Chapter 27: Reinsurance: How Insurers Insure Themselves and Why It Matters for Underwriting

"There is no such thing as a risk that is too large to write — only a risk that is too large to keep." — a maxim of the reinsurance market [constructed teaching line, in the spirit of the trade]

Overview

You have spent twenty-six chapters learning to underwrite one risk at a time: read it, grade it, price it, structure the terms, decide. Now look up from the file. The \$20 million building you are about to put on Harbor Steel's property policy is not, in any meaningful sense, your company's \$20 million. Long before you signed the quote, your company decided how much of any single fire or any single hurricane it was willing to keep on its own balance sheet — its retention — and arranged for someone else to absorb the rest. That someone is a reinsurer, and the contracts that move the surplus risk off your books shape almost every decision you make at the desk: how large a line you are allowed to write, what a catastrophe-exposed account costs you, whether you need to make a phone call before you can bind the full \$20 million, and how the combined ratio you are judged on is actually calculated. Reinsurance is the insurance behind the insurance, and the primary underwriter who does not understand it is flying with half the instruments covered.

The reason reinsurance exists is the same reason insurance exists, applied one level up. A primary insurer faces exactly the problems its policyholders face — losses it cannot perfectly predict, the occasional event far larger than any single year's premium, and a finite amount of capital standing behind a great many promises. So the insurer does what its insureds do: it transfers the part of the risk it cannot comfortably bear to a party that can, in exchange for a premium. But it does so with a sophistication the retail customer never sees — slicing risk into proportions and layers, ceding whole classes of business by treaty or one tricky risk at a time, buying protection that triggers only when a single storm hits a thousand policies at once. Get this machinery right and a mid-size regional carrier can safely write a \$20 million account it could never hold alone. Get it wrong — keep too much, buy too little, buy from a reinsurer that cannot pay — and one bad season is the last season.

This chapter builds reinsurance from the underwriter's seat outward. We start with what reinsurance is and the three jobs it does. We separate the two ways risk is ceded — by treaty and facultatively — and the two ways it is shared — proportionally and non-proportionally. We follow the money through quota share, surplus share, excess of loss, and the catastrophe cover that protects the whole book at once. Then we do the work that matters most to you: tracing how all of this changes net-versus-gross underwriting, what you are allowed to write, and what it costs. We end at the global market that carries this risk and at the hard limit of passing risk on — retrocession, and the spiral that taught the market its most expensive lesson.

In this chapter, you will learn to:

• Define reinsurance and explain the three problems it solves: capacity, catastrophe protection, and the smoothing of volatile earnings.
• Distinguish a treaty from a facultative placement and know when the treaty covers you and when you must arrange cover one risk at a time.
• Compare proportional reinsurance (quota share, surplus share) with non-proportional reinsurance (excess of loss, catastrophe XOL), and say what each does to your net account.
• Trace how the reinsurance program sets your retention, your capacity, your ceding commission, and therefore your price and your pen.
• Read a reinsurance tower, and understand retrocession and the limits of passing risk on.

Learning Paths

This is a Part V chapter: it lifts the lens from the single file to the enterprise, and every track needs it, but for different reasons.

🏠 Personal Lines: Reinsurance is invisible to your insureds but it is why a homeowners book in a hurricane state can exist at all. Weight §27.1 (why it exists) and §27.4 (catastrophe XOL) — the cat treaty is the thing standing between one storm and your company's solvency. 🏢 Commercial Lines: Weight §27.2 (treaty vs. facultative) and §27.5 (net vs. gross). The large, catastrophe-exposed account — Harbor Steel exactly — is where you will reach the edge of the treaty and have to think about a facultative placement and your net line. 📊 Analytics: Weight §27.3 and §27.4. Proportional and non-proportional structures are just different functions applied to the loss distribution; understanding how a layer transforms frequency and severity is the foundation for pricing a treaty or modeling a net book. 📜 Certification: All of it. Treaty vs. facultative, proportional vs. non-proportional, quota share, surplus share, XOL, cat XOL, ceding commission, and retrocession are core CPCU/ARe reinsurance vocabulary and recur on every exam in the subject.

27.1 What reinsurance is and why it exists (capacity, catastrophe, volatility)

Start, as always, with the decision on the desk. Your company is a mid-size regional carrier. A broker sends you a clean, attractive account that happens to need a \$20 million property limit. Your underwriting guidelines say your personal maximum line on a single property risk is, let us say, \$5 million. Do you decline a good risk because it is too big? Turn away the next one at \$8 million? If every carrier could only write risks up to the size it could comfortably absorb alone, the insurance market would be a patchwork of small bets and most large enterprises would go uninsured. Reinsurance is what makes the large risk writable by the small carrier. It is the mechanism that lets your company say yes to the \$20 million account while keeping only the \$5 million it can actually afford to lose.

