Individual indemnity: uncapped room rent, provider concentration and a creeping claim ratio
An individual indemnity product sold through tied agency has two sum-insured tiers, ₹5 lakh and ₹10 lakh, with no room-rent cap and no co-payment; the cap was removed two years ago to match competitors. Portfolio earned premium is ₹180 crore and the incurred claim ratio has risen from 71% to 94% over three years. Average claim size grew 19% last year against an estimated market trend of 11%. Analysis shows that 12% of the hospitals used generate 40% of claim cost, and their average claim is 2.3 times the portfolio average. In those hospitals, room rent, investigations and pharmacy are billed in proportion to the room category chosen. Agents say any cap will be used against them, and two competitors have launched uncapped ₹10 lakh products. The insurer has 30% quota-share reinsurance running to 31 March. Pre-authorisation audit capacity is limited.
1. Estimate how much of the 19% claim-size growth is likely to be provider-induced rather than genuine trend, and set out how you would test that hypothesis.
Read the worked answer
Estimate. 40% of claim cost comes from 12% of hospitals at 2.3 times average severity. If room rent is 25-30% of a hospital bill and investigations and pharmacy are billed in proportion to room class, the excess is roughly 0.40 x 0.28 x (1 - 1/2.3) = 6% of total cost, rising to about 10-12% once proportional ancillary billing is included. So of the 19% claim-size growth, about 11 points is market trend and 7-8 points is plausibly design and provider effect.
Tests, before acting on it.
- Compare year-on-year severity for the same hospital and the same procedure code. Genuine trend affects all providers; design effect clusters in the uncapped ones.
- Look at the distribution of room categories actually chosen within each sum-insured tier. A shift upward within a tier is moral hazard; a shift to higher tiers is anti-selection.
- Check for claim splitting and unbundling: multiple short admissions, day-care conversion, separate billing of implants.
- Check whether the 12 hospitals' case mix explains their severity; a tertiary referral centre legitimately costs more.
- Run a difference-in-differences against a peer portfolio that retained a cap. This is the cleanest test and the one a reinsurer will ask for.
Conclusion. The evidence is consistent with 6-8 points of design-driven inflation, but it is a hypothesis until the same-procedure comparison confirms it.
2. Recommend a redesign, quantify its expected effect and address anti-selection and distribution risk.
Read the worked answer
Recommendation. Reintroduce a room-rent cap at 1% of sum insured per day (₹5,000) and ICU at 2% on the ₹5 lakh tier only; leave the ₹10 lakh tier uncapped but reprice it +8% to reflect the provider intensity it attracts. Do not add a co-payment: it applies to every claim, is highly visible and would damage a segment where the insurer has no pricing power.
Quantification. Expected saving on the ₹5 lakh tier is 8-10% of claim cost. Assume 5-8% of members upgrade or lapse in response, adding 2-3 points of average cost back. Net effect 2-5%, and the anti-selection cost bites hardest in a portfolio that is 40% agency-sold, because agents will simply place the upgraders elsewhere. That is why the cap cannot substitute for a reprice: file +5-6% on both tiers at the same renewal, showing the cap as the reason the increase is smaller than trend.
Distribution. Brief agents before filing with a worked example showing that a member with a ₹1.5 lakh bill is better off under the capped design at the new premium than under an uncapped product at a competitor's re-underwritten price. Two competitors' uncapped ₹10 lakh products are a real risk; the honest answer is that the insurer cannot win an uncapped price war with a 94% ratio.
Addressable cost. Negotiate preferred-provider terms with the 12 outlier hospitals: 40% of cost concentrated in 12% of hospitals is the one supplier intervention with a fast payback.
Reinsurance. The 30% quota share halves the benefit of the saving, but the reinsurer is likely to make the redesign a condition of renewal, so the true counterfactual is not 'keep the current design'.