A natural event does not consult a company's organization chart. A flood touches, on the same day, buildings insured on property, vehicles insured on motor, sites insured on construction, goods in transit, and sometimes the company's own premises. Each of these lines measures its exposure, buys its protection and presents its results separately. Accumulation, for its part, happens on the event, and nobody is structurally tasked with looking at it there.
The first consequence is a very ordinary optical illusion: every line can sit inside its retention while the group sits outside its appetite. An event costing thirty million on property, eight on motor, six on construction and two on transit remains, for each manager, a large loss without being exceptional. The forty-six million total appears on none of the four tables, and will appear only in accounting consolidation, that is, weeks later and in a unit that no longer allows action.
The second concerns the protections bought, which do not add up as one imagines. Four programs designed line by line carry four retentions, and on a common event those retentions accumulate: the group pays the sum of the four before any protection triggers at all. The reverse reasoning is equally false, and must be said: a single protection covering the four lines would not simply be cheaper, it would change the nature of the cover, and that is not a decision a reading of accumulation is enough to carry.
The third bears on motor, the line most often forgotten in these exercises. A parked vehicle fleet is a concentration of value with neither wall nor floor: a flooded storage yard produces hundreds of total losses, with no property damage at all. And that exposure moves, being mobile: a concentration measured on garaging addresses says nothing of a fleet of new vehicles held at a port or of a delivery fleet returned to a depot for the night.
The fourth concerns construction, and it is peculiar because it is temporary. A site is an exposure that did not exist the year before and will not exist the year after, with maximum vulnerability: works not yet closed in, plant and materials stored at ground level, open earthworks that change the neighborhood's drainage. No accumulation measurement system built on a portfolio of permanent property sees those exposures appear and disappear, and they often lie, by their nature, in areas under development, therefore low ones.
One must add cross dependency, the hardest form to see. The same event can produce physical damage in one line and a loss without damage in another: goods intact but immobilized, an operation stopped by a cut road, a site suspended by inaccessibility. Those losses fall under different covers, different sublimits, sometimes different insurers, and they nonetheless share one single cause, which is precisely the definition of an accumulation.
The conduct to adopt does not require redesigning the organization, it requires an exercise few houses perform. A plausible event is taken, not an extreme one, it is laid on a map, and each line is asked what it loses on that footprint, on the same day. The exercise is short, it costs only meetings, and it almost always produces two results: a total nobody had seen, and a list of exposures nobody knew were there, a vehicle yard, a construction site, a third-party warehouse. It is the only known way of seeing an accumulation before it declares itself.
An insurance group runs a footprint accumulation exercise for the first time: a plausible flood, nothing extreme, laid over a valley where it is well established. The four lines return their figures. Property: 26 million euros, against a line retention of 20 million. Motor: 9 million, of which 6 for a fleet of new vehicles parked at a port terminal, line retention 12 million. Construction: 7 million across two development sites in low-lying ground, retention 8 million. Transit: 3 million of immobilized goods, retention 5 million. Each line manager concludes the event is absorbable. What is seen that none of them sees?
Three lines out of four are indeed inside their retention, and that is precisely what makes the finding deceptive. On this event the group pays the sum of the retentions reached before any protection really triggers: 20 million on property, plus 9, plus 7, plus 3, that is 39 million retained for a gross burden of 45. One line only triggers its protection, and for 6 million. In other words, four programs bought separately leave the group carrying almost the whole burden of an event that is in no way extreme, and none of the four tables could show it since each was correct within its perimeter. Two exposures then deserve to be looked at for their own sake, because ordinary accumulation monitoring would never have found them. The fleet of new vehicles at a port terminal is a mobile concentration of value: it appears at no garaging address, it is there this week and elsewhere next month, and it carries two thirds of the event's motor burden. The two construction sites are a temporary exposure in low-lying ground, at their maximum while they exist, invisible in a system built on permanent property, and it should be noted that their open earthworks change the neighborhood's drainage, that is, they may aggravate the burden of the property insured next door. What the exercise produces is therefore not a figure to compare with an appetite, it is a question asked at the right level: does the group accept retaining 39 million on an event of this frequency. The answer does not follow from the calculation, but the question was not being asked before a footprint was laid on a map.
- 01A natural event ignores the organization chart: every line can sit inside its retention while the group sits outside its appetite.
- 02Four programs designed line by line carry four retentions, and on a common event the group pays their sum before any protection.
- 03A vehicle fleet is a mobile concentration of value, appearing at no garaging address and visible in no property monitoring.
- 04A construction site is a temporary exposure, at its maximum while it exists, and its earthworks can aggravate the burden of property insured next door.
- 05Laying a plausible footprint on a map and asking each line what it loses on the same day is the only way to see an accumulation before it declares itself.