So we can define it precisely. Reinsurance is insurance purchased by an insurer: a contract under which one insurer (the reinsurer) agrees, for a premium, to indemnify another insurer (the ceding company) for all or part of the losses the ceding company incurs under the policies it has issued. The insurer that buys the protection and transfers — cedes — the risk is the ceding company, or cedent. Nothing about the original policyholder's contract changes; Harbor Steel never learns that part of its risk has been moved. The reinsurance contract is a separate agreement, between two insurers, sitting entirely behind the policy. This is the doctrine of privity: the reinsurer has no contractual relationship with the original insured, and the cedent remains fully liable to its policyholder whether or not the reinsurer pays. (That last clause is not a technicality. It is why the financial strength of your reinsurer — collectability — is itself an underwriting question, which we return to in §27.6.)

Reinsurance does three jobs, and naming them keeps you honest about why a given treaty exists.

THE THREE JOBS OF REINSURANCE                                  [constructed teaching example]

  CAPACITY      ─►  lets a carrier write limits larger than it could hold alone.
                    "I can offer $20M because I keep $5M and cede $15M."

  CATASTROPHE   ─►  protects against one event striking many policies at once —
                    the failure of the independence assumption (Ch. 1, §1.2).
                    "One hurricane will hit 4,000 of my homes; the cat treaty pays
                     above my retention so a single storm can't take the company."

  VOLATILITY    ─►  smooths year-to-year earnings and protects surplus, so a bad
                    year is survivable and the combined ratio is less wild.
                    "I'd rather give up some expected profit to a reinsurer than risk
                     a single year that wipes out a decade of it."

The first job, capacity, is the one a line underwriter feels most directly. Your authority to write a \$20 million limit exists only because the reinsurance program stands behind it; the program is, in a real sense, the source of your capacity. The second, catastrophe protection, addresses the single deepest flaw in the insurance model. Recall from Chapter 1 (§1.2) that the law of large numbers works only when losses are independent — when one house burning tells you nothing about whether another will. Catastrophe destroys that assumption: a hurricane does not burn one home in the pool, it floods ten thousand at once. No amount of writing "more, similar business" diversifies away a peril that strikes the whole pool simultaneously. Reinsurance — specifically the catastrophe cover of §27.4 — is how a primary insurer survives the correlated event that its own pooling cannot absorb. The third job, volatility management, is the subtlest and the most strategic: even short of a true catastrophe, an insurer's results swing from year to year, and surplus (the capital that backs every promise — Chapter 28 owns that term) is finite and expensive. By ceding the volatile tail of its results, an insurer trades away some expected profit for a steadier, more survivable, more financeable book.

📋 At the Desk Reinsurance is not free, and that is the whole point of the discipline. Every dollar of risk you cede, you pay for — and the reinsurer, on average and over time, expects to make money on the deal. So the cedent is systematically giving away expected profit in exchange for reduced variance. That is exactly the trade the original policyholder makes when they buy insurance (Ch. 1, §1.1), now run one level up: the insurer is risk-averse about its own surplus the way a homeowner is risk-averse about their house. The art is buying enough protection to survive the bad year and protect the capital, without ceding away so much margin that you are working to enrich your reinsurer. An insurer that reinsures too little gambles its solvency; an insurer that reinsures too much hands its profit to Bermuda. Neither is good underwriting management.

This framing connects to a theme that runs through the whole book: the combined ratio tells the truth. Reinsurance changes the combined ratio you are measured on, because it changes both the premium you keep and the losses you retain. A cession that protects you in the catastrophe year costs you premium in every quiet year — and the multi-year question of whether your reinsurance program is well-bought is, in the end, a combined-ratio question across the cycle, not a single-year one. Hold that thought; §27.5 makes it concrete.

27.2 Treaty vs. facultative

There are two fundamentally different ways an insurer arranges reinsurance, and the distinction governs your daily workflow more than any other idea in this chapter. The difference is simply when the decision is made and how many risks it covers at once.

Treaty reinsurance is an agreement covering a whole class or book of the cedent's business, negotiated in advance. Under a treaty, the reinsurer agrees to accept — and is obligated to accept — every risk that falls within the treaty's defined scope, and the cedent is obligated to cede it, automatically, without the reinsurer's case-by-case approval. The two parties argue once, at the annual renewal, over the terms, the price, and the scope; thereafter, every qualifying policy the cedent writes is reinsured the moment it is bound, with no further conversation. Facultative reinsurance, by contrast, is arranged one risk at a time. "Facultative" means optional — for both sides. The cedent chooses whether to offer a particular risk; the reinsurer evaluates that specific submission and chooses whether to accept it and on what terms. Each facultative placement is a small, bespoke underwriting transaction of its own.

TREATY vs. FACULTATIVE — the two ways to cede                  [constructed teaching example]

                    TREATY                          FACULTATIVE
  scope        a whole class/book              one specific risk
  obligation   automatic — cedent must cede,   optional — both sides choose
               reinsurer must accept           each risk individually
  timing       negotiated once, at renewal     negotiated per submission
  speed        instant (bound = ceded)         slow (must place before/while binding)
  best for     the bulk of ordinary business   the large, unusual, or over-treaty risk
  the UW feels it as capacity that's            it as a phone call you must make
               already there                    before you can bind the full line

For the working underwriter, the practical rule is this: the treaty is your default capacity, and you reach for facultative when the treaty runs out or won't respond. Most of what you write is covered the instant you bind it, because it falls inside the treaty your company negotiated months ago — you never think about it. You reach for a facultative placement in three situations: when a risk is larger than the treaty will hold (a \$20 million line where the treaty's per-risk capacity tops out lower); when a risk is excluded from the treaty (a class or peril the treaty carves out); or when a risk is so unusual or heavily exposed that prudent management wants a second underwriter — the facultative reinsurer — to look at it and share it. Facultative is slower and more expensive per unit of cover, and it puts a condition between you and your bind: you may not be able to confirm the full limit until the facultative reinsurer says yes.

