The Plumbing Nobody Talks About: Capacity, Reinsurance, and the Real Mechanics of Delegated Authority
Most industry commentary on capacity treats it as an abstraction. Rates are hardening. Rates are softening. Capacity is plentiful. Capacity is constrained. The language is atmospheric, as if capital moved through the insurance industry the way weather moves across a map.
In my experience, that framing misses what actually happens. Capacity is not weather. It is a set of contracts, relationships, authority grants, reporting obligations, and trust-based arrangements that get renegotiated at specific moments of the year by specific people who remember specific things about how the last cycle went. The flow of underwriting capital is deeply operational, and anyone who has sat through a treaty renewal or defended an MGA's loss ratio to a carrier capacity committee knows that the gap between industry narrative and industry reality is usually wider than the trade press suggests.
I want to walk through what actually governs the movement of capacity in 2026, because I think the market is entering a period where the mechanics matter more than they have in a decade. The combination of softening reinsurance pricing, growing delegated authority footprints, AI-driven workflow transformation, and heightened regulatory scrutiny is creating a set of conditions where operational excellence inside the delegated authority channel becomes the primary differentiator. The MGAs and carriers that understand the plumbing will outperform. The ones still treating capacity as atmospheric will find themselves surprised.
Where Capacity Actually Lives
Before discussing how capacity moves, it helps to be precise about where it sits.
In the U.S. E&S and specialty market, underwriting capital is ultimately provided by a relatively small number of balance sheets: traditional carriers, reinsurers, Lloyd's syndicates, hybrid fronting entities, collateralized reinsurance vehicles, and increasingly, insurance-linked securities structures. Each of these balance sheets has a cost of capital, a return hurdle, a regulatory posture, and an appetite envelope shaped by cat aggregation, line-of-business concentration, and rating agency concerns. These are the sources.
Between those sources and the ultimate insured sits the distribution and structuring layer. Retail brokers produce business. Wholesale brokers access non-admitted markets. MGAs and MGUs exercise delegated underwriting authority on behalf of carrier partners. Program administrators assemble books of business for specific industry verticals. Fronting carriers lend their paper and rating in exchange for fees, while retaining little net exposure.
What we colloquially call "capacity" is really a cascade of agreements connecting these layers. A single E&S policy written by an MGA might reflect a binding authority agreement with a fronting carrier, which in turn has a quota share treaty with three reinsurers, one of which has retroceded a portion of its participation to a collateralized vehicle managed by an ILS fund. The policy feels like a single product to the insured. To the people who structured it, it is a chain.
The implication, which tends to get lost in conversations about platform infrastructure, is that moving capacity faster does not mean moving paper faster. It means moving through that chain faster while preserving the information integrity each link requires. A reinsurer giving a carrier authority to delegate authority to an MGA is making a bet not just on the risk, but on the reporting, the governance, the auditability, and the behavioral discipline of everyone downstream. Speed without information integrity is not speed. It is risk transfer to the wrong party.
The 2026 Capacity Environment
The market we are sitting in today is genuinely unusual.
Reinsurance treaties for 2026 renewed at material rate decreases, pending no major losses, with the possibility of select carriers securing lower attachment points from their treaty reinsurers. At the same time, global reinsurer capital reached $760 billion by September 30, 2025, with third-party capital at $124 billion. Capital is abundant. Pricing is loosening. AM Best has moved its outlook from positive to stable, which is itself a signal that the market is transitioning out of its extended hard cycle.
For MGAs and their carrier partners, this creates a specific set of tensions. On one side, softer reinsurance pricing and ample capital mean that well-performing delegated authority platforms can receive additional PML for deployment in 2026 from existing panelists and new entrants. On the other side, as competition intensifies and rates moderate, the gap between MGAs that are truly disciplined and MGAs that have been riding a favorable market widens dramatically. Loss ratios that looked acceptable under 2023 pricing look very different under 2026 pricing. Programs that could absorb a little bit of authority creep in a hardening market cannot absorb it in a softening one.
