Per Risk Excess of Loss Reinsurance vs Cat XoL Explained
On this page
- Per Risk vs Catastrophe Excess of Loss: How to Structure Property Treaty Protection
- Key Takeaways
- What Is the Difference Between Per Risk and Cat XoL?
- What Is Per Risk Excess of Loss Reinsurance?
- What Does Catastrophe Excess of Loss Protect?
- How Do Per Risk and Cat XoL Interact?
- When Should You Use Per Risk, Cat XoL or Both?
- Which Structural Mistakes Surface Only After a Loss?
- How Should Reinsurers Underwrite a Combined Property Programme?
- How Does the Structure Affect Capital and Reporting?
- Does the Same Logic Apply to Retrocession?
- What Does the Market Look Like for 2027 Renewals?
- Checklist for CUOs and Heads of Reinsurance
- Key Terms
- Sources
- Frequently Asked Questions
Per Risk vs Catastrophe Excess of Loss: How to Structure Property Treaty Protection
Last reviewed: November 2026
Per risk excess of loss reinsurance pays when one risk, such as a single factory or warehouse, suffers a loss above the retention. Catastrophe excess of loss pays when one event, such as a windstorm or earthquake, causes an accumulation of retained losses across many risks.
Most property programmes need both covers. Whether the programme performs on the day of a loss depends less on the price of each layer than on how the two covers interlock.
This guide is for CUOs, treaty underwriters, heads of outwards and retro, and CROs. It covers how the two covers fit together, where programmes leak, and what to check before you quote, place or renew.
Key Takeaways
- Per risk XoL protects against large losses on a single risk; cat XoL protects against the accumulation of many losses from one event.
- A per risk treaty without an event limit can turn into unpriced catastrophe cover when a windstorm hits many risks.
- Per risk recoveries normally inure to the cat XoL, so the cat layers respond to the loss net of per risk recoveries.
- Changing the per risk event limit can shift millions of loss onto the cat panel without any visible change in the cat rate.
- The hours clause must define "one event" the same way in both the per risk and cat wordings.
- Reinsurers should price the whole programme, not just the layer they are quoting.
What Is the Difference Between Per Risk and Cat XoL?
Per risk excess of loss controls large single-risk volatility, while catastrophe excess of loss protects earnings and capital from accumulation events. The table below compares the two covers side by side.
| Feature | Per Risk Excess of Loss | Catastrophe Excess of Loss |
|---|---|---|
| What triggers it | Loss to one risk above the retention | Aggregate loss from one event above the retention |
| Unit of loss | Each and every loss, each and every risk | Each and every loss occurrence, within an hours clause |
| What it protects | Line size and large single-risk volatility | Earnings and capital against accumulation events |
| Typical attachment driver | Net line the cedent is comfortable retaining | Event loss the cedent can absorb within earnings tolerance |
| Limit sizing basis | Largest PML or sum insured in the risk profile | Modelled loss at a chosen return period |
| Main pricing approach | Experience plus exposure rating with exposure curves | Vendor cat models plus experience and scenario adjustments |
| Key protective clause for the reinsurer | Event limit and risk definition | Hours clause and ultimate net loss definition |
| Reinstatements | Often several, sometimes free on working layers | Usually one, paid pro rata as to amount |
| Main capital effect | Reduces net large-loss volatility and fire concentration losses | Reduces the net catastrophe scenario loss |
The more useful insight sits underneath the table. The two covers share the same underlying losses, so a change to one silently changes the loss that reaches the other.
What Is Per Risk Excess of Loss Reinsurance?
Per risk excess of loss reinsurance is a non-proportional treaty that pays the part of any single-risk loss above the cedent's retention, up to an agreed limit. It lets a cedent write larger lines than its balance sheet could carry on a net basis.
A typical working structure might be USD 8 million excess of USD 2 million, each and every risk. It responds to fires, explosions and other large single-site losses within the scope of the treaty.
How Is "One Risk" Defined?
Most property per risk wordings use some form of "risk as defined by the reassured". This gives the cedent discretion over what counts as one risk.
In principle that is reasonable, because the cedent knows how it underwrote the business. In practice it is the source of most per risk disputes.
Consider a manufacturing complex with five buildings separated by fire walls and open space. If the cedent underwrote it as one risk with one PML, a USD 15 million fire is one loss to the layer, capped at the layer limit.