⚠️ Underwriting Trap The trap is binding the full limit before the facultative cover is confirmed — issuing a \$20 million quote, having the broker accept, and only then discovering the facultative reinsurer won't take its share, or will only at a price that destroys your economics. Now your company is net on the whole \$20 million, or you must retract terms you already offered and damage the broker relationship. The disciplined move is to make the facultative placement a subjectivity (Chapter 13 owns that term) — a condition precedent to binding the full line — or to confirm the facultative before you firm up the quote. "Fac to follow" is a phrase that has cost careless underwriters their year. The reinsurance has to be in place, not intended, before the cedent is on the hook for the gross limit.

There is a hybrid worth knowing because it appears constantly: the facultative-obligatory ("fac-oblig") treaty, under which the cedent chooses whether to cede a given risk but the reinsurer is obligated to accept whatever the cedent sends. It gives the cedent flexibility and the reinsurer none, which is why reinsurers grant it sparingly and watch the ceded business closely for adverse selection — the cedent will be tempted to keep its best risks net and cede its worst, which is the same adverse-selection problem (Chapter 1, §1.4) the cedent's own underwriters fight every day, now turned against the reinsurer. Adverse selection, you will notice, does not disappear when you move up a level. It just changes seats.

27.3 Proportional reinsurance: quota share and surplus share

Having chosen how to cede (treaty or facultative), the insurer must choose what shape the cession takes. The first family is proportional reinsurance, in which the cedent and the reinsurer share premiums and losses in the same fixed proportion. The reinsurer takes an agreed percentage of the risk; in exchange it receives that same percentage of the premium and pays that same percentage of every loss, dollar for dollar, from the first dollar. Proportional reinsurance is a partnership in the outcome: if the reinsurer is on 40% of a risk, it gets 40% of the premium and pays 40% of every claim, large or small. There are two main forms.

Quota share is the simplest reinsurance arrangement that exists: the cedent cedes a fixed percentage of every risk in the covered book, and the reinsurer takes that same percentage of all premiums and pays it of all losses. A 30% quota share on a property book means the reinsurer is on 30% of every policy — 30% of the premium in, 30% of every loss out — automatically, across the whole book.

QUOTA SHARE — a 30% cession on three risks                    [constructed teaching example]

  risk     limit     cedent keeps (70%)   reinsurer takes (30%)
  A        $1.0M         $0.70M                $0.30M
  B        $3.0M         $2.10M                $0.90M
  C        $5.0M         $3.50M                $1.50M

  Premiums split the SAME way (reinsurer gets 30% of every premium).
  A $400,000 loss on risk B:  cedent pays $280,000, reinsurer pays $120,000.
  Note the weakness: on the SMALL risk A, the cedent gave away 30% it could
  easily have kept; on the LARGE risk C it still keeps $3.5M net.

Quota share is blunt but powerful. It is easy to administer, it shares the cedent's results faithfully (good and bad), and — critically — it relieves the strain on the cedent's surplus, because ceding a share of the premium written reduces the premium-to-surplus leverage (Chapter 28 develops that lever). A young or fast- growing insurer that is writing more business than its capital comfortably supports will often buy quota share precisely to fund growth: the reinsurer's capital, in effect, stands behind the ceded share of the book. Its weakness is visible in the figure: quota share treats the \$1 million risk and the \$5 million risk identically, ceding the same percentage of each, so the cedent gives away premium on small risks it could easily have kept net while still retaining a large net line on the big risks that actually threaten it. Quota share manages surplus strain and growth; it does not, by itself, solve the large-line problem.

Surplus share fixes exactly that. A surplus-share treaty cedes only the part of each risk above the cedent's chosen retention, expressed in multiples of that retention called lines. The cedent sets a retention — say \$1 million — and the treaty provides some number of lines of capacity above it. On a risk at or below the retention, the cedent cedes nothing and keeps it all. On a larger risk, the cedent keeps its retention and cedes the surplus above it, and premiums and losses on that risk are split in the same proportion as the division of the limit.

SURPLUS SHARE — $1M retention, 4-line treaty ($4M of cession capacity)   [constructed teaching example]

  risk    limit    cedent retains   ceded (surplus)   cession %    so reinsurer pays this % of losses
  A       $0.8M       $0.8M             $0               0%          0%   (under retention; all net)
  B       $2.0M       $1.0M             $1.0M           50%          50%
  C       $5.0M       $1.0M             $4.0M           80%          80%
  D       $7.0M       $1.0M             $4.0M           ~57% of $7M   ... and $2M is STILL net or fac'd!

