Every answer and its explanation appears here once you have finished the path. Each one then links to the matching glossary entry, where the concept is set out in full with its worked example.
1. A flood puts 80 centimeters of water in a workshop, 1.4 million of damage. The May 2 order recognizes eleven communes; the company's is not among them, the one across the river is. The insurer declines. What should be done, and in what order?
FIRST have the COMMUNE start the procedure, since it alone can request a supplementary order and its inaction probably explains the absence; AND IN PARALLEL isolate what falls outside the regime, because windstorm, hail and snow load are paid WITHOUT any order. The notification period, for its part, only runs from publication
The commune is the grid of the scheme, and a commune that asks for nothing gets nothing: the useful action is at the town hall, not at the insurer. The proposal challenging the declination addresses the party with no power over the order, and it is right about the unfairness without being right about the counterparty. The one waiting for a supplementary order assumes an automatic process that does not exist, since the order is the outcome of a procedure and not of an observation. The one recording the water level produces a document useful to the claim and with no bearing on recognition, which is not decided on the depth of water at one insured's premises.
Glossary entry · regime-catnat2. The director filed the claim on March 16, two days after the flood, and the order is not published until May 2. What does that chronology change about its rights?
NOTHING, and that is the point: the notification period runs from PUBLICATION OF THE ORDER and not from the loss. Rights are not lost by waiting, and filing early does not improve them either
It is the only calendar rule in the regime that favors the insured, and it is poorly known precisely because it is counterintuitive: elsewhere in insurance, delay costs. The proposal running the period from the loss is the default reading of a property insurance professional, and it is exactly the point this module corrects. The one calling the filing premature invents a formality of repetition. The one speaking of priority describes an operational reality of loss adjusting firms, saturated after an event, and not a right, which is a distinction the business interruption module takes up.
Glossary entry · regime-catnat3. Two companies separated by a river, same flood, same water depth, and only one commune recognized. How is that explained to a director without invoking a mistake?
By THE GRID: the order names COMMUNES and a period, not losses. Recognition turns on the abnormal character of the natural agent MEASURED ON THAT TERRITORY, and on whether the commune applied. Two companies separated by a street may not have the same cover on the same flood, and that is not a flaw in the scheme, it is how it is built
Being able to explain that outcome without invoking a mistake is part of the job, because a director that sees its neighbor paid will conclude its insurer was at fault. The proposal invoking the level of damage is the most spontaneous and is wrong at the heart of the scheme: the amount of the loss never enters recognition. The one invoking prevention plan zoning confuses two real and distinct mechanisms, the plan acting on DEDUCTIBLE MODULATION and not on eligibility. The one invoking history reverses the meaning of that datum, which weighs on the deductible of a recognized commune, not on its recognition.
Glossary entry · regime-catnat4. Commune A, seven companies, 480,000 euros: recognized. Commune B, six kilometers away, forty one companies, 2.1 million: declined, on 62 millimeters in twenty four hours and an eight year return period. What is the appeal by B's companies based on the scale of their damage worth?
NOTHING, and that has to be said: the test measures THE RARITY OF THE NATURAL AGENT, never the severity of the damage, and the amount of the loss never enters recognition. Ordinary rain on a saturated basin can ruin forty companies without opening the cover, and that is coherent
The scheme covers the abnormality of the hazard and not the vulnerability of the territory, which is why a saturated basin, poorly managed runoff or a badly sited business park produce considerable damage without recognition. The proposal based on the number of companies is the most attractive because it seems less crude than the amount, and it hits the same wall: density of exposure is not rarity of the agent. The one making scale subsidiary invents a hierarchy of proof, when a measurement exists here and it is what decides.
Glossary entry · periode-de-retour5. The regime's perils are not all assessed the same way. Which split holds?
