce premium fall into two groups: the ones employees notice on the day they use the policy, and the ones they do not. The second group is where the work should start.
- Room rent capping tied to sum insured. Uncapped or generously capped room eligibility drives the entire hospitalisation bill, because Indian hospital pricing for procedures, consultant fees and consumables often scales with the room category. Capping room rent as a percentage of sum insured, rather than as a flat rupee figure, keeps the cap current as the sum insured rises and is close to invisible to employees who do not elect a premium room. This is usually the highest-yield single change.
- A corporate buffer sized to actual usage. A large floating corporate buffer sits in the premium whether or not it is used. Size it from the last two years of buffer draw-down rather than from the figure carried forward since the programme was set up. Employers frequently discover they are funding a buffer that is drawn on a handful of times a year.
- Graded sum insured by band rather than a flat figure. A uniform sum insured across all grades overprices the largest and youngest segment of the population. Grading it aligns the cover with the exposure and takes cost out at the point where claims are least likely.
- A structural deductible or co-pay placed where it changes behaviour, not where it hurts. A co-pay on the whole policy is felt by every claimant and is the change employees resent most. A co-pay confined to a specific cohort, such as parents-in-law, or a deductible confined to the top layer of a large claim, removes cost with a fraction of the friction.
The lever that works on next year's number
None of the four addresses the underlying claims cost. That is the point of preventive screening and early-detection programmes: they change the burning cost component of the loading rather than the benefit structure. The effect appears at the following renewal, not this one, which is why the case has to be made a year before the money is needed. The mechanics are set out in health screening participation and its effect on loss ratio, and the wider set of spend levers in reducing employee benefit spend without cutting cover.
What a defensible counter-proposal looks like
A counter-proposal that says "25% is too high" gets a small concession and a repeat next year. A counter-proposal that reconstructs the number gets a different conversation.
The structure that works is a single page carrying four lines:
- Trend component, at the published benchmark, on the expiring benefit and the expiring census.
- Experience component, derived from your normalised incurred claims ratio, with the outlier treatment stated.
- Design and demographic component, itemised by change, each priced.
- Residual, labelled as unexplained, with a request for the insurer to attribute it.
The residual line is the one that does the work. It is difficult for an underwriter to defend an unattributed loading in writing, and putting it on the page forces one of three outcomes: the insurer explains it, the insurer reduces it, or the insurer declines to do either and you have your justification for moving the account.
Pair the counter with what you are offering in return. A renewal negotiation in which the employer only asks is weaker than one in which the employer brings design changes, a screening commitment, or a multi-year structure. Insurers price uncertainty, and a credible plan to reduce next year's claims cost is a reason to take less rate now.
Governance: what the CFO should see every quarter
Most renewal disputes are lost in the eleven months before the renewal, when nobody is looking at the numbers. Group mediclaim is an annually renewable, experience-rated contract, which means the price is being set continuously by claims that are already happening.
A quarterly pack of four items is enough:
- Incurred claims ratio to date, with the incurred-but-not-reported assumption stated and the valuation date on the page.
- Claim count and average claim value, split by employee and dependant.
- The top ten claims by value, with a note on whether any are recurring or catastrophic.
- Buffer utilisation against the buffer funded.
By month nine, that pack tells you whether the account is heading through the 80 to 85% band and gives the finance function time to decide whether the answer is a design change, a market exercise, or budgeting for the loading. It also means the claims figures at the renewal meeting are yours, dated and reconciled, rather than the insurer's presented for the first time across the table.
Group mediclaim is one of the few lines where the buyer holds the loss data in real time and usually does not look at it until the price arrives. The employers who negotiate well in a hardening market are the ones who knew in November what the June renewal was going to say.