When waiting becomes a risk

In insurance, caution is a cardinal virtue.

We value hindsight, long time horizons, data and statistical certainty. A product that is too new is, by definition, considered difficult to assess. It seems wiser to wait. To observe. To let time do its work and be confident about the measures to take.

This is a logic that can be found in most steering committees, regardless of the type of risk, where the decision is often made to wait for another quarter “to gain more hindsight”.

This posture is widely shared. It is reassuring. It conveys a sense of seriousness.

But in practice, it raises a real question: what if waiting became, in some cases, the riskiest decision? What if caution ended up costing more than measured action?

‍

The misunderstanding around “hindsight”

There is a misunderstanding in the way the market approaches the analysis of insurance products.

Two very different things are often confused:

• not yet having statistical certainty,
• and believing that no meaningful analysis is possible at that stage.

However, my experience at Wakam, WTW and Seyna has taught me that a recently launched product is not an unreadable product. It simply needs to be analysed differently.

From the very first months, certain dynamics already emerge: volumes, underwriting profiles, early loss signals, the way the product is distributed or understood. These elements do not yet allow definitive conclusions to be drawn about long-term profitability, but they do allow alerts to be raised, discrepancies to be identified and initial assumptions to be challenged.

Waiting “until there is enough hindsight” can sometimes mean postponing decisions that could have been taken earlier, differently and with far fewer consequences.

This delay is not neutral: the longer a decision is postponed, the more room for manoeuvre is lost and the greater the financial impact becomes.

‍

When inaction disguises itself as caution

In many cases, this waiting period is presented as caution. In reality, it can also be a form of inaction.

A product that performs commercially, generating the expected level of volume — neither more nor less — creates little incentive to look closely, early on, at what is happening beneath the surface. As long as premiums are coming in, the sense of urgency remains low. And the more time passes, the more imbalances can settle in without being truly questioned.

The paradox is that when analysis finally takes place, it often comes too late. At that point, decisions become abrupt: product closures, market withdrawals, termination of partnerships, significant price increases. Not because the product was fundamentally flawed, but because early warning signs were not addressed at the right time.

For example, with certain individual health insurance products distributed through brokers, we chose to wait. All stakeholders felt that the performance of this type of portfolio could not be assessed so quickly. By the time hindsight was finally considered sufficient, the situation had already deteriorated too far to allow for gradual adjustment, leading to radical decisions that could very likely have been avoided earlier.

Conversely, on other recently launched health products, very strong signals appeared in the first few months (very high volumes, atypical profiles, early technical indicators). This led to early alerts, more detailed analysis of underwriting patterns and the implementation of initial adjustment measures. This early intervention helped prevent the accumulation of risk.

The risk in these situations is not making a mistake too early. The risk is making no decision for too long.

‍

Repricing: the first reflex that often rings hollow

When a product starts to deteriorate, the first lever activated in an emergency is almost always the same: repricing. Prices are adjusted, cover is reviewed, deductibles are modified. It is logical. It is visible. It is measurable.

But this reflex often masks a deeper issue. Pricing becomes the default lever when time has not been taken earlier to understand what was really happening within the product.

In some cases, imbalances do not stem from poorly calibrated pricing, but from the way the product is distributed, managed or understood by the market. Policyholder behaviour, operational processes, data quality or claims management can sometimes play a far more decisive role than a new price point.

On breakdown or damage products, the performance of a repair network can significantly alter a product’s profitability. Profitability largely depends on how claims are handled (repair rather than replacement, intervention times, loss adjusting costs). These are operational levers that often carry more weight than any pricing change.

Fraud and the recovery of unpaid premiums also play a central role. On health portfolios comparable to those mentioned above, significant performance gaps were explained not by pricing levels, but by the handling of unpaid premiums and fraud detection, revealing process differences that were just as decisive as price.

Repricing can give the illusion of action. But repricing too late, without having explored other levers, often means correcting symptoms without addressing root causes.

It is also a sign that analysis has come after the fact, under pressure, rather than at a point where it could genuinely have guided decisions.

Acting early makes it possible to iterate through small corrective measures: by quickly testing adjustments (distribution, acceptance rules, claims handling or controlled pricing experiments), operational feedback is gathered that improves the response and limits the cost of a later, large-scale correction.

‍

Deciding earlier does not mean moving too fast

Deciding earlier does not rely on isolated intuition or a single magic indicator. It requires an organisation capable of quickly combining several perspectives on the same product.

In my experience, one of the main obstacles lies in siloed ways of working. Data is analysed in one place, actuarial work in another, claims management elsewhere. Exchanges exist, but they are sporadic and often too late. Yet early warning signs never appear within a single perimeter. They emerge at the intersection of several areas of expertise.

For example, on rent guarantee insurance products, the combined analysis of data, actuarial insights and claims management practices made it possible to identify performance gaps very early on, linked to recovery rates — a key indicator that materialises over time but strongly conditions profitability.

Conversely, in pet health insurance, comparative audits could have enabled us to identify more effective management practices and draw lessons from them. Because this cross-analysis did not take place early enough, interventions were limited to pricing adjustments, whereas observation and benchmarking could have opened up other avenues.

This is why bringing data, actuarial and claims management teams together within the same structure fundamentally changes the ability to manage products. Not because it creates new skills, but because it forces continuous exchanges. Data quality, understanding of operational processes and actuarial risk analysis are no longer addressed sequentially, but together.

This approach comes with another prerequisite: the speed and quality of data integration. Without reliable data that can be used quickly, any ambition to make early decisions is illusory. Decisions are not taken too early because of haste. They are taken too late because information arrives too slowly or in a poorly structured way, preventing early warning signs from being analysed.

This is often where everything is decided: in the ability to circulate information before imbalances become structural.

Ultimately, deciding earlier does not mean moving too fast. It means being able, as both insurer and broker, to identify sooner what deserves to be questioned.

‍

Conclusion

In insurance, hindsight remains essential. However, waiting is never a neutral decision.

For insurers and brokers alike, the challenge is not to decide without data, nor to claim an understanding of a product within its first few weeks. The challenge is to initiate dialogue early, without letting time decide in our place.

In a market where decision cycles are long and cumulative effects significant, waiting too long is not a neutral posture. It is a strategic choice, with very real consequences.

‍