QUANTIFIED THRESHOLDS for flood and drought, EXPERT REPORTS for ground movement, avalanche and earthquake. The practical consequence is that a declination is challenged on a measurement in the first case and on a method in the second
The distinction governs the nature of the appeal, and that is where it is useful: on a flood or a drought one argues a station, a quantity and a reference series, whereas on ground movement one argues a method and findings. The proposal reversing the two families is the most instructive to correct, because it follows a sound intuition, earthquakes ARE measured, and misses that measuring a hazard is not enough to build a recognition threshold enforceable commune by commune. The one announcing completed harmonization describes a desirable state and not the state of the law.
Glossary entry · periode-de-retour6. On what, and on what alone, is a flood recognition declination argued?
On THREE ELEMENTS and three only: the measuring STATION used, the QUANTITY used, and the REFERENCE SERIES giving the return period. A station six kilometers from the affected area, a twenty four hour quantity when the episode lasted three hours, or too short a series are the only three handles
An appeal that does not bear on one of those three elements has no object, and the worked case shows it: the nearest station to commune B is six kilometers away, which is a real handle, whereas the scale of damage is not. The proposal comparing with the recognized communes of the same order looks rigorous and shifts the discussion to equal treatment, when each commune is assessed on ITS measurements. The one invoking adversarial procedure looks for a procedural defect where the scheme does not work like litigation. The one relating damage to the economic fabric restates the severity test in relative form.
Glossary entry · modele-catastrophe7. A drought order recognizes one QUARTER, while shrink swell damage takes three to five years. What substantive question does that create?
That of ATTACHMENT: the damage has to be attached to a PRECISE PERIOD, and that is not a formality but the substantive question of the file. A commune recognized for the third quarter of 2022 and the third quarter of 2018, and not for 2019 or 2020, forces one to say what each episode produced
The mismatch between the order's quarterly grid and the phenomenon's multi year duration is what makes these the longest files in the regime. The proposal content with a single recognition over the period is the comfortable reading and it erases the question the expert will ask first. The one apportioning pro rata to intensity invents an accounting mechanism shaped like what one would want, and the expert does not apportion, it attaches. The one invoking limitation names a real and here secondary risk, where the period runs from publication of each order.
Glossary entry · regime-catnat-secheresse8. Cracks first photographed in September 2019, marked worsening in summer 2022, filing in March 2023. The commune is recognized for the third quarter of 2022 and the third quarter of 2018, not for 2019 or 2020. What does the 2019 photograph do to the file?
It cuts BOTH WAYS, and that must be known before producing it: it DATES the onset of the damage, which helps establish the chronology, and it establishes at the same time that part of the damage predated the 2022 episode, so that not all of it is attachable to the one recognized quarter
A chronology dated from the first cracks is the only thing that usefully shortens these files, and it has a cost to anticipate rather than discover: it also bounds what can be attached. The proposal seeing only a weakening jumps to a maintenance failure, which would presuppose that maintenance would have prevented the damage, which foundations set too high contradict. The one opening a file on 2018 reasons soundly about recognition and misses the causal link, which a one year interval does not supply. The one discarding the photograph for want of a certain date applies a stricter standard of proof than an amicable expert assessment uses.
Glossary entry · regime-catnat-secheresse9. The expert finds foundations at 0.60 meters where the 2011 soil study prescribed 1.20, and an oak four meters from the most cracked facade. What becomes of the file?
THE REGIME STOPS RESPONDING: the DETERMINING cause is not the one that played a role, it is the one WITHOUT WHICH the damage would not have occurred, and foundations at half the prescribed depth in known clay soil are such a cause. The route becomes the BUILDER, if time limits still allow, and twelve years after completion they probably no longer do
This is the module's hardest conclusion and the one to be able to state: the file is not merely declined, it is declined at the moment when the only other route is probably time barred. The proposal splitting in proportion to causes applies tort reasoning to a regime that reasons by determining cause, all or nothing. The one treating the foundation defect as an aggravating circumstance is the insured's natural plea, and it reverses the test. The one referring to the inherent defects policy aims at the right cover and hits the same calendar wall, ten years having run on a 2011 building.
