Advertise - Write For Us

If you find yourself wondering why everyone you talk to seems to be starting an MGA, you're not alone. Insurance is broadly becoming less vertically integrated. The traditional carrier controls nearly every part of the value chain through product design, pricing, underwriting, distribution, policy administration, claims management, projection of ultimate losses, and capital management. Increasingly, those functions are being separated and performed by specialized organizations. We have to ask ourselves as insurance professionals how to understand these macroscopic changes, and as actuaries how we fit into this trend, along with what new tools, skills, and broadened horizons will open for us in the process.

Managing general agents and underwriters, program administrators, fronting carriers, third-party administrators, reinsurers, and technology vendors now occupy much of the space between the insurer and the policyholder. The scale of this shift is clearest in property and casualty, where Gallagher Re's 2026 MGA Market Report estimates that MGAs now account for roughly 12.5% of total U.S. P&C premium. No longer a niche channel, a meaningful share of American P&C risk is priced by someone other than the carrier whose name appears on the policy.

The operations within that channel are increasingly modularized. An MGA may control distribution, pricing and underwriting; a fronting carrier supplies licensed paper and regulatory oversight; a reinsurer supplies much of the risk capital; and separate firms provide claims, data and policy administration. Consider a fronting specialist like State National, a licensed balance sheet and a compliance function, wrapped around programs that other organizations originate, price and largely reinsure. Together with Transverse, it accounts for 22% of Gallagher Re's program carrier composite. Despite this, neither firm resembles a traditional insurer. Each function in that chain is managed by a firm that specializes in it, compensated differently and answerable to different incentives and stakeholders. The premium dollar passes through all of them before reaching loss bearing layers. The same coverage in that report notes that roughly a third of premium assumed from the composite now comes from unrated reinsurers rather than traditionally rated balance sheets. The modern insurance enterprise is assembled from a network of interconnected specialists paired with external capital. This creates an interesting risk management challenge as incentives decentralize and professionals work together across firms on the same risk dollars.

Newsletter continues after job posts…

Life insurance appears in a stage of this model that looks similar to P&C fifteen years ago. P&C then had a substantial delegated-underwriting market but almost no industry devoted to renting balance sheets to it; the hybrid fronting carriers that now define the structure were mostly founded between 2015 and 2018. Life appears to sit at that stage now. Brokerage general agencies and independent marketing organizations rarely set mortality assumptions or bind the carrier, but we continue to see third-party distribution, with LIMRA reporting that independent channels produced 60% of U.S. individual life premium in 2024. The intermediary itself is becoming layered, per the 2026 LIMRA–NAILBA Inside the Intermediary study, 80% of intermediaries reported affiliation with a national marketing organization.

Despite intermediation in these links along the value chain, what life still mostly lacks is the balance-sheet-light operating company, and regulation, especially in the US, seems to be the biggest hurdle for this business model to succeed. State-by-state licensing, nonforfeiture and illustration requirements, and long-duration statutory reserves make it far harder to rent a balance sheet and return capital quickly to venture investors than in P&C. But the components appear to exist for a similar structure to emerge in this domain. Ethos, a licensed producer and third-party administrator, runs digital, rules-based underwriting and policy administration while its policies are issued by carriers such as Banner Life, Ameritas and TruStage, with multiple reinsurers behind them. Distribution, underwriting workflow, licensed paper and risk capital sit in four different organizations. This is close to the fronting architecture P&C has normalized, applied to term life.

This author expects to see this model applied less to traditional life lines, where competition compresses return, and more to specific, niche markets where underwriting expertise and high quality partnership with third parties play a large role, such as repatriation and remittance coverages. We should soon expect the first true wave of balance-sheet-light life structures to follow as Insurtech drives more new startups within life lines who might just decide they want to take on some of the good or niche risk they see out there.

The preceding discussion is part of a broader economic movement toward modular firms. Cloud infrastructure and digital platforms lower the cost of coordinating work across organizational boundaries. Artificial intelligence accelerates the process by compressing the manpower required for document review, coding, research, customer service, data preparation and administration. Falling and subsidized AI costs shrink the minimum efficient scale of a specialized insurance business. A new general agency can purchase policy administration, e-applications, claims workflows, commission systems and external data from vendors. AI helps a relatively small team process submissions, grow, and monitor the portfolio. Across the insurer broadly, specialized teams operate with fewer functional silos, and the individual professional sits closer to decisions that once belonged to several separate departments.

Adoption of AI among actuaries is definitively more shallow than the enthusiasm on our LinkedIn feeds would suggest. A recent SOA member survey found approximately 80% of responding actuaries using AI, but most for fewer than five hours per week. These systems are producing work product faster than the profession has built habits for checking it. This remains a bottleneck for agentic AI implementation in the actuarial function at scale. Either the efficiency compression is limited by the detailed audit of the produced work, or the work is produced efficiently but not audited to the same level of detail as work produced by the actuary. In that sense, the age of AI calls the actuary to be more efficient than ever as a communicator of their work product.

Compression makes it easier to launch and scale a program, but it remains far from easy to price and source good risk. The actuary in an intermediated structure is less likely to sit within a large, centralized department devoted exclusively to one recurring process. The greater need is for someone who can connect pricing, underwriting, distribution, capital, technology and emerging experience, who fully understands who and what makes each decision, how each participant is compensated, where the data originate and who ultimately bears the risk. Think of the processes mentioned here in your own firm, and the economics of the stakeholders in each stage of those processes. If it were intermediated across those stages, how would it change your role?

As repetitive work is compressed, the actuary’s comparative advantage moves toward problem framing, model governance, economic interpretation and communication across organizations. The profession’s opportunity is to help design and govern an insurance system in which teams are smaller, information is extensive, and responsibility is spread across a growing network of specialists. As these trends continue, there is increasing pressure for the actuary to become a respected and acknowledged leader of communication throughout the insurance value chain. Further developing yourself along this axis in the short term will best prepare you for leadership in a more intermediated world.

Austin Riis-Due, FSA, is a PhD candidate in Actuarial Science and Quantitative Finance at the University of Waterloo, having become a Fellow of the Society of Actuaries at 24, years ahead of the typical career track. His research spans stochastic control, reinforcement learning, and decentralized risk sharing, giving his writing on industry structure a rare blend of academic depth and practitioner instinct.