Reputational Risk in the Age of Social Media

I.The Asset That Became Dominant and Hardest to Insure

The shift to the intangible

The value of companies has changed in nature within a generation. Where the balance sheet of a twentieth-century industrial company rested on factories, inventories and machines, that of a contemporary firm lies essentially in intangible assets, brands, data, know-how and reputation. According to several valuation services, these immaterial assets today represent 85 to 87% of the value of large companies, and the world's fifty largest capitalizations alone would amass around 11 trillion dollars of intangible assets¹. Reputation holds a central place in this set, to the point of being regularly ranked among the foremost risks by risk-management departments. The most precious asset of the modern firm is no longer what it owns, but what people think of it, and this inversion suffices to overturn the foundations of its insurance.

The paradox of insurability

A paradox follows. The more an asset becomes decisive for the value of the firm, the more one would expect it to be covered, yet it is precisely the opposite that is observed for reputation. The reputational risk insurance market exists and is growing, from around 11.8 billion dollars in 2024 toward a projection of 30.8 billion in 2033, but it remains embryonic relative to the sums at stake², out of all proportion with the value it claims to protect. A sector leader bluntly sums up this difficulty by observing that financial markets forgive an uninsured natural catastrophe, but never an operational failure that tarnishes reputation³. The coverage of intangible assets thus lags considerably behind that of physical goods, a legacy of an insurance designed for walls and machines. To understand why requires returning to the very nature of reputation.

The legacy of an insurance of things

This lag is not a mere failure of modernization, it stems from the very genealogy of insurance. The discipline was forged around material risks, the fire of a warehouse, the wreck of a ship, hail on a harvest, and its fundamental concepts, the insured value, the loss, the appraisal, presuppose a physical object whose loss can be observed and quantified. The whole toolkit of the insurer, from the policy to the settlement, presupposes an asset that can be seen, owned and valued. When the value of the firm migrated toward the immaterial, this toolkit found itself disarmed, not for lack of will, but because its very categories no longer applied. To insure reputation is therefore not a natural extension of the trade, it is a break with the premises on which it was built, and this break explains the slowness and timidity of the offering.

II.Why Reputation Resists Insurance

A verdict, not a thing

Classical insurance was built to cover assets that share three properties, they are tangible, they are owned, and they are measurable. A building has a physical substance, belongs to its owner and possesses a value that the appraiser can quantify. Reputation satisfies none of these three conditions. It is intangible, for it has no material substrate that one could repair or rebuild. It is not owned, for it resides not in the firm but in the minds of its customers, employees, investors and the public, that is, in a multitude of third parties over which the firm has no property right. It is finally barely measurable, for want of a shared metric and a stable reference value. Reputation is a verdict that others render on you, and the company holds no title to it.

Repairing the irreparable

This singular nature defeats the indemnity mechanism. Insurance restores a lost value by paying a sum that allows the destroyed good to be rebuilt, but one does not rebuild an opinion as one rebuilds a warehouse. No indemnity compels the public to reform the favorable judgment it held, for that judgment is not for sale and cannot be bought back. The insurer can finance the means of regaining trust, it cannot guarantee the result, which depends on outside and free actors. This fundamental powerlessness joins a theme that the article in this series on works of art will take up, namely that the value of an asset may be a matter of opinion rather than of fact, and that an opinion is not indemnified like physical damage¹¹.

III.Social Media, Accelerator and Amplifier

From the episodic incident to permanent exposure

If reputation has always been fragile, social media has transformed the scale and speed of the risk. A criticism that would once have remained local and fleeting can now reach a global audience within hours, escape all control and leave a lasting trace. The speed of propagation has become the primary characteristic of contemporary reputational risk, and movements of collective denunciation, of which the MeToo phenomenon, which began in the United States in 2017, offered the most striking example, have shown how the act of a single individual can trigger an instant trial by media and ruin a reputation built over decades⁵. The firm has thus moved from an episodic exposure, punctuated by identifiable crises, to a permanent exposure, where the public's judgment forms and deforms continuously, at every instant and on every platform.

