Reinsurance Explained: How Insurance Companies Buy Insurance
Book: Structured Finance and Insurance: The ART of Managing Capital and Risk Author: Christopher L. Culp Publisher: Wiley Finance, 2006 ISBN: 978-0-471-70631-1
Chapter 9 is about reinsurance, and it is one of those topics that sounds dry until you realize how central it is to the entire financial system. This is still Part 2, Traditional Risk Transfer. Insurance companies rarely keep all the risk they underwrite. Some gets diversified away across policyholders and lines. The rest gets transferred to other firms through reinsurance.
When an insurance company buys insurance, that is reinsurance. Simple as that.
The Basic Vocabulary
The terminology is its own little world:
- Cedant: The insurance company buying reinsurance (ceding risk outward).
- Cession: The outward transfer of risk from cedant to reinsurer.
- Assumption: The inward transfer of risk by the reinsurer.
When a reinsurer buys insurance on a reinsured risk, it is called a retrocession. The reinsurer buying is the retrocedant, and the firm assuming the risk is the retrocessionaire. The retrocession market is less liquid and significantly more expensive than the primary reinsurance market.
The big reinsurers are well-capitalized, highly rated, and globally diversified. Companies like Munich Re, Swiss Re, and Berkshire Hathaway Reinsurance dominate. Their profitability is measured by the combined ratio (net losses plus underwriting expenses divided by net premium earned). A combined ratio below 100% means the reinsurer is profitable on its underwriting alone, before investment income.
Three Risks of Writing Insurance
Before understanding why reinsurance exists, you need to understand what can go wrong for a primary insurer:
Underwriting risk: Claims on a policy line exceed total premium collected. Either the pure premium was calculated wrong (average claim is higher than expected) or individual claims come in way above average.
Investment risk: The assets backing reserves decline in value before claims are filed. You set aside the right amount of money, but you invested it badly. If your reserve is in hedge fund shares and the fund tanks, you cannot pay claims.
Timing risk: Claims arrive earlier than expected. Even if you priced correctly and invested in safe assets, early claims cause losses. Culp gives a clean example: if you discount a $100 expected claim at 5% and collect $95 in premium expecting the claim in 12 months, a claim the very next day costs you $5. Not underwriting risk (the expected claim was right). Not investment risk (safe assets held their value). Pure timing.
Why Buy Reinsurance
In an M&M world, reinsurance would be pointless. But capital markets are imperfect, so:
Increased underwriting capacity. By transferring risk, the cedant can write more policies than its own capital would support. This works along two dimensions. Large-line capacity is the ability to absorb a catastrophic loss on a single policy. Premium capacity is the ability to write a large volume of policies. Reinsurance is essentially “renting the balance sheet” of another company as synthetic equity.
Increased debt capacity. Reinsurance is contingent capital. It has the same capital structure impact as issuing equity, which in turn increases debt capacity.
Reduced earnings volatility. When diversification within a policy line is inadequate, small underwriting losses can whip earnings around. Reinsurance provides synthetic diversification.
Reduced financial distress costs. Catastrophic protection for low-frequency, high-severity events reduces the probability that assets fall below liabilities.
Information acquisition. The reinsurance process is extremely information-intensive. Reinsurers doing due diligence learn a lot about the cedant, the market, and competitors. This can benefit the cedant too, because a well-informed capital provider means lower adverse selection costs.
Run-off solutions. Want to exit a business line without terminating all outstanding policies? Buy reinsurance on the whole book and the business effectively disappears from your balance sheet overnight.
Facultative vs. Treaty Reinsurance
Facultative reinsurance covers a single policy. Each deal is separately negotiated. It is flexible and customizable but expensive and time-consuming. Used for very large, unusual, or exotic risks.
Treaty reinsurance covers a group of policies that meet predefined parameters. The reinsurer cannot refuse any individual policy that falls within the treaty guidelines. This is the workhorse of the reinsurance industry, used for reinsuring entire business lines.
Facultative deals have bigger moral hazard and adverse selection problems because the cedant can cherry-pick which risks to cede. Treaty deals spread the risk more broadly.
Proportional Reinsurance Treaties
In proportional reinsurance, risk, premium, and losses are shared between cedant and reinsurer based on a ratio.