  Capacity = retention + lines = $1M + 4×$1M = $5M. Risk D's limit ($7M) exceeds the
  treaty's $5M, so the $2M excess needs another treaty layer or a facultative placement.

Notice what surplus share buys you that quota share does not: it lets the cedent keep its full retention on every risk and cede only the part that is genuinely too big. The small risk A is kept entirely; the cedent gives away nothing it could afford. The large risk C is ceded 80%, so the cedent's net exposure on a \$5 million risk is only its \$1 million retention. Surplus share, in other words, homogenizes the net book — after cession, the cedent's retained line on every risk is at or below \$1 million, so the net account behaves like a book of similar-sized risks, which is exactly what the law of large numbers wants (Chapter 1, §1.2). The price of this flexibility is administrative complexity: every risk has its own cession percentage, which must be tracked and applied to every premium and every loss. And note risk D: a \$7 million risk under a \$5 million-capacity treaty leaves \$2 million with nowhere to go — back to the cedent's net account, or out to facultative. Surplus share extends your capacity; it does not make it infinite.

📊 Model vs. Judgment A model can compute the optimal quota-share percentage or surplus-share retention that minimizes the variance of the net result for a given cost — and you should let it, because the arithmetic of variance reduction is exactly what models do well. What the model cannot decide is the strategic question behind the structure: is your company buying quota share to fund growth (a capital decision), to share the uncertainty of a new line you don't trust your own pricing on (a confidence decision), or to manage a specific accumulation (a portfolio decision — Chapter 29)? The same 30% cession can be the right answer to one of those questions and the wrong answer to another. The reinsurance structure is a financial tool, but which problem you are solving with it is a judgment about the business, and that is management's call, not the optimizer's.

A word on the most important number in proportional reinsurance from the cedent's side: the ceding commission. Because the cedent incurred the costs of acquiring and servicing the business — the broker's commission, the underwriting, the policy issuance — the reinsurer, which simply takes a share of an already- written book, pays the cedent a ceding commission: an allowance, expressed as a percentage of the ceded premium, that reimburses the cedent for those acquisition and overhead expenses (and, in a soft reinsurance market, sometimes a bit more, as the reinsurer competes for the business). Ceding commission is the allowance a reinsurer pays the cedent on proportional business to reimburse the cedent's acquisition and overhead expenses on the ceded premium. It exists only in proportional reinsurance — because only there is the reinsurer sharing the original premium and therefore the original expenses. It is also a negotiating lever: a higher ceding commission improves the cedent's net expense ratio and can make a proportional treaty a source of fee-like income, while a "sliding-scale" commission that moves with the loss ratio aligns both parties' incentives. When you hear a treaty underwriter argue about "the cede," this is the number.

27.4 Non-proportional: excess of loss and catastrophe XOL

Proportional reinsurance shares every dollar from the first one. Non-proportional reinsurance is built on a completely different principle: the reinsurer pays nothing until a loss exceeds an agreed threshold, and then pays the part above that threshold, up to a limit. The cedent and reinsurer no longer share in fixed proportion; instead the cedent keeps everything up to its retention and the reinsurer absorbs the excess. This is the reinsurance form that most resembles a deductible-and-limit structure on a primary policy (Chapter 12) — only here the insurer is the one buying the protection.

Excess of loss (XOL) is the workhorse non-proportional form: the reinsurer pays the amount of a loss above the cedent's retention (often called the attachment point or priority), up to the reinsurer's limit. XOL is written in layers, each described as "limit excess of retention" — for example, "\$4 million excess of \$1 million," meaning the reinsurer pays losses above \$1 million, up to \$4 million more (so it covers the band from \$1 million to \$5 million; a loss of \$5 million or more produces the full \$4 million recovery, and the cedent keeps the first \$1 million plus anything above \$5 million unless a higher layer responds).

A REINSURANCE TOWER — per-risk excess of loss, three layers     [constructed teaching example]

   $20M ┤                                            (above the program: NET or facultative)
        │  ┌───────────────────────────────────┐
        │  │   3rd layer:  $10M xs $10M         │   reinsurer C pays $10M→$20M
   $10M ┤  ├───────────────────────────────────┤
        │  │   2nd layer:  $5M xs $5M           │   reinsurer B pays $5M→$10M
    $5M ┤  ├───────────────────────────────────┤
        │  │   1st layer:  $4M xs $1M           │   reinsurer A pays $1M→$5M
    $1M ┤  ├───────────────────────────────────┤
        │  │   RETENTION:  cedent keeps $0→$1M  │   the cedent's net, every loss
    $0  ┴  └───────────────────────────────────┘

  A $7M loss is paid:  cedent $1M  +  layer1 $4M  +  layer2 $2M  =  $7M.  Layer 3 untouched.
  A $0.6M loss is paid entirely by the cedent — below the attachment, no recovery.

