EARTHQUAKE CATASTROPHE INSURANCE PENETRATION IN INDIA: CHALLENGES AND OPPORTUNITIES

This study employs scenario-based seismic risk modeling to assess how earthquake insurance can serve as a practical risk-transfer tool in India, where catastrophe insurance penetration remains low despite high seismic exposure. Using OpenQuake, we model two historical events on today’s exposure: the 2001 Mw 7.7 Bhuj earthquake (Kachchh, Gujarat) and the 1897 Mw ~8 Assam earthquake (Guwahati region), combining scenario shaking, exposure (GEM), and vulnerability functions to estimate building losses.

We then run counterfactual “what-if” tests using simplified proxies of three established insurance structures (CEA, JER, EQC) across penetration levels of 2%, 25%, and 80% of the highest-value residential assets, focusing on how deductibles, limits, and event caps shape outcomes.

Implications:

• Meaningful risk transfer is possible even with low uptake: in Guwahati, insuring just 2% of high-value assets can transfer a large share of losses when event caps don’t bind (e.g., 40.4% of the total loss insured under a JER-like structure in one scenario test).

• Event caps can break the system when they are too low: when event limits are enforced, performance can collapse (e.g., a CEA-like event cap results in <1% of losses insured in the tested scenarios), highlighting the need to design caps that match regional tail risk.

• Design beats slogans: the effectiveness depends heavily on deductible/limit choices (high deductibles reduce frictionless recovery and can suppress uptake; low limits underperform in high-loss scenarios).

• Diminishing returns beyond moderate penetration: moving from 25% to 80% penetration yields relatively small additional insured loss in the tests, suggesting policy effort may be better spent on smart product design and targeted uptake rather than chasing universal participation immediately.

This study was presented at the 18th World Conference on Earthquake Engineering in Milan, Italy. A more comprehensive follow-up study is currently ongoing. Please reach out if you want to read the full paper.