This is the point in the cycle where capacity discipline stops being theoretical. Carriers and reinsurers do not pull a program because of a single bad quarter. They pull it because the signals they rely on to track portfolio health become unreliable, and trust degrades below the threshold required to keep delegating. I have seen this pattern play out more times than I can count. The MGAs that lose capacity are not always the ones with the worst loss ratios. They are the ones whose reporting made it impossible for their capacity providers to explain what was happening in the book, or whose portfolio drift made the carrier feel that delegated authority had quietly become abdicated authority.
Underwriting governance is becoming a strategic asset, as loss ratio volatility, authority creep (unintentionally overstepping authority limits in underwriting decisions), and inconsistent results will reshape how capacity is allocated. That single sentence captures where the industry is headed more accurately than most of the 2026 market outlooks I have read.
What Treaty Renewals Actually Negotiate
If you have never been in a treaty renewal conversation, it is worth understanding what gets negotiated, because the outcomes determine what MGAs can do the rest of the year.
A quota share treaty between a carrier and its reinsurance panel typically negotiates four variables that matter intensely to delegated authority programs: the cession percentage, the ceding commission, the loss ratio cap or sliding scale, and the profit commission structure. These are not abstract levers. They translate directly into MGA economics.
When ceding commissions compress, which happens when reinsurers push back on underwriting quality or when broader market economics tighten, MGAs feel it immediately. Their commission revenue is a function of what the fronting or issuing carrier can afford to pay, which is a function of what the reinsurer will allow. When sliding scales tighten, the MGA's upside in good years gets capped while the downside in bad years gets sharper. When profit commissions shift from calendar-year to multi-year smoothing, the MGA's cash flow profile changes in ways that affect its ability to invest in technology, hire underwriters, or pursue new capacity.
In my experience, very few MGA founders fully appreciate how much their business model is a pass-through of treaty economics. The business feels like distribution and underwriting expertise, and it is both of those things, but the economics are shaped upstream by negotiations the MGA often does not participate in directly. One of the reasons the MGA model has been so profitable over the last several years is that reinsurance market dislocation pushed more business into delegated channels, and the resulting ceding commissions were historically generous. That era is ending. Softer reinsurance and retro pricing has encouraged higher quota share cessions and ceding commissions at renewals. While this can support capacity availability, it can also compress economics for capacity providers and raise performance hurdles for DUAEs.
What this means practically is that operational efficiency is no longer optional for MGAs. The programs that survive the next cycle will be the ones that can produce better loss ratios, better reporting, and better governance at lower internal cost. Technology is not a nice-to-have in that environment. It is the mechanism by which margin gets preserved.
The Information Asymmetry Problem
I want to spend some time on something that does not get enough attention: the information asymmetry between capacity providers and delegated authority platforms.
When a carrier or reinsurer delegates underwriting authority, they are extending a significant amount of trust. The MGA can bind business, commit the carrier's balance sheet, and create exposure obligations that the carrier will honor regardless of whether the underwriting decisions were ones the carrier would have made itself. The governance frameworks that exist around this relationship, such as binding authority agreements, underwriting guidelines, referral triggers, reporting cadences, and audit rights, are all designed to manage that asymmetry.
But here is the thing that people outside the delegated authority channel often miss: the asymmetry is real in both directions. Yes, the carrier does not see every risk the MGA binds in real time. But the MGA also does not see the carrier's full portfolio picture, does not know where the carrier's appetite has shifted since the last formal guideline update, and does not have visibility into whether a given account's bind is moving the carrier toward or away from aggregation concerns the carrier has not yet communicated.