Glossary entry · regime-catnat-secheresse10. The cracked building loses market value, its use is impaired, and part of the premises becomes unusable. What does the regime respond for?
DIRECT PHYSICAL DAMAGE, and nothing else: neither residual loss of market value, nor loss of enjoyment. Lost time enters only through business interruption cover, which is NOT compulsory and which itself presupposes covered physical damage
The regime is a compulsory extension mechanism on a property cover, not full compensation of loss: what it does not say, it does not cover. The proposal automatically including business interruption is the commonest and costliest confusion, because it is discovered after the loss, at a company that never bought the cover. The one admitting market value reasons in economic loss, which is a court's logic and not that of a property insurance contract. The one keeping loss of enjoyment picks the more defensible of the two items, and it is no more covered.
Glossary entry · principe-indemnitaire11. A broker offers to buy back the natural catastrophe deductible through a complementary cover. What should be made of that?
THE DEDUCTIBLE IS STATUTORY: it is not negotiable, cannot be removed, and NO complementary cover buys it back. This is where an insured used to contractual deductibles goes wrong, and the offer must be rejected on principle and not on price
A statutory deductible is not an expensive deductible, it is a deductible of a different nature, and the distinction is lost as soon as its amount is discussed. The proposal calling it possible but not worthwhile is the most dangerous of the three wrong ones, because it opens a price negotiation where there is nothing to buy, and because it has the shape of prudent advice. The one distinguishing floor and excess invents a contractual boundary that does not exist. The one separating property and business interruption names a real difference of form, working days on one side and a percentage on the other, and draws a false consequence from it.
Glossary entry · franchise12. On one recognized event, a company suffers 900,000 euros of physical damage and six weeks of stoppage. How many deductible charges does it bear?
TWO, which CUMULATE: a PERCENTAGE on property with a FLOOR, and a number of WORKING DAYS on business interruption. They are two mechanisms of different natures on the same event, and one does not absorb the other
Cumulation is the rule and it surprises because the two charges are not expressed in the same unit, so they are not spontaneously added. The proposal making them alternative applies a non cumulation rule that exists elsewhere in insurance and not here. The one referring business interruption to a contractual deductible mixes two true things, since business interruption cover is indeed optional, and yet its natural catastrophe deductible is statutory like the other. The one seeing only a single charge reasons per loss where the regime reasons per item.
Glossary entry · franchise13. Two subsidiaries, same group, same program, same terms, 900,000 euros and six weeks each. The southern subsidiary's deductible is tripled. What is the answer to the finance director?
That MODULATION penalizes a commune recognized several times where NO PREVENTION PLAN has been prescribed: four recognitions in five years and no plan, and the deductible is tripled. THE ONE WHO PAYS IS NOT THE ONE WHO COULD HAVE AVOIDED IT, nothing in the policy explains it, and the reason lives at the town hall
It is the regime's hardest mechanism to explain, because it makes an insured pay for a local authority's inaction, and no contractual document says so. The proposal invoking the site's own loss record is the reading of a finance director used to experience rating, and it is wrong: it is the COMMUNE's recognitions that count, not the insured's losses. The one invoking a program clause looks for the cause in the contract, where it is not, and holds out hope of a renegotiation that will not happen. The one invoking zoning confuses a prescribed plan, which lifts modulation, with classification in a flood zone, which does not.
Glossary entry · franchise14. Which public and free datum would have avoided the southern subsidiary's surprise, and when is it consulted?
The commune's RECOGNITION HISTORY, which is public: four recognitions in five years and no prescribed plan can be read in advance. It is a SITING datum, to be looked at BEFORE THE LEASE and not after the loss
Timing matters as much as the datum: once the lease is signed the information opens no decision, which is why this module calls it a siting datum. The proposal consulting the departmental list of orders nearly reaches the right source at the wrong moment, renewal no longer allowing a change of commune. The one keeping prevention plan zoning names the element that lifts modulation rather than the one that reveals it, since what costs here is precisely the ABSENCE of a plan. The one going through flood extent mapping describes physical exposure, useful and distinct from the recognition history, which alone governs the deductible.