If the cedent later argues that each building is a separate risk, the same fire can become two or three recoveries. The reinsurer's protection is not a tighter wording alone, but evidence.
The risk definition used for recoveries should match three records:
- How the risk was recorded in the cedent's underwriting system at inception
- The PML or EML basis used in the risk profile submitted for pricing
- The facultative cession record for risks above treaty capacity
If those three records do not align at underwriting time, they will not align at claims time either.
Why Does a Per Risk Treaty Need an Event Limit?
A per risk treaty needs an event limit because one catastrophe can hit the layer many times. Without a cap, the per risk panel ends up providing catastrophe cover at a working-layer price.
Take a windstorm that damages 3,000 risks and produces 20 losses above a USD 2 million retention. Without an event limit, a USD 8 million layer could pay up to USD 160 million from that single storm.
The standard answer is an event limit, also called a catastrophe limitation, commonly set at two to three times the layer limit. Some per risk layers exclude named perils altogether.
Whatever the choice, the event limit must meet three tests:
- It uses the same hours clause as the cat XoL, so both panels measure the same event.
- It is priced, with an explicit nat cat load in the per risk rate.
- It is reflected in the cat XoL pricing, because it changes the loss that reaches the cat layers.
How Are Per Risk Layers Priced?
Sound per risk pricing blends an experience view with an exposure view. Credibility weighting decides how much each view counts.
- Experience rating. Build a ten-year large-loss list, index it for inflation and sums insured, and develop it to ultimate. Then apply the current layer to each historic year to get the burning cost.
- Exposure rating. Apply exposure curves, such as MBBEFD-based or market-standard property curves, to the risk profile banded by PML or sum insured. This shows the expected layer loss from what the cedent writes today.
Credibility weighting matters most on higher layers, where experience is thin. A layer that has not been hit in eight years is not necessarily cheap; it may simply be too high for the experience to say anything.
When a portfolio has shifted towards larger industrial or energy risks, exposure rating should carry more weight. A clean loss history can hide a much heavier risk profile.
What Does Catastrophe Excess of Loss Protect?
Catastrophe excess of loss protects the cedent's net retained loss from one loss occurrence, above a retention and up to a limit. It is the principal tool for protecting earnings and solvency capital against natural catastrophes.
Some wordings also cover man-made events such as riot or explosion. The definition of the event is what makes or breaks a cat recovery.
How Does the Hours Clause Define One Event?
The hours clause sets how long one "event" can last for the purpose of the treaty. Common market practice is 72 consecutive hours for windstorm, hail and riot, and 168 hours for earthquake and flood.
The cedent can usually choose when the hours period begins. These periods vary by wording, so always check your own contract.
Three practical points matter for reinsurers:
- Long-duration events. A slow-moving flood or a run of convective storms can outlast the hours clause and split into two loss occurrences. Each then carries its own retention and may trigger a reinstatement.
- Wildfire and secondary perils. Older wordings often do not set hours periods for wildfire or severe convective storm. That ambiguity favours whoever argues it best after the loss.
- Consistency. The per risk event limit should use the same event definition. Otherwise one storm can be one event for the cat panel and two events for the per risk panel.
How Are Cat XoL Limits and Retentions Set?
The top of a cat programme is usually set at a modelled loss for a target return period. Under Solvency II this is typically 1-in-200, and rating agency capital models often test 1-in-250.
The retention reflects how much event loss the cedent can absorb within its earnings tolerance. Three cautions apply when setting both points:
- Vendor models remain stronger on primary perils, such as hurricane and earthquake, than on flood, convective storm and wildfire. Experienced buyers and reinsurers load model output for these secondary perils.
- Exposure data quality changes the answer. Poor geocoding, missing construction attributes and stale sums insured routinely understate modelled losses.
- Since the 2023 market reset, cat retentions have moved materially higher. Many cedents now retain frequency events that their programmes used to absorb.
How Do Reinstatements Work?
Cat layers typically carry one reinstatement at 100% additional premium, paid pro rata as to amount. Lower per risk layers may have several reinstatements, sometimes free.
For the reinsurer, reinstatement premium is part of the expected technical result. For the cedent, it is a post-event cash call at the worst possible moment, so both sides should model it explicitly.
How Do Per Risk and Cat XoL Interact?
In a properly drafted programme, per risk recoveries inure to the benefit of the cat XoL. The cat layers respond to the ultimate net loss, which is the gross event loss minus the inuring per risk recoveries.
