I. What a motor insurance policy actually is
Before any company or combined ratio, consider the transaction itself. A vehicle owner hands over a defined sum today in exchange for a contingent promise: that if a specified event occurs, the insurer will pay a sum unknown at the moment of contracting. What the insurer sells is not “protection” in any tangible sense. It sells the transfer of variance.
The policyholder converts an uncertain, potentially ruinous, low-probability outcome into a certain, small, budgetable cost, and the insurer accepts that variance onto its own balance sheet, holding capital against the distribution of outcomes it has assumed.
The insurer’s product is, precisely, its willingness and capacity to hold capital against a probability distribution and to be paid for doing so.
Economic value in this business is made in two distinct places, and conflating them is the single most common analytical error made about insurance.
The first is the underwriting margin: premium received minus claims paid minus the acquisition cost (commissions, dealer payouts) and administration cost of running the book. The second is the investment return earned on money the insurer holds in the interval between receiving a premium and paying a claim.
These are economically separate activities. The first is an insurance-manufacturing business whose quality is measured by the combined ratio. The second is an asset-management business run on borrowed money, the “float”, whose quality is measured by investment yield and the duration and cost of the liabilities funding it.
A firm can be excellent at one and poor at the other. Treating the reported bottom line as a single undifferentiated profit stream obscures whether a company is a disciplined underwriter cross-subsidized by nothing, or a mediocre underwriter rescued each year by investment income and realized capital gains.
Trace one comprehensive private-car policy from inception to settlement. On day one the insurer collects a premium composed of an own-damage component and a statutory third-party component, and immediately pays out a large slice, frequently a quarter to a half of a new-vehicle motor premium, as commission or dealer payout.
It books the remainder as an unearned premium reserve and earns it into income across the policy year. If the car is dented in month four, an own-damage claim is surveyed, the garage is paid, and the matter closes within weeks — a short-tail claim.
If instead the car kills or maims a pedestrian, a third-party claim is filed at a Motor Accident Claims Tribunal, contested over the victim’s income and dependency, and settled years later at an amount a judge determines. In that interval the insurer holds the premium, invests it, and earns a yield.
The entire economic question of motor insurance is whether the yield earned on money held in that interval, plus or minus the underwriting result, exceeds the cost of the capital the insurer must post to be allowed to make the promise at all.
II. The two products bundled inside one policy
Indian motor insurance is not one business; it is two businesses with almost opposite economics sold under one document. Third-party liability cover is a statutory instrument created under Chapter XI (and Section 146) of the Motor Vehicles Act, 1988.
It is compulsory, administratively priced, and carries effectively unlimited liability for death and bodily injury, because Indian law places no ceiling on the compensation a tribunal may award for a life. It is settled through Motor Accident Claims Tribunals over multi-year horizons and functions, in substance, as a social compensation mechanism that happens to be delivered through commercial balance sheets rather than through the exchequer.
Own-damage cover, by contrast, is a freely priced, short-tail property indemnity on a depreciating asset. It is high in frequency, moderate in severity, closes quickly, and is subject to open price competition and heavy discounting.
The mechanics diverge completely. A third-party claim arises when a vehicle injures or kills a person or damages property; the victim (or dependants) petitions a tribunal, which determines compensation by applying a judicial multiplier to the deceased’s or injured party’s assessed income, with an addition for “future prospects” (50%/30%/15% of income for salaried persons in the under-40, 40-50, and 50-60 age brackets respectively; 40%/25%/10% for the self-employed), a deduction for personal expenses, and standardised sums for conventional heads such as loss of estate, funeral expenses and consortium, escalated 10% every three years.
The income is itself contested, and the multiplier depends on the age of the deceased. The result is an award that is large, inflation-linked by judicial doctrine, and known with certainty only years after the accident.
An own-damage claim, by contrast, is surveyed by an assessor, subjected to part-by-part depreciation (roughly 50% on plastic panels, 30% on metal, 25% on rubber unless a zero-depreciation add-on is bought), reduced by any no-claim bonus forfeiture logic, and paid against a repair estimate or on a total-loss basis capped at the insured declared value.
The bundling is the crux. A commercial insurer cannot decline the socially priced product and keep only the commercial one. Comprehensive policies fuse both covers; standalone own-damage requires a valid third-party policy to exist; and for new vehicles the law mandates long-term third-party cover.
