Pavel Huerta, Swiss Re

Reinsurance: Fifteen years after Tohoku: preparing for the unknown

Pavel Huerta, Swiss Re

March 11 2026

The Great East Japan Earthquake and ensuing tsunami on March 11, 2011, was first and foremost a human tragedy, claiming nearly 18,000 lives and inundating over 500 square kilometres of coastline.

For the insurance industry, it was also a structural stress test. It forced us to look beyond vulnerability curves and ask a harder question: what happens when a complex system fails?

Fifteen years on, our technical capabilities are far more advanced.

Tsunami is no longer an afterthought. Catastrophe models now recognise that several connected segments of a major fault can rupture simultaneously – even off the Tohoku coast – producing powerful shaking and destructive waves.

Monitoring networks are stronger. Hazard envelopes are broader. Vulnerability assumptions are calibrated against far richer post-event data.

The models have improved. But the most important lesson from Tohoku was not about wave height – it was about uncertainty.

The scale of the shock

The 2011 event generated roughly US$35 billion in insured losses and around US$210 billion in total economic losses, making it the costliest natural catastrophe in history at the time.

Despite Japan’s stringent building standards limiting structural collapse, 2011 became a clear peak loss year for the global reinsurance sector.

For Swiss Re, the net impact from the earthquake and tsunami totalled approximately US$1.2 billion. Yet claims were paid, primary insurers remained solvent, and capacity was available at renewal.

These figures are more than statistics. What began as a domestic disaster became a globally diversified financial event – illustrating how international reinsurance capital absorbs extreme losses, stabilises cedents amid uncertainty, and preserves market function after systemic shocks.

Through multiple peak years, reinsurance has remained a reliable provider of capacity – not only in favourable markets, but precisely when capital is needed most. Continuity is not incidental; it is the core function of disciplined, globally diversified reinsurance.

The Illusion of precision

Catastrophe models are indispensable. They quantify tail risk, structure programs and optimise capital and negotiate coverage with discipline.

But they are not the reality – they are structured representations of what we currently understand.

Tohoku and other peak loss years did not invalidate modelling; they tested its boundaries.

The 2011 event showed that even in one of the world’s most earthquake-prepared nations, hazard assumptions can be exceeded. It underscored the need for stress testing beyond consensus scenarios.

At the time, few experts expected an earthquake of such magnitude on the Japan Trench off Tohoku; prevailing estimates capped the potential in the low‑8 range.

Coastal defences, designed for one level of protection, were overtopped. A nuclear incident at the Daiichi Power Pant reshaped energy policy. Semiconductor plants thousands of kilometres away suddenly mattered to insured losses.

Isolated perils in a connected world

Our industry is organised around peril categories – earthquake, wind, flood, wildfire. Even when addressing secondary perils, we often remain within a single hazard family.

The world, however, does not operate that way.

Today’s exposures are geographically concentrated, operationally interconnected, and increasingly reliant on digital infrastructure.

Power, telecoms, transport and water systems can fail together. Supply chains are optimised for efficiency, not resilience. Financial markets amplify volatility; social behaviour can alter loss outcomes.

The critical question after a major event is no longer how large the earthquake was. But how did the system respond? That distinction matters when allocating capital at the extreme.

Uncertainty is structural

Every major catastrophe in the last two decades has revealed a blind spot. Tohoku highlighted tsunami underestimation and infrastructure interdependency.

Christchurch, which caused insured losses of ~US$30 billion, exposed the scale of liquefaction risk, significantly increasing residential and infrastructure claims.

The Thailand floods, which caused insured losses of ~US$20 billion, revealed global supply chain concentration. Major impacts on electronics, automotive and hard drive manufacturing caused global business interruption far beyond Thailand.

The Covid-19 pandemic, which caused estimated insured losses of ~ US$40- 50+ billion globally, demonstrated non-physical business interruption exposure.

Recent extreme weather events have shown how hazard clustering challenges independence assumptions.

These were not failures of the modelling. They were reminders that models operate within defined boundaries. Reality does not.

The next surprise is already somewhere in today’s portfolio. We simply do not yet know its form. That is not cause for alarm. It is a reminder of what reinsurance is designed to do and what it should not do.

Capital is for the unknown

Reinsurance exists to absorb volatility beyond the expected. Primary insurers manage frequency and moderate severity within diversified; reinsurers sit at the edge – providing capital where uncertainty is greatest.

If tail risks were fully known, reinsurance would not be required. The tail remains long precisely because uncertainty is structural, with implications for how we think about return periods, correlations, event independence and aggregate management in a world of compound risk.

Capital models rely on structured scenarios. But capital buffers must also recognise that the next systemic interaction may not resemble the last. The cost of capital at the extreme is therefore not just a function of expected loss. It is a function of uncertainty around that loss

Manage the extreme, not the mean

In a competitive environment, it is tempting to anchor on recent losses and model averages. But catastrophe risk is defined by extreme, peak years.

The real test is what happens in the tail. That means looking beyond expected losses and asking harder questions: how much capital is exposed in a truly severe scenario? How much uncertainty sits around the model? Where might our assumptions be wrong?

A 1-in-200-year loss is not a precise number – it is an estimate in a world that keeps changing.

Ignoring uncertainty around severity, interconnectivity and how fragile interconnected systems can become, or how losses can spread beyond the original footprint, runs the risk of underestimating the exposure we are taking on.

This is not theoretical. It is fundamentally about protecting capital so that we can continue to be there when clients need us most.

A business imperative

The issue is not whether models are useful, but how much confidence we assign to their precision when allocating capital at the edge.

Fifteen years after Tohoku, our view of tsunami risk is stronger. Datasets are richer. Vulnerability modelling is more robust. The event also prompted further strengthening of Japan’s already high disaster mitigation standards, including the incorporation of very low-frequency, high-impact scenarios into national planning.

Yet Japan remains one of the world’s most earthquake‑prone countries, and the potential for another peak seismic loss year is ever-present.

Reinsurers have a vital role in preparing for what comes next. By working closely with insurers, corporates and governments, we can design solutions that expand earthquake coverage and gradually narrow protection gaps over time – combining capital, expertise and partnership to strengthen resilience.

The lessons from the Great East Japan Earthquake remain as relevant as ever: hazard models must evolve, mitigation must strengthen, and risk transfer must scale with exposure. Peak loss years are not anomalies; they are structural features of catastrophe risk.

Reinsurance exists to absorb shocks, enable recovery, and sustain confidence when uncertainty is greatest. The Great East Japan Earthquake was a defining test of that purpose. Our responsibility now is to ensure that when the next peak event arrives, the system – and society – is even more resilient.

The next major loss will not look like 2011. It never does.

Our task as reinsurers is to ensure that discipline, structure and capital allocation reflect that reality.

That is not pessimism. It is prudence – and the foundation of a resilient reinsurance market.

Pavel Huerta is the head property underwriting globals for APAC at Swiss Re.

MORE FROM: Comment
Partner Content
Join the mailing list Designed to help you stay ahead, our daily newsletter provides a round-up of the latest news and insights, delivered straight to your inbox.
Print Edition
Subscribe to receive our quarterly print magazine.