Read the tower from the bottom up, because that is how a loss climbs it. Every loss starts in the cedent's retention — the cedent always pays the first dollar up to its attachment point. A loss that stays small never reaches the reinsurance at all; the cedent eats it net. A loss that pierces the attachment point spills into the first layer, and if it is large enough, into the second and third. This is the single most important diagram in reinsurance, and you should be able to take any loss figure and walk it up the tower: who pays what, and where it stops. Note the structural consequence for the cedent: XOL caps the cedent's loss per event at the retention (plus anything above the top of the program), which is precisely the protection against severity that proportional reinsurance does not give. The cedent keeps all the small losses (which it can afford and which it would rather not pay a reinsurer to handle) and offloads only the large ones (which threaten it). XOL is severity protection; quota share is not.

Two mechanical features of XOL change how you read it. First, reinstatement: an XOL layer typically provides its limit a limited number of times per year. After a loss exhausts (or partly exhausts) a layer, the cedent usually pays a reinstatement premium to restore the cover for the rest of the term — and a layer may be "one reinstatement," meaning it can pay twice in a year and no more. In an active catastrophe season, running out of reinstatements is a real and frightening exposure: the second or third storm can arrive with the cat cover already spent. Second, the retention is per-occurrence, so a year of many medium losses, each below the attachment, can hurt the cedent badly even though no single loss ever reached the reinsurance — the cedent retains the frequency even when it is protected against severity. An underwriter who buys only high-attaching XOL is protected against the one huge loss but fully exposed to a bad frequency year.

📋 At the Desk The single most consequential reinsurance an underwriter on a catastrophe-exposed book lives inside is the catastrophe XOL (cat XOL): an excess-of-loss treaty that responds not to a single large risk but to a single large event — one hurricane, one earthquake, one wildfire — that strikes many policies at once. Where per-risk XOL attaches above a single policy's loss, cat XOL attaches above the aggregate loss to the cedent's whole book from one catastrophe occurrence. The cedent retains the first slice of any catastrophe (say the first \$10 million of event loss), and the cat tower pays the layers above it up to the program's top. This is the instrument that addresses the failure of independence (Ch. 1, §1.2) head-on: it is bought not because any single account is too big, but because one event correlates thousands of accounts at once. The size of the program — how high the tower goes — is driven by the catastrophe model's estimate of the probable maximum loss (Chapter 30 owns PML and the return-period thinking); the cedent buys cover up to some return period (a "1-in-100-year" or "1-in-250-year" event) and retains, or goes bankrupt on, anything above it. When you place a coastal account like Harbor Steel into the book, you are consuming some of that cat tower's capacity, and that consumption is a real cost the price must reflect.

🔍 Check Your Understanding 1. A cedent has a per-risk XOL tower: retention \$1M, then \$4M xs \$1M, then \$5M xs \$5M. A single fire causes a \$6.5M loss. Walk the loss up the tower: who pays what? 2. Why does a quota share protect a young insurer's surplus while an excess-of-loss treaty does not — and why does XOL protect against a single catastrophic loss while quota share only shares it? 3. Why is "running out of reinstatements" a more frightening phrase on a catastrophe XOL than on a per-risk XOL?

27.5 How reinsurance shapes net vs. gross underwriting

Now the payoff — the section that turns this whole chapter from background into something you use at the desk. Every figure you have learned to compute about a risk exists in two versions, and confusing them is one of the most common and dangerous errors a developing underwriter makes. The gross position is the risk as written — the full \$20 million limit, the full premium, the full potential loss, before any reinsurance. The net position is what your company actually keeps after cessions — the retained limit, the premium net of reinsurance cost, the loss net of recoveries. Gross is what you offered the insured; net is what your company is actually on the hook for. Both matter, and they answer different questions.

GROSS vs. NET — the same $20M property account                  [constructed teaching example]

                              GROSS (as written)     NET (what we keep)
  property limit offered          $20,000,000          $5,000,000   (retention)
  ceded to surplus-share treaty        —              $11,000,000
  ceded to facultative                 —               $4,000,000
  ──────────────────────────────────────────────────────────────
  a total fire loss of $20M:      we pay $20M gross,   we pay $5M net
                                  recover $15M from reinsurers

  Premium works the SAME way: we collect the gross premium, pay away the
  reinsurance cost, and KEEP the net premium against our NET losses.

Here is the principle, and it is worth stating flatly because so much follows from it: your capacity, your appetite, and ultimately your pen are governed by your net retention, not your gross limit. You can offer \$20 million because the reinsurance program lets you cede \$15 million of it; what your company is actually betting is the \$5 million it keeps. This reframes several decisions you have already learned to make:

• Capacity and authority. Your "line size" — the maximum you may write on one risk — is really a statement about net retention plus the treaty capacity behind it. When you reach the edge of the treaty (the \$20 million account against a treaty that holds less), you must arrange facultative cover to bring your net back down to the retention — exactly the trap of §27.2.
• Pricing. The premium you charge is a gross number, but the economics that matter to your company are net: gross premium, minus the reinsurance cost (the cat-treaty cost allocated to the account, the facultative premium, the surplus-share cession net of ceding commission), against the net losses you retain. A catastrophe-exposed account that looks adequately priced gross can be a money-loser net once the cost of the cat cover it consumes is charged against it. This is why pricing follows risk (the book's fourth theme) must mean net risk: the rate has to be adequate for what you keep after paying for the reinsurance the risk requires, not merely for the gross exposure.
• Combined ratio. The combined ratio your management is judged on is a net number — net losses and net expenses over net premium. Reinsurance changes it in both directions: in a catastrophe year, recoveries pull the net loss ratio down dramatically (that is the cover working); in every quiet year, the ceded premium with no offsetting recovery pushes the net combined ratio up relative to gross. A well-bought reinsurance program should lower the volatility of the combined ratio across the cycle even if it slightly raises its average — and judging the program means judging it across the cycle, not in any single year.

