Prairie Wholesale: Is $60M Blanket Really Fully Insured?
Prairie Wholesale operates three distribution warehouses: Winnipeg (building $22M, stock $14M, plant $6M), Regina ($11M building, $7M stock), and Saskatoon ($16M building, $9M stock). All sit within 400 km, share one electrical utility and one sprinkler contractor, and Winnipeg and Regina are both inside the same river floodplain. Annual gross profit is $11M at Winnipeg, $5M at Regina and $7M at Saskatoon; indemnity periods are 12 months. The Winnipeg site carries $1.5M of extra expense and $1M of debris removal exposure. The program is a single blanket property limit of $60M with a $250k deductible and a $10M flood sublimit shared across all three sites. The broker calls the program 'fully insured' because $60M exceeds the largest single building value.
1. Is the program 'fully insured'? Test the blanket limit against plausible loss scenarios and explain what the blanket structure hides.
Read the worked answer
No. The blanket limit is shared, so it is only as good as the worst combination of events.
A single Winnipeg fire that runs the full indemnity period produces roughly: property $42M + business interruption $11M + extra expense $1.5M + debris $1M = $55.5M. That consumes 92.5% of the $60M limit, leaving $4.5M for Regina and Saskatoon - which sit on the same floodplain, draw from the same utility, and depend on the same sprinkler contractor. The correlations are the point: these are not independent risks, so a two-event year is plausible and the limit cannot respond.
Flood is worse. The shared $10M sublimit sits against about $60M of floodplain-exposed property value (Winnipeg $42M plus Regina $18M), before any business interruption. That is not a rounding gap; it is a structural shortfall.
What the blanket hides: per-location adequacy, the absence of contingent business interruption cover for the shared utility and contractor, and the fact that no one has compared the limit to a defined maximum foreseeable loss. 'Limit exceeds largest building' is a building-value test applied to a total-loss question.
2. Estimate a maximum foreseeable loss for a Winnipeg fire and explain how it differs from a probable maximum loss. Show your arithmetic.
Read the worked answer
Maximum foreseeable loss (MFL) assumes the worst credible outcome: protection impaired or delayed, maximum stock on hand, spread unchecked, full indemnity period consumed, plus all the ancillary costs.
Winnipeg MFL ≈ $42M property + $11M gross profit (12 months) + $1.5M extra expense + $1M debris = $55.5M, before demolition, code upgrade, or stock at seasonal peak.
Probable maximum loss (PML) assumes protection performs as designed - sprinklers control, damage is confined, supply is recovered on a shorter timeline. A defensible PML here might be $20M–$28M depending on the sprinkler and compartmentation review.
The two numbers answer different questions. PML is a pricing and retention number: how large a loss should we expect in a normal bad year? MFL is a solvency and limit number: how large could the loss be before the balance sheet is at risk? Setting a limit against PML leaves the tail uninsured. Setting it against MFL is expensive, so the correct output is an explicit, documented decision about which portion of the MFL–PML gap the company is choosing to retain, approved by the board rather than buried in a broker summary.
3. What exposure data would you require before renewal, and what structural changes would you recommend?
Read the worked answer
Data to demand: COPE detail per building (construction, occupancy, protection, exposure); sprinkler adequacy and impairment history; flood elevation and historical crest data per site; monthly stock values to catch seasonal peaks; business interruption worksheets validated by finance rather than by the broker; a dependency map for the shared utility, sprinkler contractor and any single-source suppliers; and a contingent business interruption exposure estimate.
Structural changes:
- Replace the blanket with scheduled per-location limits set at or near each site's MFL, or retain a blanket but demonstrate in writing that it covers a two-site scenario.
- Separate flood sublimit sized to floodplain exposure, or a parametric flood cover for a defined trigger.
- Add contingent business interruption and utility service interruption cover, which the current program does not appear to address.
- Raise the Winnipeg deductible and use the saving to buy limit - accept frequency, protect against severity.
- Document the retained gap in the risk register with an owner and a board-approved tolerance, so the gap becomes a decision instead of an oversight.