Peak Re | From climate modeling to market opportunity: gorging a new clarity on Southeast Asia’s climate risk

September 1 2025

Southeast Asia’s rapid economic growth stands in stark contrast to its persistent insurance protection gap. This is not a simple problem of capital availability; it is a crisis of clarity.

The immense, volatile, and persistent nature of climate risks such as typhoons and floods in the region creates a fog of uncertainty that hinders the confident deployment of capital.

To unlock this vital market and serve these vulnerable economies, the industry must move beyond its current tools and forge a new, more accurate understanding of the risk.

The region’s economic dynamism is undeniable, with countries like Vietnam, Indonesia, and the Philippines experiencing robust GDP growth rates averaging 5-7% annually in recent years. Yet, this progress is overshadowed by a staggering insurance protection gap.

According to recent estimates, the Asia-Pacific region could account for nearly half of the global uninsured risk, potentially reaching US$1.86 trillion by 2025. In emerging Asia-Pacific nations, the total protection gap has doubled since 2013 to US$740 billion in 2023, driven largely by underinsurance against natural catastrophes.

Southeast Asia, in particular, bears the brunt of weather-related extremes, where the proportion of uninsured economic losses is high.

Climate change, with global temperatures already 1.1 degrees Celsius above pre-industrial levels and projected to rise 2.5–3 degrees Celsius by 2100 under scenarios where current mitigation efforts, technological advancements, and policy enforcement remain unchanged, exacerbates this vulnerability, intensifying the frequency and severity of extreme weather events in the region.

Typhoons are forming closer to coastlines, intensifying more rapidly, and persisting longer over land due to warmer oceans absorbing more water vapor and heat.

Recent events like Typhoon Yagi in 2024, which devastated Vietnam, the Philippines, and surrounding areas with unprecedented flooding and landslides, highlight the escalating threats.

 

According to report by World Meteorological Organization (WMO), over 80% of Asia’s disasters in recent years have been flood- and storm-related, resulting in more than 2,000 fatalities and affecting nine million people directly.

Without clearer risk insights, reinsurers and insurers struggle to price policies accurately, leaving communities exposed and capital sidelined.

The foundational flaw: why current industry models fall short

For years, the reinsurance industry has built its risk views on the foundation of catastrophe models.

While these models have served as valuable tools, they could benefit from further alignment with advancing science and the evolving climate, to provide an even more refined view of risk. Building on these foundations with updated approaches could help minimise potential surprises and support more precise pricing.

Traditional catastrophe (cat) models rely on historical data and statistical probabilistic simulations to estimate losses from events like typhoons and floods. While this approach has provided a baseline for risk assessment, these models often fail to fully incorporate the dynamic effects of climate change, such as shifting storm trajectories and increased intensity.

Transitioning to physics-based climate modeling offers the potential to significantly improve accuracy by incorporating dynamic climate processes.

However, current climate models face substantial challenges due to high computational costs, which limit the ability to produce large ensemble simulations necessary to capture rare “black swan” events – extreme occurrences that fall outside typical historical patterns and pose significant risks in regions like Southeast Asia.

This opportunity for enhancement manifests in real-world challenges: insured losses from hurricanes, floods, wildfires, and severe storms have exceeded US$100 billion annually since 2020 and are projected to surpass US$200 billion in 2025

In Southeast Asia, low insurance penetration – often below 5% for property risks – amplifies the gap, as evolving risks encourage a need for more adaptive models to facilitate capital inflow.

The result? Opportunities to refine premiums, stabilise portfolios, and expand coverage in high-risk areas, helping to narrow the cycle of under protection.

The new imperative: we must invest to update our risk views in more efficient ways

The insurance industry is at an inflection point. We can no longer afford to be passive consumers of insufficient modeling. The new imperative is to become active creators of risk intelligence.

This requires a fundamental shift in mindset and capability – from simply using models to actively forging new tools that can deconstruct, analyse, and rebuild climate data into a view that is purpose-built for the business of risk transfer. This is the difficult but necessary work required to achieve true clarity.

