Acquisition Cost in the Rate: How Commission Loading Enters Technical Pricing

Ask a broker what the commission on a placement is and you get a number. Ask an underwriter and you get a component of an equation, sitting in a specific place, doing arithmetic work.

That difference explains most of the mutual incomprehension in an Indian commercial rate negotiation. The broker experiences commission as income and the rate as a separate variable the underwriter controls. The underwriter experiences them as two terms in one construction, where moving either moves the other, in a direction most brokers do not assume. This post is written from the second chair.

The occasion is a divergence in the published numbers. Non-life commission expense reached roughly INR 47,266 crore in FY2024-25, against roughly INR 39,601 crore the year before, close to 19 percent growth. General insurance premium over the same period grew about 8.5 percent. Distribution cost grew at more than twice the rate of the premium base it is paid out of.

For a regulator, that gap is a distribution-reform question. For an actuary it is narrower: it says the acquisition loading assumed in last year's rates was wrong, and the difference had to come from somewhere.

Every commercial rate an Indian insurer quotes is a version of one build.

1. Expected loss cost. What the risk is expected to cost in claims over the policy period, from the account's own experience where credible and portfolio burning cost where not, adjusted for exposure change, claims inflation and trend. The only component about the risk.
2. Expense loading. The cost of doing business other than acquiring the account: underwriting, policy administration, claims handling, technology, compliance, the branch.
3. Acquisition cost. What it costs to get the account through the door. Commission or brokerage is the largest and most visible element.
4. Profit and risk margin. The return required for putting capital behind the volatility, net of expected investment income.

The premium is whatever makes those four add up. What brokers rarely internalise: components two, three and four are all expressed as percentages of the premium, while component one is expressed in currency. Loss cost is the only component that does not scale with the answer. The rest are shares of the number being solved for, which makes the technical premium a fixed-point calculation rather than an addition.

Write the build out. Let the technical premium be P, the expected loss cost be L in currency, and let c, e and m be acquisition cost, other expenses and profit margin, each expressed as a fraction of premium.

The premium has to cover all four:

P = L + cP + eP + mP

Solve for P and you get the form every pricing actuary uses:

P = L divided by (1 minus c minus e minus m)

Commission is not added to the rate. It is divided into it, and that single fact drives everything else.

Take an account with an expected loss cost of INR 100, other expenses at 15 percent of premium, and a required margin of 5 percent.

• At 10 percent commission, the technical premium is 100 divided by (1 minus 0.10 minus 0.15 minus 0.05), which is 100 divided by 0.70, or INR 142.86.
• At 15 percent commission, it is 100 divided by 0.65, or INR 153.85.

Five points of commission did not add five percent to the rate. It added 7.7 percent.

The gearing worsens as the denominator shrinks: on an account already carrying heavy expense and a thin margin, every additional point costs the client proportionally more, because each point is recovered from a smaller residual. This is why an underwriter reacts badly to a commission request on a stressed account in a way that looks disproportionate. It is not temperament. It is division.

The uncomfortable corollary: a broker negotiating a higher commission is negotiating a higher rate for their own client. Not one for one. More than one for one. If the rate holds, the money comes from somewhere else.

Now the part that appears in no textbook build, because it is a failure of process rather than a feature of the model.

The acquisition loading in the rate is a pricing assumption, set at portfolio level, often annually, by the actuary. The commission actually paid is a transaction outcome, set at account level, continuously, by whoever is in the room. Different numbers, different people, different clocks, and in most Indian general insurers nobody owns the reconciliation. The gap opens in predictable places:

• The rate is filed and the deal is done later. The pricing basis assumes a portfolio-average acquisition ratio; the individual placement is negotiated against a competitor's quote at a rate the account executive needs to win. The commission conceded was not in the filing.
• Variable components are invisible to the build. The rate is loaded for the grid rate. Volume-linked, retention-linked and campaign-linked components accrue later, at portfolio level, never touching the account's pricing sheet.
• The mix moves and the loading does not. A portfolio loading is an average over an assumed mix. Realised yields in the Indian market run roughly 8 to 10 percent on large risks, 12 to 15 percent on mid-market package business, and 15 to 20 percent on retail health and miscellaneous retail. A book drifting from the first band to the third has repriced its own acquisition cost without any underwriter making a decision.

The FY2024-25 divergence is this gap aggregated across an industry and printed. Something was loaded for in the rates written that year. Something else was paid.

Rate adequacy is narrower than the phrase suggests. A rate is adequate when it covers the expected loss cost, the expenses actually incurred, the acquisition cost actually paid, and the cost of the capital supporting it. Not the expenses assumed. The ones incurred.

