Art. 20 · What your balance sheet can absorb · 3 min read
Can your reserves absorb a bad year?
No year is average. The number that matters is how often a year costs more than you have available.
Once you have a reserve and a stated bar, the two can be put together. Premiums are budgeted, so leave those aside; what the reserve exists to meet is the unplanned part of the year: the excesses you pay, and the losses you carry uninsured. Add those up across a year and ask how often the total exceeds what you have available.
That frequency has a name worth knowing: how often the reserve is saturated. If you have £120,000 available and the total outflow exceeds it in six years out of a hundred, then you are running a business that expects to have a serious problem roughly every seventeen years. Whether that is acceptable is the question from Article 18, but now it has a number attached.
Two traps to avoid when you or anyone else does this properly.
Do not add up every worst case at once. A "bad year" is not fire plus flood plus a liability claim plus a cyber incident all landing together. Those are largely independent, and the chance of all of them in one twelve-month period is astronomically smaller than the chance of any one. Adding worst cases produces a number so frightening it prevents decisions rather than informing them.
But do not assume everything is independent either. One storm can produce property damage, business interruption and a liability claim from the same afternoon. A recession can bring credit losses, a claim spike and a fundraising problem at once. Where a single cause touches several lines, they move together, and that is precisely when reserves get tested.
Use it
- Ask for the breach frequency, not the average. "How often does this exceed what we have?" is the decision-grade question.
- Test one change at a time. Excess up, reserve down, limit up. Watching the breach count move tells you which lever is actually load-bearing.
- Look for common causes. List the single events that would hit more than one policy. That short list is where your correlation risk lives.