Manipulation and the fake

To this acceleration is added an alteration of the very matter of reputation, information. The public's growing mistrust, the manipulation of opinion by state and non-state actors, and the proliferation of false content combine to make the risk harder still to grasp. A reputation can now be attacked not by a proven fact, but by a fabrication, a doctored video or a coordinated disinformation campaign, against which the truth of the facts offers a slower protection than the virality of the lie. This dimension, which another article in this series devoted to synthetic content will extend, adds to reputational risk an adverse and intentional component, where the firm is no longer only exposed to its own mistakes, but to deliberate attacks exploiting the mechanics of the platforms¹².

Measuring the invisible

This omnipresence of the risk has paradoxically opened the way to a beginning of measurement. The same platforms that propagate crises produce a massive data trail, where algorithmic analysis attempts to read the state of reputation in real time. Specialist providers aggregate open sources and apply artificial intelligence to detect weak signals, identify the leading indicators of a crisis and quantify the evolution of sentiment toward a brand⁶. This quantification, still imperfect, is the condition of any insurability, since one can neither price nor indemnify what one cannot measure. Social media therefore plays a dual role, at once the principal vector of reputational risk and the source of the data that allows its measurement to be sketched, in a movement where the instrument of harm becomes also the instrument of its assessment.

IV.The Obstacle of Endogeneity and Moral Hazard

A risk one carries within oneself

The deepest difficulty, however, lies in a property that radically distinguishes reputation from classical perils, its endogenous character. A hurricane or an earthquake are external events, independent of the insured's will, which makes them insurable. The reputational crisis, on the contrary, most often arises from the firm's own behavior, from the misconduct of an executive, a questionable decision or, more subtly, a clumsy response to an initial incident. Practitioners insist on this point, reputation is played out less in the triggering event than in the way the firm reacts to it, and it is not so much what it says as what it should have left unsaid that often propagates the crisis⁴. Reputational risk is thus, for a large part, a risk that the firm carries within itself.

The limit of moral hazard

This endogeneity collides head-on with the principle of moral hazard, which forbids insuring a risk whose realization depends on the insured's conduct. One cannot cover a firm against the consequences of its own faults without prompting it to laxity, and that is why reputation policies strive to distinguish the suffered crisis from the provoked one, an exercise of great difficulty given how far the two intertwine. Reputation thus sits on the frontier of the insurable, divided between an exogenous part, external attacks and accusations, which one may envisage covering, and an endogenous part, the insured's own faults, which by nature escapes risk pooling.

This endogenous part also goes beyond the initial fault and extends to the management of the crisis, where the firm remains master of its choices and therefore responsible for their effect. A late response, a clumsy denial or a communication perceived as insincere can turn a minor incident into a lasting disaster, so that the magnitude of the loss depends largely on decisions made after the triggering event. This plasticity of the damage, shaped by the insured's behavior even in its aftermath, is foreign to classical perils, where the extent of the harm is fixed by the event itself. It makes pricing uncertain and settlement disputable, for it becomes hard to distinguish the share of the loss attributable to the attack from the share aggravated by the firm's poor reaction.

V.The Market's Responses and Their Limits

Two imperfect approaches

Faced with these obstacles, the market has developed two families of responses, which circumvent the difficulty rather than resolve it. The first, indemnity-based, does not cover reputation itself but the costs of responding to the crisis, the fees of communications experts, legal advisers and brand-rehabilitation specialists, to which are sometimes added business interruption cover and rapid liquidity. Insurers such as Beazley, Liberty or AIG have structured such offerings, often backed by consultants and real-time data monitoring¹⁰. The second approach, parametric, developed in particular by Steel City Re, pays a predefined sum when reputation indicators, sometimes combined with the variation of the share price, cross an agreed threshold⁷. Each provides a real response, but none truly insures reputation.

