Reinsurance changes who absorbs loss, but it does not automatically change who owes the policyholder

The first correction is the most important one. Reinsurance is not a side promise sold to the customer. It is a behind-the-scenes contract between insurance companies. The NAIC says the cedent transfers risk to the reinsurer, and that the reinsurer’s obligation arises only when the original insurer’s liability has been incurred. IFRS 17 makes the same economic point from an accounting angle: the holder of reinsurance does not normally get to reduce what it owes the policyholder just because it expects recovery from a reinsurer.

That means ordinary indemnity reinsurance usually changes the insurer’s backend economics, not the policyholder’s direct legal counterparty. The customer still faces the original insurer unless the deal is assumption reinsurance or another structure that actually transfers the policy relationship.

The money path is therefore not just premium in, claim out. It is usually premium received by the insurer → reinsurance premium or treaty settlement sent to a reinsurer → gross liability still measured at the insurer → reinsurance recoverable or reserve credit recognized if the treaty qualifies → reinsurer may retrocede part of the risk again → final net result depends on collectibility, collateral, treaty wording, and capital support.

That is why reinsurance belongs in a “follow the money” series. It changes capacity, volatility, reserving presentation, liquidity strain, and group capital — but only if real risk moves in a way the books and the supervisor will recognize.

Modern reinsurance mechanics were built in layers: contract law, solvency law, and group supervision

Not all reinsurance structures move money the same way

Follow one insurance dollar through the reinsurance machine

The balance-sheet bridge for a reinsured dollar

The table below is simplified, but it matches the direction of the statutory, accounting, and prudential source material.

Credit for reinsurance is the hinge between a treaty and actual accounting relief

The U.S. rule is blunt. Model Law #785 says credit for reinsurance is allowed to a domestic ceding insurer as either an asset or a reduction from liability only when the legal and regulatory conditions are met. That is why the public statements spend so much space on authorized status, certified status, reciprocal jurisdictions, collateral, funds held, and overdue balances.

The NAIC reinsurance page also shows why this matters in practice. If the cedent uses a licensed or otherwise qualifying reinsurer, the cedent may receive more favorable credit treatment. If the cedent uses an unauthorized reinsurer, collateral questions become central unless the reinsurer qualifies through another recognized route. The annual statement then forces those issues into the open: P&C Schedule F tracks provisions for unauthorized reinsurance, overdue reinsurance, collateral deficiency, and total provision for reinsurance; life and health Schedule S track unauthorized, certified, and restated net-credit positions as well.

This is one of the most important “follow the money” lessons in the whole industry. A ceded liability is not the same thing as collected cash. It is first a legal entitlement under a treaty, then an accounting position, and finally a solvency question about whether the cedent can really rely on that support when stress hits.

Funds withheld and modified coinsurance show why ceded risk and supporting assets do not always travel together

Life and some health reinsurance get much harder to read once funds withheld and modified coinsurance enter the picture. Issue Paper No. 74 explains the basic idea: under a coinsurance with funds withheld treaty, the reinsurer establishes reserve liability on its share of the reinsured policies, but the ceding company withholds assets from the reinsurer to offset future obligations. Current NAIC reporting materials say that when the mean reserve changes, money plus treaty interest moves between cedent and reinsurer according to the agreement.

The plain-English consequence is important. Risk can be ceded while the supporting assets remain on, or economically tied to, the ceding company’s side of the structure. That can make the headline story sound cleaner than the real operating story. It also raises questions about who controls the investments, who bears investment risk, how restrictions are disclosed, and whether affiliated or related-party concentrations are hiding inside the ceded structure.

That is exactly why NAIC added Schedule S Part 8 for life and similar reporting changes for other blanks. The regulators were not adding a decorative schedule. They were reacting to a real transparency problem: risk and assets can be split in ways that older public schedules did not show clearly enough.

Group structure matters because reinsurance can move strain across entities without moving it out of the enterprise

One of the easiest mistakes in insurance analysis is to stop at the legal entity that issued the policy. That is not enough. Model Act #440 defines enterprise risk as any activity, circumstance, or event involving one or more affiliates that could materially harm the insurer or the holding company system. Form F then requires the ultimate controlling person to report those risks at the system level. ORSA adds the same message in risk-management language: the group must assess solvency and capital from a group-level perspective, not only one entity at a time.

The global frameworks say the same thing more formally. Solvency II requires group eligible own funds to remain at least equal to the group Solvency Capital Requirement, and it designates a group supervisor to coordinate the system. ComFrame goes even further by requiring a group-wide strategy for reinsurance and other forms of risk transfer that covers gross and net retention, reinsurer credit risk, and the mix of traditional and alternative risk transfer.

So when a block is ceded to an affiliate, the right question is not “Did risk leave this balance sheet?” The right question is “Where did it go next inside the group, and what happened to capital, liquidity, and concentration after it moved?”

Retrocession is reinsurance for reinsurers, and it adds a second hidden chain behind the first one

The NAIC says reinsurers may also buy reinsurance protection, and that this is called retrocession. The basic reason is simple: even a reinsurer can become too concentrated in catastrophe, mortality, longevity, casualty-tail, or counterparty risk if it keeps everything it assumed. Retrocession is the reinsurer’s own version of capacity and volatility management.

IAIS ICP 13 also reminds you that reinsurance is broader than traditional treaties. It includes other forms of risk transfer such as capital-markets structures. That matters because some peak risks do not stop at one cedent-to-reinsurer handoff. They can keep moving through retrocessionaires, special-purpose vehicles, or capital-markets investors.

But each extra layer adds another thing that can go wrong. EIOPA’s 2024 supervisory statement says it is important to assess the actual risk mitigation taking place, especially where third-country reinsurance is involved. That is regulator language for a simple truth: a tower of treaties is only useful if the protection still works when you need to collect.

When reinsurance stops being true transfer and starts looking like financing

This is where the industry gets most interesting. Not every contract that calls itself reinsurance gets full reinsurance accounting. Issue Paper No. 104 says contracts that do not transfer both underwriting risk and timing risk are handled through deposit accounting in the P&C context. Issue Paper No. 162 then sharpens the point by saying accounting credit for contracts that pass risk transfer is only for the amount of risk ceded.

Life and health have their own version of the same problem. The A-791 framework and the 2025 working materials say a ceding insurer may not reduce liabilities or establish an asset if, in substance, the cedent has to reimburse the reinsurer for negative experience or the treaty contains other features that let the reinsurer claw the risk back. That is why current NAIC work keeps revisiting YRT combinations, experience refunds, and risk-limiting features. The label on the cover page is not enough.

The plain-English test is simple. If the cedent still carries most of the downside through side agreements, forced recapture, excessive repricing, negative experience reimbursement, or thin-risk windows, the supervisor may treat the deal more like financing than real transfer.

What commonly goes wrong when people explain reinsurance too casually

What can be directly observed, what can be calculated, and what remains unknown

Plain-English glossary