Historically, this two-way asymmetry has been managed through a combination of personal relationships, quarterly reviews, and the willingness of both parties to pick up the phone. That works when everyone has been in the business for twenty years and knows each other's reputations. It does not scale. It does not work for new MGAs trying to access capacity without pre-existing relationships. It does not work for carriers trying to manage a diversified delegated authority portfolio across dozens of MGA relationships. And it particularly does not work when the business involves complex or emerging risks where appetite is genuinely uncertain on both sides.
The platforms that will define the next decade of delegated authority are the ones that solve this asymmetry structurally rather than relationally. That means real-time visibility into bound risk for capacity providers, continuously updated appetite signals for underwriters, and governance frameworks that function at the pace of the business rather than at the pace of quarterly reporting. Appetite is dynamic and shifts as the portfolio evolves, requiring real-time visibility and adaptive decision support. Any platform that treats appetite as a static PDF is building for the last market, not the current one.
The Real Cost of Fragmented Workflow
One of the things I have become increasingly convinced of is that most of the inefficiency in the E&S and specialty market is not about technology per se. It is about workflow fragmentation across parties who each have slightly different information, slightly different incentives, and slightly different tools.
Consider what happens when a retail broker submits a risk that might fit in delegated authority. The retail broker is working from their agency management system and whatever carrier appetite information they have been able to assemble. The wholesale broker receives the submission, matches it against their knowledge of market appetite, and routes it to MGAs or carriers they think will engage. The MGA receives the submission, ingests it into their underwriting system, checks it against their guidelines, potentially refers it to their carrier partner if it sits outside clean authority, and ultimately produces a quote. The carrier, behind the scenes, has its own portfolio management system tracking aggregate exposures. The reinsurer, further behind, has its treaty reports showing what is being ceded.
At each handoff, information is lost or transformed. The risk characteristics the retail broker entered get flattened into whatever fields the wholesale broker's system captures. The MGA's underwriter may or may not see the full submission history. The carrier may or may not know in real time what the MGA has bound. The reinsurer definitely does not know in real time.
According to Federato's 2025 State of Underwriting, underwriters spend 26% of their time on unwinnable deals, and 25% of submissions fall outside the portfolio's appetite. That statistic is not surprising to anyone who has worked in underwriting operations. It is the direct consequence of a market where appetite signals do not travel well across the chain, and where most of the intelligence about what actually fits lives in the heads of individual underwriters rather than in the systems the business runs on.
The opportunity here is not to eliminate the chain. The chain exists for structural reasons, and wholesale brokers, MGAs, and fronting carriers all provide real value. The opportunity is to make the information flow match the capital flow. If capital can move from a reinsurer's balance sheet to an insured's policy through four intermediaries, information about portfolio composition, appetite shifts, and exposure accumulation should be able to move back up that same chain at the speed the business actually requires.
Where AI Changes the Math
I want to be careful here, because I have been consistent in saying that AI as a value proposition is not a differentiator. Outcomes are. Every insurtech platform and every incumbent carrier is now claiming AI capabilities. Most of those claims are either narrow automation being rebranded, or pilots that have not made it past 30 days in production.
But the underlying shift is real. WTW's March 2026 survey found that insurers using sophisticated analytics achieved combined ratios six percentage points lower and premium growth three percentage points higher than slower adopters. Six points of combined ratio is not a marginal gain. That is the difference between a program that gets renewed with better terms and a program that loses capacity.
What actually changes when AI works correctly inside the underwriting workflow is not that humans get replaced. It is that the information asymmetry problem I described earlier becomes tractable. When appetite is embedded dynamically in the submission intake process rather than sitting in a PDF, when portfolio composition is visible at the point of bind rather than at the end of the quarter, when authority creep can be detected in real time rather than in a post-incident audit, the governance relationship between MGAs and capacity providers fundamentally improves.
The pattern that does not work is AI as a black box producing decisions that humans cannot explain. If AI is introducing appetite recommendations or decisions that feel more like a black box, you create a new kind of ambiguity. The answer might be faster, but it's harder to trust. It's harder to explain. In a channel built on trust between carriers, reinsurers, MGAs, and brokers, explainability is not a nice-to-have feature. It is the product.