Insuring the echo rather than the thing

The limits of these arrangements are structural. The indemnity approach finances the reaction to the crisis, not the restoration of the reputation, which remains uncertain. The parametric approach, by triggering on the share price, insures the financial echo of the reputational harm rather than the harm itself, and reopens the question of basis risk, since the price may fall without a reputational crisis or remain stable despite a degraded reputation. To this is added a decisive capacity constraint, for the usual limits of these policies, of the order of 10 to 50 million dollars, remain negligible relative to the intangible assets of a multinational, which run into billions⁸. Reputation insurance therefore protects a fraction of the risk, and only by proxy, by covering its financial and operational manifestations rather than its substance.

To these limits is added an immaturity of the market itself. For want of a dominant model and of comparable historical data, reputation policies differ profoundly from one insurer to another, in their trigger, their scope and their valuation method, to the point that there is scarcely a shared standard. This heterogeneity complicates purchasing for firms, which struggle to compare offerings built on distinct logics, and it hinders the formation of a reinsurance market capable of providing the missing capacity. As long as the measurement of reputation and the definition of the loss are not further standardized, these products will remain bespoke constructions, useful to some firms but unable to reach the scale and liquidity of the major non-life lines, toward which the market nonetheless aspires to raise them.

VI.What Can Insurance Really Cover?

From indemnity to prevention

At the end of this analysis, it appears that insurance cannot insure reputation in the strict sense, because reputation is a verdict that money does not reverse. This observation, far from condemning these products, shifts their function. The value of reputation policies lies less in risk transfer than in risk management and prevention, for they provide above all access to crisis consultants, permanent monitoring of weak signals and a discipline of governance. The insurance component is almost secondary relative to the service component, and it is in this that these offerings come closer to a subscription to a response capacity than to a classical indemnity. They fall under the letter G of the ESG acronym, that of governance, in that they institutionalize the firm's preparation for its own crises⁹.

This shift from indemnity to service also explains the market's growth despite its limits. If firms buy these policies even though they do not restore reputation, it is because they find value elsewhere, in access to a crisis expertise hard to build internally, in the discipline that underwriting imposes, and in the signal of seriousness that a prepared governance sends to investors. The reputation policy is thus worth as much for what it organizes upstream as for what it pays downstream, and its usefulness is measured less by the eventual indemnity than by the resilience it helps to build. It is an insurance whose principal benefit is realized before the loss, and not after, reversing the usual logic of the trade.

Conduct, the only cover that holds

There follows a conclusion that goes beyond the insurance market alone. The best protection against reputational risk is conduct, for a reputation is defended first by the acts that found it and by the quality of the response to crises. Insurance can accompany this defense, finance its means and cushion certain of its financial effects, but it cannot substitute for it, and any firm that believed it could buy its reputational peace of mind through a contract would mistake the nature of the risk. Reputation thus marks a frontier of insurability of a new kind, where insurance ceases to indemnify a loss in order to finance a behavior, and where the insurer's role slides from that of payer toward that of prevention partner. As the economy dematerializes and value lodges in the intangible, this frontier will become central, and the question will no longer only be how to insure reputation, but to admit that certain assets, because they reside in the minds of others, are protected by virtue more surely than by contract.

This lesson holds well beyond reputation, for it foreshadows the challenge that the whole set of immaterial assets poses to insurance. Data, the brand, trust, market position share with reputation this quality of existing less in a balance sheet than in a relationship, and of resisting for this reason the categories of an insurance born for tangible goods. The sector will have to invent, for these assets of a new kind, hybrid instruments combining risk transfer, prevention service and algorithmic measurement, in the image of what reputation policies already sketch. Reputation is thus only the first of these immaterial assets to test the limits of the insurance model, and the way the market learns to cover it will prefigure its capacity to protect an economy that has become, for the most part, invisible.