Chartered Property Casualty Underwriter (CPCU)

Set tolerances before growth begins; if coastal accumulation approaches the catastrophe limit, pause new writings or buy protection rather than treating premium targets as permission.

Apply it: Example: coastal limits approach the board's event cap; pause new binding until catastrophe modelling confirms net exposure.

Compare expected margin with concentration and volatility; when one segment dominates results, diversify limits and distribution across locations and products before adding volume.

Assign an executive who can accept or mitigate exposure; without a named owner, a risk can remain on the register while controls fail.

Refer a submission before binding when it exceeds delegated authority; retaining the analysis and proposed terms lets the authorized reviewer make a traceable decision.

Apply it: Example: a submission exceeds an underwriter's authority; refer the complete file instead of splitting it to bypass approval.

Test hazard, control quality, data reliability, and price adequacy together; weak loss-prevention evidence can justify tighter terms even when historical claims are low.

At renewal, use changed values, claims, and catastrophe exposure to revise pricing or terms; repeating last year's decision can leave the insurer underpaid.

Read declarations, forms, endorsements, and exclusions as one contract; a summary cannot resolve conflicting operative wording or override a later endorsement.

Apply it: Example: a later endorsement changes an exclusion; reconcile both clauses and document the operative coverage position.

State the relevant facts and clauses, then explain ambiguity and rationale; this creates a reviewable coverage position instead of relying on a broker paraphrase.

Communicate decisions accurately and preserve contemporaneous notes; omitting an assumption or promising an outcome can mislead customers and impair governance.

Divide incurred losses, loss-adjustment expense, and underwriting expense by earned premium in this simplified ratio; an 83% result signals underwriting profit for that period.

Apply it: Example: $83m of incurred loss, LAE, and expense over $100m earned premium is an 83% simplified combined ratio.

Compare reserve changes by accident year and line; concentrated adverse development can reveal assumptions or claim trends that a current-year aggregate hides.

Separate frequency, severity, trend, and mix before changing rates; otherwise a temporary shift in business composition may be mistaken for deteriorating underlying risk.

Aggregate locations under credible event scenarios; several individually acceptable properties can produce an intolerable portfolio loss when one storm affects them together.

Apply it: Example: three adjacent zones share one storm path; aggregate their insured values before approving regional growth.

Model attachment, limit, reinstatement, and net retention under the treaty; gross catastrophe exposure alone overstates protection available after wording constraints.

Stress capital against severe but plausible outcomes; if losses breach appetite or solvency buffers, reduce accumulation or strengthen reinsurance before expanding.

Trace model inputs to their source and test completeness; biased or stale data can produce precise-looking but unreliable pricing and underwriting decisions.

Apply it: Example: a postcode proxy changes pricing outcomes; validate data and fairness before approving model use.

Test whether sensitive proxies create disparate outcomes and whether their use is justified; statistical correlation alone does not establish fair or lawful treatment.

Disclose material assumptions and preserve the analysis; if pressure continues to distort results, use the designated ethics or governance escalation route.