In almost all property insurance, you insure first and value later. The contract sets a sum insured, and on the day of the loss an adjuster establishes what the property was really worth: the indemnity principle holds that the insured should be put back where it stood, no more and no less. Fine art insurance reverses that order. The parties agree in advance on an amount that will be paid on total loss, without appraisal and without discussion. The act of underwriting therefore already contains the settlement of the claim, delivered years ahead, and that is the peculiarity everything else follows from.
The reason for the reversal is not commercial convenience. A work has no observable intrinsic value: it is worth what the last comparable sale made, what the artist's standing has become, what a recent exhibition did to its reputation. Deferring valuation to the day of the loss means deferring a debate nothing will settle, about an object that no longer exists. Agreed value removes that debate by moving it to a moment when the object is still there, when it can be looked at, compared and documented. This is not a shortcut: it is the only way to contract over property whose value is an opinion.
One then has to read what market wordings actually say, because they are more candid than practice. They state that the value is fixed for the purposes of the contract only, with no representation as to the price the property would fetch if sold. In other words, the parties do not claim to have found the true value: they choose a number they agree to hold as true, so as to be able to bind. It is a convention, not a finding, and that distinction carries the whole module. A convention cannot be challenged on the ground that it departs from reality, since it never claimed to describe it.
The consequence meets the indemnity principle head on, and that is deliberate. If the artist's standing has fallen thirty percent between inception and the fire, the insurer still pays the agreed amount, and the insured receives more than a sale would have produced. Some wordings go further and follow a rise beyond the stated figure. The line therefore knowingly departs from a rule that governs all the rest of the insurance contract, and it does so because the alternative, a contested appraisal of a destroyed work, produces litigation rather than justice.
Then comes the point almost nobody reads before needing it. Agreed value governs TOTAL LOSS. It does not govern partial loss. On a tear, a scratch, a tide mark from damp, the contract does not pay a fraction of the agreed figure: it pays the cost of restoration, then the depreciation that remains after it. Depreciation is expressed as a percentage, and a percentage requires a base. Depending on whether the contract computes it on the agreed value or on the market value of the day, the indemnity changes entirely once the two have diverged. A collector insured forty percent above the market is therefore perfectly protected against fire and much less so against a tear, which is the exact opposite of the intuition.
That asymmetry produces two opposite pathologies, and they are not corrected the same way. Over-valuation, when it is old and the market has fallen back, opens a gap between what the object is worth and what its destruction pays, and that is not a textbook hypothesis in a line where the object often sits alone in a locked room. Under-valuation only shows on a partial loss, where too low a base erodes a depreciation that is nonetheless real. Between the two, the only discipline that holds is periodic revision of the agreed figure, supported by a dated appraisal, because an agreed value is a photograph and a photograph ages.
That leaves what agreed value never buys. It removes the argument about the amount, not the argument about whether a loss occurred, nor the one about how the damage is classified, nor the one about who owns the object. A work claimed by the heirs of a dispossessed family stays claimed whatever figure the contract carries. A work whose attribution collapses keeps its agreed value without having suffered any damage at all. The convention bears on a number, and a number only answers the question put to it. Everything that follows in this certification turns on what agreed value leaves open.
A collector insures a painting in January 2021 for an agreed value of 2 million euros, on the strength of a December 2020 appraisal. By 2024 the artist's standing has fallen back and the work would trade around 1.4 million. The contract provides that on partial damage the insurer pays the cost of restoration plus the depreciation assessed after it, capped at 10% of the sum insured. Two scenarios arise. In March, a fire destroys the painting. In June, in a variant of the same file, the painting is not destroyed but torn during a move: restoration costs 60,000 euros and the appraiser assesses residual depreciation at 8%. The collector assumes both scenarios settle on the same base, since it is the same policy and the same work.
The two scenarios do not settle on the same base at all, and that is the mistake a competent professional makes because the policy displays only one figure in large type. On the fire, agreed value governs: the insurer pays 2 million with no appraisal, even though the market would only pay 1.4 million. The 600,000 gap is not an anomaly, it is the accepted product of a convention that departs from the indemnity principle, and there is nothing to argue about. On the tear, agreed value governs nothing: restoration is paid, 60,000 euros, then depreciation of 8%. The question that decides the amount is then the base, and nothing else. On the agreed value, 8% is 160,000 euros. On the market value of the day, it is 112,000. The wording does cap depreciation at 10% of the sum insured, that is 200,000 euros, which bites in neither case and creates the illusion that the cap settles the matter: it settles only the ceiling, not the base. The 48,000 difference therefore turns on a sentence the collector never read, and the paradox is complete: he is over-protected on the loss that destroys and under-protected on the one that damages. Revising the agreed value, which he neglected because the market was falling and he saw no interest in it, would have brought the two bases together and removed the question altogether.
- 01Agreed value is a convention, not a finding: wordings themselves say it holds for the purposes of the contract only, with no representation as to sale price.
- 02It governs total loss and nothing else: on partial damage you return to restoration cost plus depreciation.
- 03Depreciation is a percentage, so it has a base: agreed value or market value of the day, and the gap between them decides the indemnity.
- 04An agreed value is a photograph and a photograph ages: without dated revision you are over-protected on fire and under-protected on a tear.
- 05It removes the argument about the amount, never the one about whether a loss occurred, how the damage is classified, or who owns the object.