The regulatory architecture thus forces a commercial insurer to sell a loss-making, politically priced social product as the condition of accessing the freely priced, potentially profitable one, and layers on top the Motor Third-Party Obligation regulations, which compel every insurer to write a minimum quantum of this business. The commercial line is effectively held hostage to the social one.
III. Unit economics and the float
The industry’s headline metric is the combined ratio on net earned premium, and it is insufficient on its own. It compresses two segments with order-of-magnitude different economics into one number and says nothing about the float that partly redeems the loss.
The correct unit is per vehicle-year, split by cover and by vehicle class. The industry’s FY26 motor combined ratio was approximately 128.0% on a financial-year IGAAP basis (deteriorating from 123.7% in FY25; the nine-month figure was 128.1% for 9MFY26 against 123.8% a year earlier). That number is a blend.
Within it, own-damage runs far closer to breakeven, the best private writers report own-damage loss ratios in the high-60s to low-70s percent, while third-party loss ratios for the weaker parts of the industry, especially public-sector and commercial-vehicle pools, have historically run at loss ratios well above 100% and at times above 200%.
The dispersion is enormous: ICICI Lombard’s own motor loss ratios illustrate the well-run end, 9MFY24 motor third-party at 64.6% and own-damage at 64.9%, and in Q1FY27 own-damage at 67.6% and third-party at 70.6%, against company guidance of a 65-67% overall motor loss ratio, whereas the industry aggregate combined ratio of 128% reveals how much worse the poorly-run tail must be.
Two-wheelers, private cars and commercial vehicles differ starkly: commercial-vehicle third-party is the epicentre of loss, private car is the best mix, and two-wheelers combine low premium with poor loss economics and near-total non-compliance.
Now construct the float. The insurer holds two pools of policyholder money: the unearned premium reserve (short) and the reserve for outstanding claims including incurred-but-not-reported liabilities (short for own-damage, very long for third-party, whose claims can settle five to ten years after the accident). The whole of general insurance in India sits on a large investment book relative to its net worth, ICICI Lombard alone carried investment assets of ₹584.21 billion as of March 2026 against an investment leverage (net of borrowings) of 3.47x, earning a realized investment yield including capital gains that has run around 8-9%.
This is the annuity inside the business: management borrows from policyholders at a “cost” equal to the underwriting loss and invests the proceeds at the bond-plus-equity yield.
The equation that decides whether motor is worth writing is therefore not “combined ratio below 100.” It is whether the underwriting loss, expressed as a percentage of the float generated, is smaller than the investment yield on that float.
A combined ratio of 110% on a line that generates float equal to one year of premium, at an 8% yield, is roughly value-neutral before the cost of equity capital; the same 110% on a short-tail line that generates only a few months of float is value-destroying.
Third-party, being long-tail, generates far more float per rupee of premium than own-damage and can therefore tolerate a higher combined ratio, which is precisely the economic argument insurers use to keep writing it. But at an industry motor combined ratio of 128%, the underwriting loss is roughly 28 points of premium.
Even generous float assumptions, third-party float of perhaps two to three years of premium at an 8% yield, recover only a portion of a 28-point loss once the short-tail own-damage half of the book (which generates little float) is blended in.
The industry as a whole is writing motor below its economic breakeven.
Finally, the accident-year versus financial-year distinction matters more here than anywhere else in Indian general insurance: a financial-year combined ratio flatters or punishes the current year by the release or strengthening of reserves struck for prior accident years, and in a long-tail line like third-party those prior-year movements can dominate the reported number. A financial-year motor combined ratio is only as honest as the reserve estimates embedded in it.
IV. Reserving, reserve releases, and whether reported profit is real
In the long-tail half of motor insurance every rupee of reported profit is a management estimate until a tribunal converts it, years later, into a cash fact.
A third-party claim is incurred today, reported later, litigated for years, and settled at a figure a judge determines by applying a multiplier to a contested income.
The insurer must therefore hold reserves — case reserves for reported claims and IBNR/IBNER reserves for claims incurred but not yet reported or not yet fully developed — and the adequacy of those reserves is simultaneously the largest liability on the balance sheet and the least verifiable number in the accounts. Book value in this business is itself an estimate.