The ultimate net loss clause and the inuring reinsurance clause in the cat wording must state this explicitly. If they do not, the order of recoveries becomes a negotiation after the event.
Worked Example: How the Event Limit Moves Loss Between Panels
The illustration below is simplified and uses USD millions. It shows how one per risk term changes what the cat panel pays.
Programme:
- Per risk XoL: 8 xs 2, each and every risk, event limit of 3x the layer limit (24)
- Cat XoL Layer 1: 40 xs 10
- Cat XoL Layer 2: 100 xs 50
Event: a windstorm produces a gross loss of 90 across about 3,000 risks. Four risks suffer large losses of 12, 9, 6 and 4, and the remaining 59 falls on smaller risks below the per risk retention.
| Step | Calculation | Result |
|---|---|---|
| Per risk recovery, risk 1 | min(12 − 2, 8) | 8 |
| Per risk recovery, risk 2 | min(9 − 2, 8) | 7 |
| Per risk recovery, risk 3 | 6 − 2 | 4 |
| Per risk recovery, risk 4 | 4 − 2 | 2 |
| Total per risk recovery | 21, within event limit of 24 | 21 |
| Ultimate net loss to cat XoL | 90 − 21 | 69 |
| Cat Layer 1 recovery | min(69 − 10, 40) | 40 |
| Cat Layer 2 recovery | 69 − 50 | 19 |
| Cedent's net retained loss | 90 − 21 − 40 − 19 | 10 |
Now change one term. At renewal, the per risk panel pushes the event limit down from 3x to 1x the layer limit, which is 8.
| Item | Event limit 3x | Event limit 1x | Change |
|---|---|---|---|
| Per risk recovery | 21 | 8 | −13 |
| Ultimate net loss to cat XoL | 69 | 82 | +13 |
| Cat Layer 2 recovery | 19 | 32 | +13 |
| Cedent net | 10 | 10 | 0 |
The cedent's net position is unchanged. The cat Layer 2 panel now carries 13 more on this event, and nothing in the cat submission necessarily flags it.
The lesson for reinsurers: you cannot price a cat layer in isolation from the per risk structure beneath it. Any change to the per risk retention, limit or event limit is effectively a change to the cat layer's attachment.
Your submission review should therefore require the full inuring structure, not just the layer being quoted.
When Should You Use Per Risk, Cat XoL or Both?
The right structure depends on line size, catastrophe exposure and how much earnings volatility the cedent can carry. The table below maps common portfolio situations to the structure that usually fits.
| Portfolio Situation | Structure That Usually Fits | Why |
|---|---|---|
| Large commercial and industrial lines, low nat cat exposure | Per risk XoL, with a high-attaching cat XoL | Single large fires drive volatility more than events do |
| Personal lines with small, uniform sums insured in a cat zone | Cat XoL, often alongside a quota share | Individual losses rarely pierce a per risk retention |
| Mixed commercial book in a cat-exposed region | Per risk XoL plus cat XoL, with an event limit and clear inuring | Both large single risks and accumulation are material |
| Risks whose PML exceeds the per risk top limit | Facultative reinsurance above the treaty | Treaty capacity alone leaves a net gap |
| Frequent mid-sized events eroding earnings below the cat retention | Aggregate XoL or a lower cat layer | Each event stays under the retention, but the year adds up |
These are starting points, not rules. The final structure should be tested against modelled events and the cedent's actual large-loss history.
Which Structural Mistakes Surface Only After a Loss?
Seven structural mistakes cause most disputes, leakage and unexpected net losses in combined property programmes. Each one is cheap to fix at renewal and expensive to discover after an event.
Per Risk Cover With No Event Limit
The per risk panel ends up providing unpriced cat cover. The cat panel can be underpriced too, because its modelled losses assumed per risk recoveries that a later event limit may remove.
Mismatched Hours Clauses
The per risk event limit and the cat XoL measure "one event" differently. One storm then becomes two events for one panel and one event for the other, and the inuring calculation breaks down.
Cat Retention Set on a Gross Basis
The cat retention is calibrated to gross modelled losses rather than losses net of per risk recoveries. The effective retention is then higher than the board intended, or the programme buys cover it does not need.
A Gap Above the Per Risk Top Limit
The largest PMLs in the portfolio exceed the per risk top limit, and facultative cessions are not placed consistently. A single large fire can then leave a net loss well above the stated retention.