Quota Share Treaties (QST)
A quota share treaty uses a fixed percentage. If the split is 75/25, the cedant keeps 75% of premium and pays 75% of every claim. The reinsurer gets 25% of premium and pays 25% of every claim. The reinsurer also pays a ceding commission to compensate the cedant for originating the business.
Important detail: the sharing is per dollar of claims, not based on the order claims are filed. A $75,000 claim on a 75/25 QST does not mean the cedant pays everything. The cedant pays $56,250 and the reinsurer pays $18,750.
QSTs are good for increasing capacity and reducing earnings volatility. Two primary carriers can even swap portions of their portfolios through a reciprocity treaty to improve diversification.
Surplus Share Treaties (SST)
In a surplus share treaty, the cedant sets a fixed dollar retention for each policy, and anything above that goes to the reinsurer. If the retention is $75,000, small policies are fully retained by the cedant. A policy with a $100,000 limit is split 75/25. A policy with a $150,000 limit is split 50/50.
SSTs create large-line capacity but give less UPR relief than QSTs. They also have bigger adverse selection problems because the cedant chooses the retention per policy and can cherry-pick what to cede.
Excess of Loss (XOL) Reinsurance
This is where it gets really good. In XOL reinsurance, the reinsurer pays only when losses exceed a certain amount. Small losses stay with the cedant.
The notation is “$A XS $Y” meaning the reinsurer covers up to $A in losses, but only after the cedant has already absorbed $Y. The lower attachment point is $Y. The upper attachment point is $A + $Y.
Vertically Layered Programs
Multiple reinsurers can cover different loss layers. Company A covers $1M XS $2M. Company B covers $5M XS $3M. Company C covers $2M XS $8M. Each reinsurer’s liability depends on the total cumulative losses. The highest layer is called the catastrophic layer because of the extremely low probability of losses reaching it.
Per-Occurrence vs. Aggregate
XOL can be structured per-occurrence or in aggregate. A per-occurrence treaty pays per individual event. An aggregate treaty covers cumulative losses across all events in a year.
This distinction was at the heart of a major legal dispute after September 11, 2001. Swiss Re, Chubb, and Allianz argued the attacks should be classified as a single occurrence (one terrorist attack) rather than two (separate planes hitting separate buildings). For Swiss Re, the difference was between $3.6 billion and $7.2 billion in liability. The court ruled it was a single occurrence.
Aggregate XOL / Stop Loss Treaties
These cover accumulated small losses that individually fall below per-occurrence thresholds. If a homeowner’s insurer has per-occurrence coverage at $125,000 and gets 10 claims of $100,000 each, no single claim hits the per-occurrence treaty. But an aggregate $500K XS $500K treaty would kick in after the fifth claim. Aggregate treaties are very effective for smoothing earnings but they are expensive and usually include significant co-insurance provisions.
Horizontal Layering and Blended Cover
In a horizontally layered program, multiple reinsurers share the same loss layer, splitting costs. A blended cover combines vertical and horizontal layering. For example, two reinsurers split claims from $2M to $8M 50/50, and a third reinsurer covers the catastrophic layer above $8M alone.
Syndication
Big reinsurance deals are often syndicated. A lead underwriter negotiates the treaty and then shares the risk across syndicate members. This is different from a fronting arrangement where one reinsurer takes the whole deal and then retrocedes most of it.
The credit risk is different. In a syndicated deal, the cedant is exposed to all five syndicate members individually. If one cannot pay, the cedant loses that share. In a fronting deal, the cedant has credit exposure only to the fronting reinsurer. If one of the retrocessionaires behind the scenes fails, the fronting reinsurer still owes the full amount.
The pricing differs too. Retrocession costs are typically higher than direct reinsurance rates, and those costs get passed to the cedant in the fronting arrangement. But if the fronting reinsurer is AAA-rated and the syndicate members are not, the cedant might happily pay extra for the better credit quality.
My Take
This chapter is dense but rewarding. The XOL notation and layering structures are the building blocks for understanding structured finance products later in the book. If you can picture a vertically layered XOL program with attachment points, you are already halfway to understanding CDO tranching.
The 9/11 example is a powerful illustration of how much the definition of a single “occurrence” matters. A seemingly academic distinction about event definitions translated into billions of dollars of liability.
And the distinction between syndication and fronting is the kind of practical detail that matters enormously in practice but rarely makes it into textbooks. Credit risk is not just about whether your counterparty can pay. It is about how many counterparties you are exposed to and how that exposure is structured.