Glossary entry · risque-climatique-physique15. In one night, 128 kilometer per hour gusts tear off the roof and rain destroys high level stock for 340,000 euros; the stream overflows and destroys ground level stock for 260,000. How many deductibles, and of what nature?
TWO deductibles of DIFFERENT natures on the same event: one CONTRACTUAL and negotiated on the windstorm share, the other STATUTORY and fixed on the flood share. Water coming down from the roof and water rising from the ground follow different regimes, although they mix in the same building
The same building, the same night, and two legal regimes: that is the configuration the module works, and it is settled by splitting rather than unifying. The proposal keeping two statutory deductibles has the right number for the wrong reason, and it applies the regime to a peril, windstorm, that falls outside it. The one unifying under the order is the reflex of the operator in the worked case, and it puts an already vested claim to sleep. The one keeping a single contractual deductible reasons per event, which is the logic of a property policy and not of a loss with two causes under distinct regimes.
Glossary entry · delai-carence16. The operator files a single 600,000 euro claim while waiting for the natural catastrophe order. What does that wait cost it?
It PUTS AN ALREADY VESTED CLAIM TO SLEEP: the WINDSTORM share is payable IMMEDIATELY, with no order, and three hundred and forty thousand euros wait for nothing on a decision that does not concern them
Cash is the real stake of this module: a company stopped for six weeks cannot afford to wait three months for a recognition that covers only half its loss. The proposal invoking the notification period on the windstorm share names a real and secondary risk, and it is right about the starting point, which makes separating the two shares all the more urgent. The one fearing a global declination overstates the effect of a single presentation, which can be corrected. The one seeing no cost forgets that the order will not only change the deductible, it will decide half the loss.
Glossary entry · peril-secondaire17. The next day, the operator has the site cleared and cleaned. What does that cleaning destroy?
THE PROOF OF THE SPLIT between the two causes: the DEPOSIT LINE left by the water, which says how high it rose, and the POSITION OF DAMAGE IN THE RACKING, which distinguishes stock wetted from above from stock submerged from below. The gust record survives, because it sits with a third party. Three days are enough to erase the other two
The split is not reconstructed, it is observed, and it is observed on traces that operations have every reason to remove the next day. The proposal treating the split as reconstructible from records is the most tempting, because the gust record really exists: it establishes the wind, never the share of stock the wind wetted. The one invoking adversarial assessment names a real and procedural consequence of the same mistake. The one about quantifying stock points to a lesser collateral loss, inventory being reconstructible from the accounts when the deposit line is reconstructible from nothing.
Glossary entry · risque-climatique-physique18. How is the split by cause requested from the expert, and what methodological safeguard must be required?
By requesting it EXPLICITLY and in writing, with an UNDETERMINED category rather than an eyeballed split: an item honestly classed undetermined can be negotiated, an item split by eye is poorly defended on both sides
A split that is not requested is not made, and a split made by eye produces figures nobody can defend, which reopens the discussion months later. The undetermined category is the methodological move that makes the report usable. The proposal apportioning pro rata to the deposit line takes the right proof and turns it into an arithmetic rule, when the line states a height and not a causation. The one leaving the expert free confuses independence of assessment with absence of instructions, which is the insured's choice. The one appointing two experts doubles the cost and produces two reports that will contradict each other on the mixed items, precisely the ones that matter.
Glossary entry · principe-indemnitaire19. A company on high ground suffered no physical damage, and stops for nine weeks because the flood cut its road and drowned its supplier. What does ordinary business interruption cover open for it?