Cholamandalam MS created an additional inflation-linked motor third-party reserve of ₹150 crore in 9MFY26, pushing its combined ratio to 116.2%. HDFC Ergo’s loss-reserving triangle “showed an unfavourable development in FY2024 due to the change in the actuarial assumptions, factoring in a higher road mortality rate and an increase in accident frequency”, direct evidence of prior-year reserves proving inadequate against rising award trends.
And the historical precedent at the public insurers is stark: IRDAI once terminated the actuary membership of National Insurance’s appointed actuary over FY2015-16 under-reserving, where certified motor third-party IBNR of ₹3,030 crore compared with a “most likely” estimate of ₹7,293 crore — a shortfall of ₹4,263 crore in one segment in one year.
The recurring ICRA framing is that “the uncertainty regarding the extent of claims is relatively higher in the long-tail Motor-TP segment,” which “could result in uncertainty regarding the level of future claims in relation to the past reserves made for this segment.”
In plain terms: some private insurers have been releasing prior-year reserves into current profit on the strength of favourable development, while the structural risk is that award inflation and settlement acceleration make those releases premature.
On 11 June 2026, in Shishu Pal @ Shish Ram v. Surjeet & Ors. (2026 INSC 634), the Supreme Court created a new compensation head — “loss of domestic care” — valuing a homemaker’s unpaid work at a notional ₹30,000 per month, escalating 10% cumulatively every three years, and in that case enhanced an award from ₹8.43 lakh to ₹62.78 lakh.
ICICI Lombard’s management assessed that this single judgment would raise the industry’s motor third-party loss ratio “in the range of 12 to 15 percent” (CEO Sanjeev Mantri), took a ₹165 crore reserve charge on already-earned premium in Q1FY27 (which alone added 2.8 percentage points to its combined ratio and helped drive a 46% fall in quarterly profit to ₹403.17 crore, from ₹747.08 crore; even excluding this and two fire losses, adjusted profit still fell 23% to ₹575 crore), and the industry estimates the claim-outgo impact ramping from 3-4% initially to 10-12% “as courts across the country adopt the interpretation more widely over the next 12-18 months.”
The same judgment reportedly pressed for faster disposal of motor appeals (with the Court noting High Court appeal pendency averaging around eight years and encouraging a disposal cap), which accelerates the conversion of reserves into cash and compresses the interval over which the float is held.
The critical point for the analyst is that this is not a one-time charge against one insurer; it is a permanent re-basing of the multiplicand in the compensation formula, applied retroactively to every open accident year, landing on reserves that were struck before the standard existed.
The honest conclusion is reported non-life profit attributable to motor is a stack in which genuine underwriting profit is thin to negative, float/investment income is the largest true economic contributor, realised capital gains episodically flatter the bottom line, prior-year reserve releases have quietly supported some private insurers’ current earnings, and the entire edifice is about to be re-baselined twice in opposite directions — upward in reported profit by Ind AS 117 discounting, and downward in economic reality by the retroactive homemaker-judgment reserve strengthening and the acceleration of settlements.
An investor who takes the FY27 reported combined ratio at face value, on whichever basis it is presented, without asking how much is discounting and how much is reserve adequacy, will be systematically misled. The correct posture is to demand the pre-discounting, accident-year number and to treat any reserve release in a rising-award environment as suspect.
V. The master variable — the regulated tariff and its political economy
Third-party pricing is the one price in Indian general insurance still set administratively.
No third-party tariff revision has in fact been implemented since FY2019-20.
The operative rates remain those extended from FY2019-20, themselves carried over from FY2018-19 (₹2,072/₹3,221/₹7,890 for the three private-car bands), because IRDAI in early 2019 extended the prevailing rates “until further orders” and no subsequent order ever came. CNBC-TV18 confirmed in mid-2025 that “Motor TP premiums have not been increased for 4 years” and that IRDAI had sought an average 18% increase; as of the ICICI Lombard Q1FY26 commentary, “none announced.”
The rate has therefore been frozen for approximately six years while medical inflation ran roughly 13-14% annually (among the highest in Asia), repair costs climbed 30-40%, and tribunal awards compounded by judicial doctrine.
The disciplined reading is that this is not bureaucratic delay but a revealed political preference.