The pattern that works is AI as a workflow enabler that makes underwriter judgment more consistent, portfolio signals more legible, and governance more real-time. That is a very different thing than automated decision-making, and in the delegated authority channel specifically, the distinction matters enormously. An MGA whose AI makes their book behave more consistently with their stated appetite is an MGA whose capacity providers will give them more authority next year. An MGA whose AI replaces underwriter judgment with opaque scoring is an MGA whose capacity providers will pull authority the first time a class hits a bad loss ratio.
What This Looks Like for Practitioners
For retail brokers and wholesale brokers, the implication is that submission quality becomes a differentiator in a way it has not been for decades. When carriers and MGAs have tools that can detect whether a submission is actually in appetite before an underwriter touches it, the brokers who produce clean, well-structured submissions will get faster quotes, better terms, and more attention. The brokers who do not will get routed to the bottom of the queue by systems that know the difference.
For MGAs, the implication is that operational infrastructure is now the thing capacity providers care about. The pitch that used to work, which was essentially "trust us, we have good underwriters," no longer works against competitors who can show their capacity providers real-time portfolio dashboards, automated guideline compliance monitoring, and audit-ready documentation of every binding decision. The MGAs winning new capacity in 2026 are not the ones with the best pitch decks. They are the ones whose capacity providers can see the book as clearly as the MGA itself can.
For carriers, the implication is that the delegated authority channel is not going away, but the governance model is being rebuilt. For carriers, the priorities are clear: protect combined ratios through disciplined underwriting, deploy capacity strategically rather than broadly, and invest in the audit and governance capabilities that provide early warning when portfolio performance drifts. Carriers that treat delegated authority oversight as a compliance function will underperform carriers that treat it as a portfolio management function. These are different things, and the distinction is starting to show up in combined ratios.
For reinsurers, the implication is that the quality of information coming up the chain from delegated authority programs is improving faster than most reinsurers have built the capability to consume it. Treaty negotiations in the next three years will increasingly reward programs that can demonstrate governance through data rather than through assertion, and will penalize programs that cannot. The reinsurers that build the capability to consume structured program data at scale will be in a position to offer differentiated terms. The ones that cannot will be price-takers.
For insurtech founders and operators, the implication is that the platform opportunity is larger than most of the current category leaders have articulated. The prize is not replacing any single layer of the distribution chain. It is building the infrastructure that allows capital to move through the chain faster while preserving the information integrity every party requires. That is a much harder problem than submission triage or document extraction, and it is the problem that will define which platforms actually matter in five years.
The Takeaway
The capacity conversation in 2026 is not really about whether there is enough capital. There is plenty. It is about whether the operational infrastructure connecting that capital to policyholders is good enough to allocate it intelligently. The MGAs, carriers, reinsurers, and platforms that treat this as a capital flow problem will continue to build dashboards and portals. The ones that treat it as an information flow problem will build the infrastructure that defines the next decade.
I have spent my career inside this system, and what I am most convinced of is that the next wave of winners will not be the ones with the biggest balance sheets or the longest track records. They will be the ones who understand that capacity is a relationship, that relationships depend on trust, that trust depends on information, and that the quality of information is now a solvable engineering problem for the first time in the history of this industry.
That is a very different insurance market than the one most of us came up in. It is also, in my view, a significantly better one.
Fabio Faschi is an Insurance leader, Founder of Scholarus AI and Hogglet.com, National Producer, Board Member of the Young Risk Professionals New York City chapter and Committee Chair at RISE with over a decade of experience in the insurance industry. He has built and scaled over a dozen national brokerages and SaaS-driven insurance platforms. Fabio's expertise has been featured in publications like Forbes, Consumer Affairs, Realtor.com, Apartment Therapy, SFGATE, Bankrate, and Lifehacker.