NOTHING by that route: business interruption PRESUPPOSES COVERED PHYSICAL DAMAGE STRIKING THE INSURED. A company stopped without its own damage opens nothing, however real its loss, and one must then look to NAMED EXTENSIONS, which are other covers and not that route
This is the condition directors discover afterwards and that closes half of corporate files after a flood: the loss is entirely real and the cover entirely inapplicable. The proposal treating inaccessibility as automatically covered names the right object and files it in the wrong place, since it is precisely an extension to be bought. The one opening everything on proof of causation reasons in tort, where causation suffices, and not in property insurance, where physical damage is required. The one splitting the stoppage at the order's date imports a calendar rule that bears on notification and not on the measure of loss.
Glossary entry · perte-exploitation20. The company did buy a supplier failure extension naming its supplier. The insured site's commune is not recognized, the supplier's is. Does the extension respond?
YES, and this is the peculiarity to remember: supplier failure in natural catastrophe DEPENDS ON AN ORDER MADE ON A COMMUNE WHERE THE INSURED OWNS NOTHING. The covered physical damage is the SUPPLIER's, and it is the supplier's commune that must be recognized
This configuration is what makes the supplier failure extension so peculiar under the regime: the trigger is an administrative act on a territory foreign to the insured, entirely outside its monitoring and yet verifiable by reading the order. The proposal requiring the insured's commune applies the main cover's rule to an extension that exists to get around it. The one conditioning the extension on the supplier's indemnification adds a condition the extension does not carry, and it describes a check that is useful in practice. The one demanding an express clause asks the extension to restate what it is.
Glossary entry · carence-fournisseur21. The flood cut the only heavy goods access road for eleven days. By what route is that stoppage indemnified?
By a NAMED extension for denial of access, distinct from supplier failure, SUB LIMITED and TIME BOUNDED, which responds only if it appears in the contract. Eleven days may well fall under a deductible expressed in days or above a modest sub limit
A cut road and a drowned supplier are two distinct extensions, often sub limited and bounded, and a company that has one does not necessarily have the other: they must be named separately before anything is promised. The proposal going through ordinary business interruption holds sound causal reasoning and hits the requirement of the insured's own physical damage, which it does not have. The one invoking the regime on the road confuses the damaged property, which belongs to the authority, with the insured, which only suffers prevented use. The one looking to neighbors and third parties recourse aims at a liability cover, which answers for damage caused and not damage suffered.
Glossary entry · carence-fournisseur22. The policy's indemnity period is twelve months, calibrated on the experience of an isolated loss. Why can it turn out to be short after a natural event?
Because a natural event is COLLECTIVE: three hundred companies in the same basin call THE SAME EXPERTS, the same contractors and the same machine suppliers THE SAME WEEK. A term calibrated on an isolated loss measures a queue that did not exist in the calibration
This is the difference in kind between a loss and a catastrophe, and it is written in no clause: the term bought is right for a workshop fire and short for the same company when the whole basin burns at once. The proposal running the period from the loss while awaiting recognition describes a real and secondary effect, and it is corrected by filing early, which another module in this course establishes. The one taking the deductible from the end of the period invents a mechanism. The one announcing a statutory reduction to six months invents a rule, and it has the shape of the regime's real constraints, which makes it plausible.
Glossary entry · perte-exploitation23. A portfolio's modeled average annual loss goes from 3.1 to 5.8 million euros. The committee proposes raising the zone's rate by twenty-five percent. What step must come before the decision?
Rerun the calculation changing ONE TERM at a time: same exposure with the new model, then new exposure with the old model. A gap attributed to the wrong term buys the wrong remedy
Hazard, vulnerability and exposure belong to three different owners and are corrected by three different actions: changing scientific source, visiting the risk, or cleaning a file. Raising a rate where capital values needed fixing charges the same square meter twice, once through the corrected value and once through the rate. The answer comparing to actual experience confuses a distribution with three draws. The one deferring to the vendor asks it to explain a gap partly caused by the insurer's own portfolio, which the vendor does not know. The one waiting a year assumes time will reveal what a decomposition gives in a day.
