In an exclusive CNBC-TV18 interview published on September 30, IRDAI Chairman Ajay Seth said there is currently no economic rationale for the managing general agent model, and that a party can be an insurer or a distributor but cannot be both at a time. The regulator's approach, he said, is intended to ensure that the roles of manufacturers and distributors remain clearly separated. Those two sentences are the most consequential thing the chairman has said about market structure, and they land in the same week that Policybazaar's parent told an analyst call that the managing general agent model recognised by the new Insurance Act was one of the routes it was exploring to replace the distribution income its own business model depends on.

Context: what the interview adds to the consultation paper

The interview covered the same 160-page paper, Recalibrating Economics of Insurance Distribution, and much of its content is about the chairman's case for it in numbers: costs at 38 percent now against 8 percent a few years ago, over 44 percent profit after tax described as too high for a financial services sector, and the cost of doing business at one of the largest private insurers rising from under 12 percent to 18 percent. His framing of the goal is that the sector was that efficient seven years ago and should be able to return to that position, with the stated constituency of the shock being policyholders rather than anyone else.

The structural statements sit alongside that. The chairman set out two architectures explicitly. Insurance Distribution Entities, the entity-level category that replaces brokers, master agents and similar groupings, are designed as an open architecture, meaning a broker can hold relationships with more than one insurer. Insurance Distribution Persons, the individual-level category, can sell only from insurers with which they have a tie-up. He separately said that on managing general agents there is currently no economic rationale for the model, and that the intended discipline is that you can be an insurer or a distributor but not both at a time.

Implication: why an MGA is the hardest test in the paper

  • An MGA captures the underwriting margin and the expense line at the same time, and that is precisely the problem. Expenses of management is calculated for the insurer. If the same commercial group manufactures the risk, designs the product, sets the price and distributes it, it can choose where the cost sits. A distributor fee paid by the insurer to an affiliate is inside the insurer's expense measurement, but an MGA earning its economics through the product margin rather than through a disclosed distribution fee is not. The paper's stated intention of looking at all payments to distributors is aimed at exactly this, and the chairman's flat statement that the two roles cannot coexist is the regulator saying the structural route out is separation, not disclosure.
  • The chairman's position and a named company's stated strategy are now in direct tension. If the model has no economic rationale in the regulator's view, a distributor platform cannot build a plan around entering it. That is a materially different outlook from the one reported from the Policybazaar analyst call in the same week, where the MGA model was named among the diversification routes being explored alongside credit life, monetising existing services and setting up an in-house insurer. Watch for an IRDAI clarification on the MGA definition in the draft regulations, because a definitional carve-out is the only thing that would reconcile the two positions.
  • Open for entities, closed for persons, is a structural bet on who survives. The distinction matters more than it first appears. An open architecture means a multi-insurer broker can shop across the market and be paid by several insurers, which the commission caps treat unfavourably by paying multi-insurer channels less than tied ones. A closed architecture means an individual agent is locked to a single insurer. Read together, the two statements tell a person selling insurance which side to be on: the regulator is pricing the independent multi-insurer channel low and building its architecture around single-insurer ties. For anyone considering a career in distribution, that is the operative signal, and it points away from independence and towards agency.
  • Role separation at the point of sale is the part a policyholder actually feels. If the final regulations hold the line that a party cannot underwrite and distribute, then the seller of your policy is more likely to be someone who is not carrying the risk. That reduces the conflict at the moment you are deciding, which is the moment it matters. It does not by itself tell you anything about commission rates, and it is the one part of this consultation that could improve the buyer experience without any change in price.
  • The unresolved piece is enforcement, not principle. Vertical separation is easy to state and hard to police, because ownership can be routed through holding structures and affiliations. The consultation paper's proposal to make the identity and conduct record of sales personnel available through the Public Insurance Registry is the mechanism that would make separation verifiable, and that registry is the piece a buyer should follow, because it is the only part of this reform that a policyholder can use directly.

Action

For a policyholder, the question this interview makes newly answerable is short: ask the person selling you whether the company they work for also carries the risk. If the answer is yes, that is a conflict of interest the regulator has now said in public it does not accept, and you are entitled to think about it before you sign. The question costs nothing and the answer is usually available in a sentence.

For the industry, the thing to watch is drafting rather than principle. Specifically: how an MGA is defined, whether an MGA's remuneration is inside or outside the expense measurement base, and whether the tie-up requirement for individual distribution persons is drafted as a prohibition or as guidance. The first two decide whether the model is viable at all. The third decides whether open architecture for entities means anything in practice.

Watch item: the chairman is scheduled to meet insurance chief executives in Delhi on October 5, and written feedback closes on October 25. If the MGA point is picked up in the written responses, which it now should be, the exposure drafts will carry the definition and the answer will be legible in the text. If it is not picked up, the absence is itself informative, because it would mean the regulator's stated position on MGAs never reached the drafting.