Attachment Basis Mismatch
The per risk treaty runs on a risks-attaching basis while the cat XoL runs on losses-occurring. This is workable, but recovery calculations must track it carefully.
At renewal or cancellation, the inuring cover for a loss may sit in a different treaty year with a different panel. Without a clear mapping, recoveries get missed or double-counted.
Risk Definition Not Matched to Underwriting Data
The cedent's policy administration system records locations differently from how the risk definition treats them. After a loss, nobody can prove whether the claim was one risk or three.
Recovery Calculations Held in Spreadsheets
Event limits, inuring order, reinstatement premiums and panel shares are calculated manually across several files. After a large event, time pressure turns small errors into material recoverable disputes and delayed collections.
How Should Reinsurers Underwrite a Combined Property Programme?
Reinsurers should underwrite the whole programme, even when quoting only one layer. The data request and renewal questions below make that practical.
What Data Should Reinsurers Request?
| Data Item | Why It Matters |
|---|---|
| Risk profile banded by PML and by sum insured | Drives exposure rating and shows line-size drift |
| Ten-year large-loss list, indexed, with cause of loss | Drives experience rating; shows which losses were event-related |
| Flag on each large loss: single-risk or part of an event | Tests how much nat cat has historically reached the per risk layer |
| Geocoded exposure with construction, occupancy and year built | Determines whether the cat model output can be trusted |
| Full programme structure, including inuring covers and event limits | Needed to price any single layer correctly |
| Line guide and facultative cession policy | Shows how risks above treaty capacity are handled |
| PML methodology and the definition of a risk in the underwriting system | Tests whether the risk definition will hold at claims time |
Which Questions Should Reinsurers Ask at Renewal?
- Has the share of the portfolio above the per risk top limit grown, and is facultative support placed for every such risk?
- Which per risk losses in the last five years came from nat cat events, and would the proposed event limit have capped them?
- How does the cedent's own cat model view compare with the vendor view, and what loads apply for secondary perils?
- Are the per risk and cat hours clauses identical?
- Who calculates recoveries after an event, and on what system?
These questions also feed the wider use of data and models in treaty decisions. We cover that in more depth in AI in reinsurance underwriting.
How Does the Structure Affect Capital and Reporting?
The choice between per risk and cat structure changes capital and financial reporting, not just the technical result. Three frameworks matter most.
Solvency II
Under the Solvency II standard formula, cat XoL reduces the net loss in the natural catastrophe sub-module scenarios. Per risk XoL is especially relevant to the man-made fire risk concentration scenario.
That scenario is based on the largest concentration of fire exposure within a 200-metre radius. A well-structured per risk treaty reduces the net loss from it.
Internal model firms capture both effects directly. Either way, the inuring order modelled for capital must match the inuring order in the wordings.
Rating Agencies and Counterparty Risk
Rating agency capital models stress net catastrophe losses at high return periods. They also charge for reinsurance recoverables according to each counterparty's credit quality.
As a result, a cheaper panel with weaker security can cost more in capital than it saves in premium.
IFRS 17
Under IFRS 17, reinsurance contracts held are measured separately from the underlying contracts. Many annual property XoL treaties qualify for the premium allocation approach.
Loss-recovery components and reinstatement premiums still need careful treatment. A recovery calculated incorrectly under the treaty will also be reported incorrectly.
Does the Same Logic Apply to Retrocession?
Yes. Reinsurers face the same structural choice on their own outwards programme.
Per risk retro protects large single-risk lines written on a facultative or per risk treaty basis. Cat retro and industry loss warranties protect the accumulated event position across the assumed book.
The inuring order, event definitions and basis mismatches described above apply equally. They are often harder to manage, because the underlying data arrives later and in less detail.
For event-triggered alternatives where speed of payment matters more than indemnity precision, see AI in parametric cat insurance for reinsurers.
What Does the Market Look Like for 2027 Renewals?
Since the 2023 reset, cat retentions have moved materially higher, and cedents have retained more frequency and secondary-peril loss. With capacity returning and pricing softening through 2025 and 2026, the pressure has shifted to terms.
Cedents are seeking lower attachments, broader hours clauses, more generous per risk event limits and aggregate cover for frequency loss.
Softening markets usually show up in terms and conditions before they show up in rate. The per risk event limit and the hours clause are typically where discipline erodes first.
As the worked example shows, an erosion on the per risk side can land on the cat panel without a visible change in the cat rate. Reinsurers that underwrite the whole programme, not just the layer in front of them, will see that drift before it reaches their results.