Glossary entry · modele-catastrophe24. For a single 46 million euro site, a model returns an average damage ratio of eleven percent. The finance director wants to set the retention on it. What is missing?
The DISPERSION around that mean: across ten thousand buildings of the class, eleven percent is the right figure; on this one, the question is which side of the mean it falls, and the mean does not say
A damage curve returns a class average, and using it on a single site is precisely what it is not built for: models attach secondary uncertainty to each point, and that is what fails to circulate because a mean fits in a table while a distribution needs a page. The three other answers name real and necessary checks. Return period, valuation basis and deductible all have to be established, and none answers the question asked, which is what a single figure can size.
Glossary entry · principe-indemnitaire25. A director refuses an increase: two floods described as hundred year events in three years would prove the modeling wrong. Three answers are possible. Which can turn against him?
NON-STATIONARITY: if the fitted series describes a catchment that urbanization has changed, the stated return period is too long, so the event is MORE frequent than announced
All three do dismiss the argument, and only one changes the direction of the file: a series fitted on a catchment that no longer exists produces a wrong number with the same confidence as a right one, and the error runs the damaging way. The answers on the gauge and on hazard versus loss are accurate and neutral. The one concluding that none turns misses what makes the third valuable: it does not merely defend the increase, it says where to look, on which series and which station the number was fitted, and that question goes to the vendor rather than the client.
Glossary entry · periode-de-retour26. A hundred thousand year stochastic catalog is built on forty years of observations. What does it add, and what does its length not measure?
It adds RESOLUTION and no new information about the physics: it knows what forty years knew, expressed more finely. Its length only reduces simulation noise
Where history offered one storm of 1999, the catalog offers three thousand neighboring ones, some passing fifty kilometers further south where the portfolio sits: that is resolution, and it is valuable. What it does not manufacture is physical knowledge. The answer on temporal depth is the one most often said and confuses simulating with observing. The one on independence describes an ASSUMPTION of the catalog, not a contribution, and a contestable one since a storm season often brings several. The one on robustness inverts the relationship: extrapolation sits upstream of the draw, and lengthening the draw does not correct it.
Glossary entry · modele-catastrophe27. A note returns 140 million euros OEP and 205 AEP at the hundred year return period. A per-event structure of 110 above 30 is proposed. Which error is the most dangerous, and why?
Sizing an ANNUAL protection by reading a PER-EVENT curve: the protection looks sufficient, it is on the isolated loss, and it proves short in the year three medium events follow each other without any clearing the retention
The two errors are symmetric and cost both ways; the overestimate shows, since unnecessary height is paid for, and the underestimate does not, because nothing in the figures presented flags it. The 65 million gap is not imprecision: it says that on this portfolio the share of the hundred year burden coming from REPETITION is of the same order as that from the largest event. The answer calling it simulation noise confuses a structural gap between two measures with the randomness of a draw. The one on return period names an elementary check which, here, has already been made.
Glossary entry · pml28. Two models return 118 and 167 million euros on the same portfolio. What temptation does the discipline of a documented house view protect against?
Against the view chosen AFTER seeing the figures, always chosen for comfort: the low model when growth is pressing, the high one when a protection is negotiated
What makes a view defensible is not its value but that it is stated BEFORE the results, applied without exception, and revised on criteria rather than on mood. A house that arbitrates twice a year in its own interest does not have a view of risk, it has an instrument of justification, and it turns on the day a loss requires explaining why the low curve was used. The answer on a single model describes a real dependency risk, unrelated to the order of decisions. The one on the mean names one of the two reflexes to refuse, and refusing it silently is not enough. The one on publication handles a communications question.
Glossary entry · risque-climatique-physique29. A portfolio is geocoded to the postcode centroid. Why does the error not average out across many contracts?