⚠️ Underwriting Trap The catastrophe-exposed account that is "profitable gross but ruinous net." A coastal property priced to a rate that looks adequate against its gross expected loss can still destroy value once you charge it for the slice of cat-treaty capacity it consumes and the net volatility it adds. Underwriters who price and measure themselves gross fall in love with premium volume that is quietly unprofitable net. The discipline — and it is the hardest discipline in a soft market, with the broker pushing for the order — is to ask of every catastrophe-exposed risk: does this account earn its keep on a NET basis, after the cost of the reinsurance it forces us to buy? If the honest answer is no, the right answer is a higher price, tighter terms, or a decline — regardless of how attractive the gross premium looks. This is the same rate-adequacy discipline you learned in Chapter 11, now stated in net terms.

There is a portfolio dimension here too, which Chapter 29 develops: reinsurance interacts with the shape of the whole book. A surplus-share treaty that homogenizes net line sizes lets the law of large numbers work on the net account; a cat treaty's retention determines how much of a single event the company keeps; the ceding commission on a quota share shows up in the net expense ratio. The underwriter who thinks only about the single file, and never about the net book the file becomes part of, is missing the level at which reinsurance actually does its work.

27.6 The reinsurance market: Lloyd's, Bermuda, and the big reinsurers

Where does all this risk actually go? Reinsurance is a global business, and the risk you cede from a single plant in Port Hadley is pooled, layered, and spread across a worldwide market of capital that exists precisely to absorb the peaks that individual insurers cannot. Knowing the geography of that market matters, because who your reinsurer is — and whether they will still be standing and solvent when the loss comes — is itself an underwriting question. Reinsurance is sold mainly through three overlapping channels.

A few facts about how this market behaves are worth carrying. First, the reinsurance market has its own underwriting cycle (Chapter 3 owns that term) — hard markets after big losses, soft markets when capital is plentiful — and it is more violent than the primary cycle. After a major catastrophe destroys reinsurer capital, reinsurance prices spike and capacity contracts; in soft years, abundant capital drives prices down. That cycle flows straight through to you: when the reinsurance market hardens, your cost of cat cover rises, which must be passed into primary prices, which is part of why a hard reinsurance market produces a hard primary market in catastrophe lines. Second, capital has increasingly entered reinsurance from outside the traditional industry — insurance-linked securities (ILS) such as catastrophe bonds, which let capital- market investors take on insurance catastrophe risk for a return uncorrelated with the stock market. This "alternative capital" has reshaped the property-cat market, and it is worth knowing it exists even though its mechanics are a topic of their own.

⚖️ Compliance Corner The most important regulatory and credit fact about reinsurance is collectability — and it sits at the heart of insurance law's good-faith framework (Chapter 4). Remember the privity point from §27.1: the cedent remains fully liable to its policyholder regardless of whether the reinsurer pays. So if your reinsurer becomes insolvent or disputes the claim, your company still owes Harbor Steel every dollar of its covered loss — and now must pay it with no recovery. This is why regulators care intensely about reinsurance security: a ceding insurer generally may only take credit for reinsurance (treat the ceded risk as off its books for capital and reserving purposes) when the reinsurer is authorized, or is unauthorized but posts collateral (a trust or letter of credit), under the NAIC's credit-for-reinsurance framework. The practical lesson for the underwriter: the financial strength rating of your reinsurer (AM Best and the other rating agencies — Chapter 3) is part of your risk. A cheap cession from a shaky reinsurer is not a bargain; it is a deferred loss. You are only as protected as your weakest reinsurer is solvent.

This is the place to advance the book's fifth theme — technology augments underwriters; it does not replace them — in the reinsurance context. The pricing of catastrophe reinsurance is among the most model-intensive activities in all of insurance: the cat models (Chapter 30) that estimate event losses are what reinsurers price against, and the structuring of ILS depends entirely on modeled probabilities. Yet the reinsurance market remains, at the top, a relationship and judgment business — the lead reinsurer who knows a cedent's book and management, the broker who places a difficult program, the underwriter who decides whether the model's view of an unmodeled or poorly-modeled peril is trustworthy. The models set the floor of the conversation; experienced judgment about the cedent, the data, and the model's blind spots decides the deal.

27.7 Retrocession and the limits of passing on risk

If an insurer can reinsure, can a reinsurer reinsure? Yes — and the answer reveals both the elegance and the hidden danger of the whole system. Retrocession is reinsurance bought by a reinsurer: a reinsurer (the retrocedent) transfers part of the risk it has assumed to another reinsurer (the retrocessionaire), for exactly the same reasons a primary insurer reinsures — to manage its own capacity, catastrophe exposure, and volatility. Retrocession lets the reinsurance market spread a single catastrophe's risk even more widely around the globe, so that no single reinsurer holds an untenable concentration. In principle, this is the pooling logic (Chapter 1) extended to its furthest reach: risk diffused across the maximum possible base of capital.