Checklist for CUOs and Heads of Reinsurance
Use this checklist before you quote, place or renew a combined property programme.
- The risk definition matches how risks are recorded in the underwriting system and the risk profile
- The per risk event limit is set, priced, and uses the same hours clause as the cat XoL
- Per risk recoveries explicitly inure to the cat XoL in the ultimate net loss and inuring clauses
- The cat retention is calibrated net of per risk recoveries
- Facultative support is in place for every risk whose PML exceeds the per risk top limit
- Attachment bases (risks-attaching vs losses-occurring) are mapped across treaty years
- Reinstatement premiums are modelled for both cash flow and technical result
- Recovery, allocation and reconciliation run on a controlled system rather than spreadsheets
For the last point, a centralised reinsurance treaty management platform holds terms and settlement calculations in one place. Bordereaux reconciliation then allocates recoveries across panel shares and checks them against the records.
Key Terms
Per risk excess of loss (per risk XoL): A treaty that pays the part of a loss to one risk above the retention, up to a limit.
Catastrophe excess of loss (cat XoL): A treaty that pays the part of the cedent's total net loss from one event above the retention, up to a limit.
Event limit (catastrophe limitation): The maximum a per risk treaty pays for all risks affected by one event. It is usually a multiple of the layer limit.
Hours clause: The wording that defines how long one loss occurrence can last, such as 72 hours for windstorm.
Ultimate net loss (ULN): The loss the cedent actually retains after inuring reinsurance recoveries and salvage. Cat XoL layers respond to the ULN.
Inuring reinsurance: Reinsurance whose recoveries are deducted before another treaty responds. Per risk XoL usually inures to the benefit of cat XoL.
PML (probable maximum loss): The largest loss a single risk is reasonably expected to suffer. It drives the per risk risk profile and line sizes.
Reinstatement premium: The additional premium paid to restore a layer's limit after a loss, usually pro rata as to amount.
Risks-attaching vs losses-occurring: Risks-attaching covers losses on policies incepting during the treaty period. Losses-occurring covers losses that happen during the treaty period, whatever the policy start date.
Sources
Frequently Asked Questions
What is per risk excess of loss reinsurance?
Per risk excess of loss reinsurance is a non-proportional treaty that pays when the loss to one risk exceeds the cedent's retention, up to an agreed limit. It protects the cedent against large single-risk losses such as a major factory fire.
What is the difference between per risk and catastrophe excess of loss?
Per risk excess of loss responds to a large loss on one risk. Catastrophe excess of loss responds to the total of many retained losses from one event within an hours clause. Per risk controls large-loss volatility, while cat XoL protects earnings and capital.
Can a per risk excess of loss treaty respond to a catastrophe event?
Yes, unless nat cat is excluded. If a windstorm causes several losses above the per risk retention, each one can recover from the per risk layer. That is why most per risk treaties carry an event limit that caps total recoveries from one event.
What is a typical event limit on a per risk treaty?
A per risk event limit is commonly set at two to three times the layer limit for any one event. Some per risk treaties exclude named perils entirely and leave catastrophe exposure to the cat XoL programme.
Do per risk recoveries reduce the loss to the cat XoL?
Normally yes. Per risk recoveries usually inure to the benefit of the cat XoL, so the cat layers respond to the ultimate net loss after per risk recoveries. Both the ultimate net loss and inuring clauses should state this explicitly.
How is per risk excess of loss reinsurance priced?
Reinsurers price per risk layers by blending experience rating, based on an indexed and developed large-loss burning cost, with exposure rating using exposure curves on the cedent's risk profile. A separate load is added where nat cat can hit the layer.
How are cat XoL limits and retentions set?
The cat XoL top limit is usually set at a modelled loss for a target return period, such as 1-in-200 for Solvency II or 1-in-250 for rating agencies. The retention reflects how much event loss the cedent can absorb within its earnings tolerance.
Does a property insurer need both per risk and cat XoL?
Most commercial property books in catastrophe-exposed regions need both. Per risk XoL supports large lines, while cat XoL protects against accumulation. Books with small, uniform sums insured may rely mainly on cat XoL.
Hitul Mistry
CEO, Insurnest
An InsurTech leader with more than a decade of experience across insurance and technology, focused on solving business problems with the help of technology. Has worked with brokers, insurance carriers, and reinsurance firms across the India, UAE, and US markets.
View LinkedIn profile →