Ask who bears the cost of India’s road-accident compensation system and who is protected from bearing it. The vehicle owner — a vast, organised, politically salient electorate spanning every truck operator, taxi driver, and two-wheeler household in the country, is protected from price increases.
The insurer, roughly a third of whose capacity is state-owned, bears the residual. And here the political economy becomes self-reinforcing: the same exchequer that would absorb the political cost of a tariff hike is the exchequer that recapitalises the state insurers when their solvency collapses.
Government can either raise the price on voters or refill the insurers’ capital from the budget; it has consistently chosen the latter, because a recapitalisation is an opaque, back-page balance-sheet transfer while a tariff hike is a front-page tax on every vehicle owner. The freeze is the rational output of that incentive structure, not an accident of it.
Consensus treats the tariff as “a hike always eighteen months away,” whereas the more defensible reading is that the freeze is the permanent architecture of a system in which vehicle owners are politically protected and insurers are the designated, recapitalisable bearer of the compensation burden.
Model the three regimes. Under an indefinite freeze, third-party economics deteriorate further every year as awards compound against a flat price, own-damage cross-subsidy and investment income remain the only offsets, and the state insurers require perpetual recapitalisation — the segment is a levy, not a business.
Under a single catch-up revision (the 18-25% range floated), the arithmetic still lags: with awards having compounded well over 50% cumulatively over six years and the homemaker judgment adding a further 12-15% to loss ratios, even a 25% one-time hike merely narrows, rather than closes, the gap, and does nothing to prevent renewed erosion thereafter.
Only a return to annual indexation, the formula IRDAI once committed to, linking revisions to inflation and claims experience, would make third-party a genuinely commercial line.
VI. The Supreme Court intervention, enforcement, and the uninsured pool
In an order dated 4 August 2026 Supreme Court found that approximately 56% of vehicles on Indian roads — about 16.54 crore of 30.48 crore — are uninsured, and issued structural directions: integrating Automatic Number Plate Recognition cameras with the VAHAN and Insurance Information Bureau databases to generate automatic e-challans, a fuel-linkage proposal (no fuel for uninsured vehicles), a four-layer policy structure, and an extension of mandatory third-party tenure for new vehicles from three to four years for cars and from five to six years for two-wheelers. The directions are live and their implementation — particularly the fuel-linkage and ANPR e-challan machinery — remains to be operationalised and should be tracked rather than assumed.
Analysed on their own terms, the consequences are genuinely double-edged. Extending mandatory tenure on new vehicles pulls premium forward and lengthens the float, favourable, and it improves persistency, but it simultaneously locks in a frozen-tariff price for a longer duration on each new policy, which is unfavourable precisely because the price is inadequate.
VII. The cost side — severity inflation, repair economics, and the electric transition
Third-party severity is driven by judicial award inflation, medical costs, and settlement pace — all now compounding, and all recently accelerated by the homemaker judgment and the Court’s push for faster disposal.
Own-damage severity is driven by parts prices, labour, vehicle complexity, and increasingly by electrification. On repair economics, the modern vehicle is a worse insurance risk per rupee of sticker price than its predecessor: sensors, cameras and radar are concentrated in the bumpers and windscreen, the most common impact zones, so that a minor collision now triggers not a panel swap but the replacement and recalibration of expensive electronics; integrated body structures mean localised damage propagates into larger repairs.
Advanced driver-assistance systems cut both ways, reducing accident frequency while raising the severity of each accident that does occur, so the net effect on loss cost is ambiguous and depends on the maturity of the fleet.
Electric vehicles sharpen every one of these dynamics and add an asymmetry the current tariff does not reflect.
On the third-party side, the regulated premium carries a 15% discount for EVs, a subsidy embedded in the frozen schedule. On the own-damage side, EV premiums run materially higher, commonly 20-40% above the equivalent internal-combustion car on the own-damage portion, with Policybazaar putting the gap at 20-25% (for example, insuring a Tata Nexon EV costs roughly ₹9,388 more per year than the petrol Nexon) — because the lithium-ion battery pack represents 40-60% of the vehicle’s insured declared value, cannot usually be repaired once its casing is compromised (a stone-strike denting the casing “often cannot be repaired safely”), and must be replaced at a cost of roughly ₹3-8 lakh — frequently exceeding the entire value of an older conventional car, and in some models (a Tata Nexon EV pack at around ₹6.5 lakh, versus ₹15,000-25,000 for a petrol panel repair) exceeding the on-road price of a petrol hatchback.