But there is a serious failure mode, and it is one of the most instructive cautionary tales in the industry. When risk is passed from insurer to reinsurer to retrocessionaire and onward, it can — if the chain is opaque and the participants are also reinsuring each other —come back around. A reinsurer can unknowingly end up retroceding risk to a chain that eventually reinsures the reinsurer's own original cession, so that the same underlying catastrophe loss circulates through the market and lands, magnified, back on the parties who thought they had passed it on. This is the spiral: a self-referential loop of retrocession in which a loss is amplified as it cycles through interconnected participants who have lost sight of the ultimate exposure they hold. The London market's excess-of-loss "LMX spiral" of the late 1980s and early 1990s is the canonical real example, and it is the subject of this chapter's second case study; a handful of catastrophe losses, amplified through a tangle of retrocession, produced losses far larger than the original events and threatened the participants who had believed themselves protected. The lesson is permanent: passing on risk is not the same as making it disappear, and a market that loses track of where its risk ultimately resides has not diversified its catastrophe exposure — it has hidden and concentrated it.

🤖 Model vs. Judgment The spiral is what happens when everyone trusts that someone else's model and someone else's underwriting has handled the risk, and no one holds the full picture. Each participant in the chain may have priced its own slice perfectly using a perfectly good model — and the system can still be catastrophically mispriced, because the models priced individual transactions while the danger lived in the connections between them, which no single model could see. This is the deepest version of the book's central tension: not "the model versus the underwriter" on one risk, but the limits of any model when the real exposure is systemic, opaque, and emergent. The judgment that protects against a spiral is not a better cat model; it is the underwriting discipline to understand and limit your aggregate exposure to interconnected risk, to know who is ultimately behind the cover you bought, and to refuse the seductively profitable retro business whose ultimate exposure you cannot see. Some risks are managed not by pricing them better but by declining to be the last one holding them.

The general principle that closes this chapter is a humbling one. Reinsurance is one of the most powerful tools in insurance — it is what makes large risks writable, catastrophes survivable, and earnings stable. But it does not abolish risk; it relocates it, and relocation has limits. The risk must ultimately rest on someone's capital, and the system is only as sound as the capital and the clarity at the bottom of the chain. A primary underwriter who treats reinsurance as a magic eraser — "I can write anything, I'll just reinsure it" — has misunderstood the tool. Reinsurance lets you write more than you could hold; it does not let you write what no one, anywhere, should hold. The judgment about which risks belong in the system at all still has to be made — and it is made, first and most consequentially, at the primary underwriter's desk.

🗂️ The Underwriting File

Ceding Harbor Steel's catastrophe exposure. You have built Harbor Steel's property terms (Chapter 19): a \$20 million building on an agreed-value basis, \$8 million of equipment, \$10 million of business income, a 5% named-windstorm deductible, and an ACV roof endorsement until the warranted replacement. Now look at the same account from the height of the enterprise, and ask the two reinsurance questions this chapter taught.

First: the per-risk question — is the \$20 million line within our capacity, or do we need facultative? Recall your company's net retention on a single property risk is on the order of \$5 million [constructed teaching figure], with a surplus-share treaty providing capacity above it. Harbor Steel's \$20 million building limit is large but, against a multi-line treaty with several lines of capacity over a \$5 million retention, it is plausibly within the treaty — the surplus share absorbs the \$15 million above your retention, and your net line on a total fire loss is your \$5 million retention, not the gross \$20 million. If the treaty's per-risk capacity tops out below \$20 million, the excess goes to a facultative placement, which becomes a subjectivity on the quote (Chapter 13): you confirm the fac before you firm the full limit, never "fac to follow." Either way, the line is writable — the point is that your bet is the net retention, and the gross \$20 million is shared.

Second: the catastrophe question — how is the named-windstorm exposure ceded? This is the one that actually drives the economics. Harbor Steel sits in a named-windstorm zone on the Gulf Coast; a hurricane that hits Port Hadley will not hit only this plant — it will strike every account your company writes in that zone at once. That correlated exposure is ceded not risk-by-risk but through the catastrophe XOL treaty: the company retains the first slice of any single catastrophe event (its event retention) and the cat tower pays the layers above it. Harbor Steel's contribution to the company's modeled event loss consumes capacity in that cat tower — a real cost (the slice of cat-treaty premium the account should bear) that the net price must reflect, even though the named-storm wind deductible already shifts the first layer of any wind loss to the insured.