Because aggregation SMOOTHS intensities: highly exposed sites receive too low an intensity and protected ones too high. On the mean the error appears to cancel, on the TAIL it does not, since the tail is made of the sites smoothing pulled back toward the mean
An aggregated portfolio reports an extreme loss that is too low, and it does so SYSTEMATICALLY: this is not noise that numbers erase, it is a directed bias, and it falls precisely on the part of the distribution used for sizing. The answer on varying areas describes a real irregularity that modulates the size of the bias without changing its direction. The one on rural failure describes a geocoding defect, a different subject. The one on a flat penalty credits the model with behavior it does not have: it calculates without complaining, and that is exactly what makes the defect invisible.
Glossary entry · accumulation-cumul30. A model's catalog groups damage over seventy-two hours, the treaty uses one hundred and sixty-eight. What follows, and when is it discovered?
The OEP curve read does not describe the same thing as the structure it is sizing, and the gap appears ON THE DAY OF A LOSS, not before
The definition of an event is not given by nature, it is given by a text, and there are two: the contract's and the catalog's grouping convention. When they do not coincide, the figure used for sizing measures something other than what is being sized, and nothing flags it before settlement. The answer on a conversion factor invents a tool. The one limiting the effect to the aggregate curve inverts the mechanism, since grouping is precisely what defines an event, hence the OEP. The one saying the treaty prevails is right on the law and misses the point: the treaty does prevail, and it is the measurement that becomes false.
Glossary entry · peril-secondaire31. A site zoned blue has its first floor at 2.80 meters, no basement, its chillers on the roof. Another, outside any zone, sits at the foot of an urbanized slope with its electrical cabinets in a semi-buried room. The committee surcharges the first. Where is the method wrong?
It reads a map built to answer the planner's question: a regulatory zoning says what may be BUILT, it does not say what an already built structure will lose. What decides vulnerability is the first floor level and the position of technical equipment
Three maps exist, regulatory, hazard and tariff, built for three purposes, and none was made to answer the underwriter's question. The site outside the zone that ends up under water is no exception: runoff appears on no fluvial flood map, and a slope urbanized for fifteen years is a slope whose absorption has been removed. The answer on severity stays inside the map instead of leaving it. The one on the compulsory cover is accurate and misses, selection then working through the other covers and through acceptance. The one on loss history asks for a record two sites cannot provide.
Glossary entry · regime-catnat32. A slope slides after a wet winter, an order recognizes the state of natural disaster, and the expert notes that a drain prescribed in the geotechnical study was never installed. What is the FIRST step?
Preserve the evidence: have the state of the slope, the wall and the missing drain recorded BEFORE any make-safe earthworks, which will destroy the evidence of causation within days
The scheme answers for damage whose DETERMINING CAUSE is the abnormal intensity of a natural agent, and the expert's finding names a second one: the two routes have different deadlines, counterparties and evidence, and opening the wrong one costs months that are not recovered. The two answers choosing a route immediately make the same mistake in opposite directions. The one awaiting the expert looks the most reasonable and is the most expensive: during the wait, make-safe works will have erased what the expert was to rule on, and the second route may be closed by the time it is opened.
Glossary entry · regime-catnat33. An insurer measures flood accumulation by department. A sub-catchment mapping shows forty-three percent of values exposed to the same event. What follows about the departmental limit?
It was not too high, it measured SOMETHING OTHER than what it thought: a flood follows a catchment, and no limit set on an administrative grid will ever see that concentration
Counting by department measures accumulation with an instrument whose graduations are not those of the phenomenon, and the result can err in either direction. The answer calling it too high keeps the wrong unit and corrects only its level, which will reproduce the defect at the first redistribution of the portfolio. The one making it an upper bound keeps a role it does not fill, since it can also understate. The one declaring it useless for every peril overgeneralizes: the relevant unit is the peril's, catchment for flood, a wide grid for storm, clay formations for subsidence.