Specialised labour, high-voltage certification, and the absence of a competitive independent-garage network remove the insurer’s usual leverage over repair cost.
The pricing is being set against a thin claims dataset, and as the EV share of the parc compounds the risk is that today’s premiums prove to have been struck on optimistic loss assumptions, a classic pattern in which a new vehicle technology is underpriced until its claims mature.
VIII. Market structure, subsidised capacity, and the coming capital regime
The largest motor writers were ICICI Lombard at roughly ₹11,552 crore, New India Assurance at ₹10,728 crore, Tata AIG at ₹9,805 crore, United India at ₹9,024 crore, and Bajaj Allianz at ₹7,278 crore.
Two of the top five — New India and United India — are state-owned, and the public-sector trio outside New India carries catastrophic solvency. At end-FY26, United India’s solvency ratio was approximately -136% (from -65% a year earlier), National’s -111% (from -67%), and Oriental’s -163% (from -103%), against a regulatory minimum of 150%.
These are not merely weak balance sheets; they are deeply negative, meaning liabilities exceed admissible assets by a wide margin, and yet these insurers continue to underwrite motor business at scale. The estimated capital required to restore them is variously put at ₹15,000-17,000 crore in the orienting data, though official/industry communications to the Department of Financial Services have at times pegged the collective need closer to ₹12,000-13,000 crore (roughly ₹2,000-3,000 crore each); either figure is very large relative to their earning power.
When a material share of industry capacity is supplied by entities that are not required to earn a return on capital, are insolvent on a regulatory measure, and are recapitalised from the exchequer when they fail, the marginal price of motor own-damage is set by a bidder indifferent to underwriting profit.
Rational private insurers describe exactly this: heavy discounting and a lowest-price bidding dynamic in own-damage, sustained by capacity that has no cost-of-capital discipline. The persistence of a 128% industry motor combined ratio is, in significant part, the fingerprint of subsidised irrationality.
Two reforms could change this, and one could entrench it. The Sabka Bima Sabki Raksha (Amendment of Insurance Laws) Act, 2025 received presidential assent on 20 December 2025, took effect on 5 February 2026, and permits 100% foreign direct investment under the automatic route plus a composite licensing framework; the enabling FEMA Non-Debt Instruments amendment was notified on 2 May 2026.
In parallel, a risk-based capital framework is being introduced alongside the Ind AS transition. The constructive case is that risk-based capital finally forces loss-making, undercapitalised capacity to price risk properly or exit, and that 100% foreign ownership brings disciplined global underwriting capital.
The bearish case is that recapitalisation of the state insurers simply continues — as it has through repeated cycles, with the Cabinet approving ₹12,450 crore across FY20 for National, Oriental and United India and a further roughly ₹5,000 crore tranche in FY22, perpetuating the irrational capacity, while 100% FDI merely adds more competitors chasing the same compulsory volume, intensifying the own-damage price war rather than disciplining it.
Which of these dominates is the second most important variable in the sector after the tariff to watch.
IX. Moat, honestly — where durable advantage actually lives
Motor own-damage has most of the characteristics of a commodity: a legally mandated product, an aggregator-driven price-comparison distribution layer, and a claim that any competent insurer can pay.
Yet a persistent gap between the best operator and the industry does exist and must be explained rather than assumed away or dismissed. Rank the candidate sources of advantage by durability. Most durable is claims-cost control, owned or contracted garage networks, parts procurement at scale, salvage recovery, and fraud detection, because it compounds with volume and is hard to replicate quickly; this is where a genuine, persistent loss-ratio edge can live.
Next is risk selection through data and granular pricing, which is real but decays as telematics and analytics diffuse to all players. Distribution capture through OEM and dealer relationships, bancassurance, and agency scale is valuable but increasingly expensive, since the dealer extracts much of the economics at the point of sale (₹20,000-30,000 of commission on a ₹50,000 new-car policy is common).
Expense-ratio advantage from scale and digital operations is real — Go Digit’s management-expense ratio to gross written premium is among the lowest in the industry, but is a cost edge, not a franchise.
The test is whether any of this shows up as a persistent loss-ratio or expense-ratio gap over a full cycle rather than in a single lucky year.