What this settles, and what it doesn't. It settles that the catastrophe exposure is ceded to the cat XOL treaty and that the per-risk net line sits within retention (with facultative on standby if the treaty caps below \$20 million). It does not yet settle the precise cat load, the account's full capital cost, or whether the coastal-property accumulation in the Port Hadley zone still has room — those are the next three chapters: the capital charge (Chapter 28), the portfolio and concentration fit (Chapter 29), and the cat model's PML/AAL contribution and zone aggregate (Chapter 30). Running disposition: catastrophe exposure ceded to the cat XOL treaty; net per-risk line within retention (facultative on standby for any over-treaty slice); the net-of-reinsurance cost must be carried in the price. The account is becoming writable not just as a single file but as a manageable piece of the company's net, reinsured book — which is exactly the shift Part V is teaching you to make.

Conclusion

Reinsurance is the insurance behind the insurance, and it shapes the primary underwriter's pen far more than its low profile suggests. It does three jobs — it provides capacity to write limits larger than a carrier could hold alone, it provides catastrophe protection against the correlated event that defeats the law of large numbers, and it smooths volatile earnings and protects the surplus that backs every promise. It is arranged two ways — by treaty, which covers a whole book automatically, and facultatively, one risk at a time, when the treaty runs out or won't respond. And it is shaped two ways — proportionally (quota share and surplus share, which share every dollar from the first), and non-proportionally (excess of loss and the all-important catastrophe XOL, which pay only above a retention). Each form does something specific: quota share manages surplus strain and growth; surplus share homogenizes the net book and solves the large-line problem; XOL caps severity per event; cat XOL stands between one storm and the company's solvency.

The lesson that matters most at the desk is that almost every quantity you have learned to compute exists in two versions, and your company's economics live in the net one. Your capacity is your net retention plus the treaty behind it; your price must be adequate for the risk you keep after paying for the reinsurance it forces you to buy; your combined ratio is a net number that a good program makes steadier across the cycle. The catastrophe-exposed account that is profitable gross but ruinous net is one of the most expensive traps in the business, and the discipline to price net — especially in a soft market — is the same rate-adequacy discipline that runs through the whole book. We closed with the system's hard limit: reinsurance relocates risk, but it does not abolish it, and a market that loses track of where its risk ultimately rests, as the LMX spiral showed, has hidden its catastrophe exposure rather than diversified it.

We have now seen how Harbor Steel's catastrophe exposure leaves your balance sheet for the cat treaty and how its net line sits within your retention. The next chapter asks the question standing just behind reinsurance: how much capital does this account, net of its reinsurance, tie up — and does it earn the cost of that capital? Reinsurance and capital are two answers to the same question — how much risk can an insurer safely hold? — and Chapter 28 takes up the second half.

Key Terms

• Reinsurance — insurance bought by an insurer; a contract under which a reinsurer, for a premium, indemnifies a ceding insurer for all or part of the losses on the policies the cedent has issued.
• Ceding company (cedent) — the insurer that transfers (cedes) risk to a reinsurer; it remains fully liable to its own policyholders regardless of whether the reinsurer pays.
• Treaty vs. facultative — treaty reinsurance covers a whole class or book automatically and obligatorily; facultative reinsurance is arranged optionally, one specific risk at a time.
• Quota share — a proportional treaty in which the reinsurer takes a fixed percentage of every risk in the book, receiving that percentage of premium and paying it of every loss from the first dollar.
• Surplus share — a proportional treaty that cedes only the part of each risk above the cedent's chosen retention (in multiples called lines), splitting premium and loss in the same proportion as the limit.
• Excess of loss (XOL) — a non-proportional form in which the reinsurer pays the part of a loss above the cedent's retention (attachment point), up to the reinsurer's limit; written in layers ("limit xs retention").
• Catastrophe XOL — an excess-of-loss treaty that responds to the aggregate loss to the cedent's whole book from a single catastrophe event, above the cedent's event retention; the cover that addresses correlated catastrophe loss.
• Retrocession — reinsurance bought by a reinsurer, transferring part of assumed risk to another reinsurer; spreads catastrophe risk further but can, if opaque, create a loss-amplifying spiral.
• Ceding commission — the allowance a reinsurer pays the cedent on proportional business to reimburse the cedent's acquisition and overhead expenses on the ceded premium.

Spaced Review

1. A cedent buys a per-risk excess-of-loss tower: it retains \$1M, then layer 1 is \$4M xs \$1M and layer 2 is \$5M xs \$5M. A single loss of \$8M occurs. Walk it up the tower — who pays what, and is any layer exhausted? (§27.4)
2. Distinguish quota share from surplus share, and explain which one a fast-growing young insurer would buy to relieve surplus strain and why. (§27.3)
3. (From Chapter 1.) Reinsurance's catastrophe cover exists to address the failure of a specific assumption that makes the law of large numbers work. Name the assumption and explain why a hurricane violates it. (§27.1; Ch. 1, §1.2)
4. (From Chapter 11 / the recurring pricing-discipline question.) An underwriter calls a coastal account "clearly profitable" because its gross rate exceeds its gross expected loss. Why might the account still hurt the combined ratio, and what must the price be adequate against instead? (§27.5; Ch. 11)
5. (From Chapter 4.) Your reinsurer becomes insolvent the month after a covered loss. Who still owes the original policyholder, and what does this imply about how you should weigh a reinsurer's financial-strength rating when you choose your cover? (§27.6; Ch. 4)