Glossary entry · accumulation-cumul34. Eighteen contracts in four other departments carry a supplier failure extension naming a logistics site in the exposed sub-catchment. Why does that accumulation escape every geographic grid?
Because it is not read on a map but in the EXTENSIONS BOUGHT: these insureds are not in the same place, they depend on the same place, and nobody aggregates extensions by named site
An insurer may have a single plant in a valley and end up with fifteen claims, because fourteen of its insureds depend on it for one part. The fix is a bounded data task, aggregating extensions by named site, and it belongs to no department by default. The answer on a separate line moves the problem to another table that will see it no better. The one on sub-limits treats a per-item cap as a reason to ignore a total. The one on ownership confuses what the insurer owns with what its insureds depend on, which is exactly what the extension is about.
Glossary entry · carence-fournisseur35. A year looks excellent on subsidence. Why is that reading dangerous, and on which two attributes is the exposure actually measured?
Because the burden may already exist and simply not be reported yet: a dry summer is reported the following spring. Exposure is measured on sensitive clay SOIL crossed with BUILDINGS whose foundations are insufficiently anchored
This peril produces no images and has become one of the heaviest items of the scheme: defects appear the following autumn, become visible in winter, are reported in spring. Treating that lag as luck is the error that makes reserves rise three years running. The three other answers each name a plausible lag mechanism, appeal, smoothing, deductible, and above all exposure attributes that are the wrong ones: neither altitude, nor age, nor density says whether a soil shrinks and whether foundations are anchored, and almost no insurance file carries the second of those two attributes.
Glossary entry · regime-catnat-secheresse36. A house grows fast in an exposed region where property rates are twenty percent below average. Growth, premium and expense ratio all improve together. What do none of these indicators see?
That the cover's income follows a MARKET RATE and the exposure a GEOGRAPHY: the surcharge being a percentage of the property premium, twenty percent less is collected for an exposure that does not depend on local price levels
The surcharge rate is set by regulation, uniform across the territory: an underwriter does not price this risk, it accepts or declines the whole policy. Doubling in a low rate region therefore doubles an exposure while collecting a less than proportional income, and the gap shows only if someone measures it. The answer on a short history is true and a commonplace of the field, applying to any region. The one on the expense ratio is accurate and bears on a different indicator. The one on competition describes a market movement unrelated to the mechanism of the scheme.
Glossary entry · franchise37. The public annual protection's threshold is expressed as a percentage of retained premiums, which have risen thirty-four percent in three years. What has changed, and who decided it?
The company raised its own retention by the same proportion WITHOUT any decision being taken: a structure unchanged on paper has changed in fact, and the protection now triggers later
This is the kind of drift a purely net reading never shows, all the more quiet because nobody intended it. The answer seeing nothing states the mechanism's INTENTION, which is defensible, and concludes there is nothing to watch: it remains to be known whether the retention thus raised still matches what the house accepts to carry, and that question belongs to the board. The one saying it triggers earlier has the direction wrong. The one attributing the change to the scheme moves to a third party an effect produced by the company's own growth.
Glossary entry · stop-loss38. An event costs 30 million in property, 8 in motor, 6 in construction and 2 in marine. Each line stays within its retention. What does the group discover, and when?
That it pays the SUM OF THE FOUR RETENTIONS before any protection responds, and it discovers this in accounting consolidation, weeks later, in a unit that no longer allows action
A natural event does not consult a company's org chart, and each manager sees a large loss without it being exceptional: the forty-six million total appears on none of the four reports. The answer having the programs complement each other describes the opposite of what happens. The one invoking an inter-line accumulation clause invents a mechanism that does not exist by default. The one isolating motor keeps a sound observation from the module, a parked fleet is a concentration with no wall and no floor, and uses it to dismiss the other three lines, which is exactly the error the footprint exercise exists to avoid.
Glossary entry · perte-exploitation