The durable edge, where it exists, is narrow, resides mainly in claims-cost machinery and underwriting discipline, and belongs to very few operators.
X. Where profitability accrues — a taxonomy of the listed universe
Profitability in this sector is unevenly distributed and always has been, and the taxonomy matters more than any single company.
Among multiline private insurers, ICICI Lombard is the clear standout: FY26 profit after tax of ₹2,772 crore (up 10.5%), return on equity of 18.1%, combined ratio of 103.4%, and critically its return on equity is delivered despite underwriting losses in most segments, rescued by investment income and realised capital gains (capital gains net of impairment were ₹9.33 billion in 9MFY26; investment yield including gains around 8-9%). Its motor book is disciplined but still runs a combined ratio above 100; the group’s profitability is a float-and-investment story wrapped around better-than-industry underwriting.
Go Digit, the motor-concentrated digital insurer, illustrates the other pole: its combined ratio widened to 107.2% in Q1FY27 from 104.6%, profit after tax fell 37.5%, and its motor market share slipped to 5.6% as management deliberately shrank unprofitable private-car and commercial-vehicle business, a candid admission that motor at current prices does not pay.
GIC Re, the state reinsurer, carried a full-year FY26 combined ratio of 106.02% (improved 2.79 points), with its motor-dominated “miscellaneous” segment the principal drag, and delivered its FY26 net profit of ₹8,392 crore substantially on investment income and a strengthening solvency ratio of 4.21x.
The state direct insurers outside New India are loss-making and insolvent.
XI. Valuation — lenses in tension
On normalised through-cycle underwriting, the question is what combined ratio a good operator can sustain once the tariff normalises and reserves are adequate.
For a best-in-class multiline insurer, a through-cycle combined ratio around 102-104% is plausible, which, combined with float income at an 8-9% yield on 3.5x investment leverage, supports a mid-to-high-teens return on equity.
For a motor-concentrated insurer at a frozen tariff, the sustainable combined ratio is above 105% and the return on equity depends almost entirely on investment income, which is not a franchise the market should pay a premium for.
On a float-and-book-value lens, the best private insurer trades at a rich multiple of book because its return on equity is elite and durable, but book value in a long-tail line is itself a reserve estimate, and the homemaker judgment plus Ind AS 117 discounting mean that reported book is about to become less comparable and less reliable, arguing for a discount to stated book precisely where the reserves are least conservative.
On a scenario-weighted tariff lens, the value gap between “continued freeze,” “single catch-up revision,” and “restored annual indexation” is the widest single driver of segment value; a motor-heavy insurer is worth materially more under indexation than under a permanent freeze, while a diversified insurer is far less sensitive.
What is the market implicitly assuming?
The premium multiples on the best private insurer (analyst targets have valued ICICI Lombard around 30x forward earnings) imply an expectation of eventual repricing, continued reserve adequacy, and a benefit from the accounting transition.
Where a premium multiple is being paid, what has to go right is: a real tariff revision that at least matches claim inflation, reserves that do not require strengthening as the homemaker standard propagates, and an Ind AS transition that does not reveal larger economic liabilities beneath the discounting gloss.
The margin of safety is thin for a motor-pure exposure.
XII. Risks and falsification
Discipline requires separating cyclical headwinds from structural impairment and naming, for each, the observable that would confirm it.
Cyclical: weak vehicle sales, a soft own-damage pricing cycle, and a bad catastrophe year.
Structural, and far more dangerous: permanent political refusal to reprice the tariff (confirmed by continued absence of a MoRTH gazette notification through successive fiscal years); permanent subsidised loss-making capacity (confirmed by further recapitalisation of the negative-solvency state insurers rather than their restructuring or exit); disintermediation by manufacturers, dealers, and platforms (confirmed by falling insurer take-rates and rising dealer/aggregator commissions); and systematic under-reserving that unwinds over a decade (confirmed by adverse prior-year reserve development in the ICRA triangles and by the scale of reserve strengthening forced by the homemaker judgment).
What breaks a bearish thesis, equally: sustained enforcement (ANPR e-challans, fuel-linkage) that genuinely expands the insured pool at repriced rates; risk-based capital that forces undisciplined capacity to exit rather than be refilled; and a demonstrated, persistent loss-ratio advantage at one or two operators that survives a full cycle.