That distinction is where an insurer loses money without anyone doing anything obviously wrong. Take the same account priced at INR 142.86 on a 10 percent acquisition assumption, then placed at 15 percent commission because the market required it, with the rate holding at 142.86:

• Commission paid: 15 percent of 142.86 = INR 21.43
• Other expenses: 15 percent of 142.86 = INR 21.43
• Available for losses and profit: 142.86 minus 42.86 = INR 100.00
• Expected losses: INR 100.00
• Underwriting profit: zero

Priced at exactly breakeven before a single claim behaves unexpectedly. Nothing went wrong with the risk, the loss cost was estimated correctly, and the rate was calculated correctly on its own assumption. The assumption simply was not the deal that got done.

The rule is exact: when the rate does not move, every point of additional acquisition cost is a point of underwriting profit, one for one. The gearing that made five points cost the client 7.7 percent on the rate side becomes a flat one-for-one transfer on the margin side when the rate is frozen. The client is protected by the market. The underwriter is not.

This is why an underwriter judged on loss ratio and one judged on combined ratio behave differently, and why only one is judged on their actual job.

Loss ratio is incurred claims over earned premium. On the account above it is 100 divided by 142.86, or 70 percent, whether the commission paid was 10 points or 15. Commission is not a claim. It does not appear.

Combined ratio adds the expense side: claims plus all expenses, including acquisition, over earned premium.

• At 10 percent commission: 70 percent loss ratio plus 25 percent expense ratio (14.29 plus 21.43, over 142.86) = 95 percent combined. Five points of underwriting profit.
• At 15 percent: 70 percent plus 30 percent (21.43 plus 21.43, over 142.86) = 100 percent combined. Nothing.

Same risk, same claims, same rate. Five points of combined ratio, entirely distribution.

The regulatory frame closes the loop. The IRDAI (Expenses of Management, including Commission, of Insurers) Regulations, 2024, effective 1 April 2024, cap an insurer's aggregate expenses at roughly 30 percent of gross written premium for general insurers and roughly 35 percent for standalone health insurers. That ceiling sits on the expense side of exactly this equation. An insurer near it cannot concede acquisition cost anywhere without taking it back somewhere, which is why a commission conversation with a constrained carrier feels like a conversation about something other than the account.

The build says the rate should rise 7.7 percent for five points of commission. The market says the rate is what the competitor quoted. Both are true, and the difference goes to one of four places.

1. The margin. The default, and the arithmetic above. Five points of commission, five points of combined ratio, continuing until the account is at 100 and then past it. Visible, honest, finite.
2. The loss cost assumption. The dangerous one. When the required rate will not clear the market, the pressure travels backwards up the build and lands on the only component with interpretive latitude. The trend assumption softens. The large-loss load gets described as conservative. Nobody records a decision to underprice, and the reserve strengthening arrives three years later attached to no particular quarter.
3. The terms. The rate holds and the cover shrinks: a higher deductible, a tightened warranty, a sub-limit that was not there last year, a narrower business interruption indemnity period. The most rational response, and the one the client feels last, at the claim.
4. Someone else's account. Cross-subsidy from a segment with room, sustainable as long as that segment stays uncontested.

Which is why the FY2024-25 divergence is a pricing story before it is a distribution story. Ten points of gap, absorbed somewhere. Some was margin, some was terms, and some, in some portfolios, was option two.

Nothing here says brokers should ask for less, only that both parties should know what is being asked.

For the broker. A commission request is a rate request with a multiplier on it, and the multiplier depends on how much room is left in the denominator. If the underwriter is not moving the rate, the ask is coming out of the account's margin, and therefore out of the underwriter's appetite at the next renewal, or the terms at the next claim.

For the underwriter. The reconciliation between the loading assumed and the commission paid is somebody's job and in most Indian insurers it is nobody's. The minimum viable discipline is an acquisition-cost variance report at segment level: loaded ratio against paid ratio, including variable components, quarterly, mix effect separated from negotiation effect. Without it the actuarial function is repricing against an assumption distribution has already abandoned.

On the reform conversation. IRDAI has signalled a consultation paper on commission rules, and press reporting in July 2026 attributes to it staggered or trail structures, remuneration differentiated by distributor effort, product-wise caps by complexity and tenure, and tighter disclosure. The paper had not been published as of the date of this post, and every one of those ideas is a proposal. The reported figure of up to roughly 40 percent of premium on some life and health products is an observed market level, not a cap and not a commercial-lines number. The FY26 and FY27 expense ceilings are anticipated rather than notified.

If acquisition cost does move from an upfront concentration to a payment spread across the policy life, the loading stops being a single ratio applied at inception and becomes a stream whose present value depends on persistency. Persistency becomes a pricing assumption rather than a retention statistic, a harder problem than the one the industry has today, landing on the same function that is currently not reconciling a single static ratio.

The underwriter holds both sides. The rate the client pays and the commission the broker earns are one equation viewed from two chairs, and the only party sitting where both are visible is the one being asked to concede.