Traditional physics-based climate models struggle between computational costs and meaningful statistics.

The idea setting would be running large-ensemble, high-resolution climate models, whose resolution would be sufficient to resolve the dynamics of extreme weather, and whose ensemble size is large enough to capture black swan events.

However, high computational costs prevent us from achieving those two things simultaneously. Recent development of data-driven climate models utilises the power of AI and has created AI emulators of weather and climate.

Huge ensembles using such AI climate emulators have been conducted for capturing extreme events, whose probability is so low that that they can be easily missed in deterministic, physics-based model simulations.

Of course, AI models learn the physical principles from data, which still need to be generated by traditional, physics-based weather and climate models, so some ways of combining AI and physics-based models hold the key to revolutionising catastrophe models in the context of climate change.

Scientists around the world have been working for such research areas, but to accelerate progress and deliver mature products for operational use, the industry should join this adventure in this early stage and collaborate with scientists.

The breakthrough: new clarity unlocks market opportunity

Achieving this new clarity through advanced climate modeling directly translates to commercial breakthroughs for the reinsurance industry.

· Precision underwriting: It allows the market to move from broad assumptions to granular, accurate pricing, creating more sustainable and technically sound portfolios. In anticipation of AI-enhanced models in near future, underwriters can assess risks at the property level, incorporating real-time climate data to price policies that reflect true exposure, reducing adverse selection and improving profitability.

· Strategic capital allocation: With a truer understanding of tail risk, capital can be deployed with greater efficiency and confidence. Clearer views of black swan events enable better reserve setting and reinsurance placements, minimising surprise losses and optimising returns on capital.

· Sustainable expansion: It provides the trusted foundation needed for the industry to expand its footprint in Southeast Asia, developing new products and partnerships to meaningfully close the protection gap. Initiatives like parametric triggers for flood or typhoon events, backed by AI-validated data, can attract new clients in underserved markets, fostering inclusive growth.

Our shared future: A commitment to the path forward

Forging a new perspective on climate risk is a significant yet vital endeavor for the future. It demands investment, specialised expertise, and an unwavering dedication to innovation.

At Peak Re, we are actively collaborating with academia to integrate the latest scientific insights on climate change and cutting-edge technological advancements into practical risk assessment tools and models.

This effort translates innovative climate risk modeling into actionable strategies, empowering not only our organisation but also our clients. We support insurers by refining their portfolios, enhancing underwriting accuracy, and crafting tailored solutions to address Southeast Asia’s unique climate challenges.

However, this journey transcends any single company. Our industry’s collective success and social responsibility hinge on delivering this enhanced clarity to the world’s most complex risks in context of climate change. By uniting in this challenge, we can forge a more resilient future and tap into the vast opportunities that await.

References

1. PwC (2025). “Protection gap could reach US$1.86tn by 2025, with Asia-Pacific region accounting for almost half of all uninsured risk”. https://www.pwc.com/bm/en/press-releases/insurance-in-2025-and-beyond.html

2. insureERM (2023). “Emerging Asia-Pacific protection gap doubles since 2013 to $740bn.” Available at: https://www.insureerm.com/news/emerging-asia-pacific-protection-gap-doubles-since-2013-to-740bn.html

3. Garner, A.J., Samanta, D., Weaver, M.M. et al. Changes to tropical cyclone trajectories in Southeast Asia under a warming climate. npj Clim Atmos Sci 7, 156 (2024). https://doi.org/10.1038/s41612-024-00707-0

4. Geneva Association. “Safeguarding Home Insurance: Reducing exposure and vulnerability to extreme weather“, https://www.genevaassociation.org/publication/climate-change-environment/safeguarding-home-insurance-reducing-exposure-and

5. NOAA (National Oceanic and Atmospheric Administration). “Hurricanes and Climate Change.” This foundational resource explains how warmer sea surface temperatures and increased atmospheric moisture contribute to more intense storms. Available at:

https://www.gfdl.noaa.gov/global-warming-and-hurricanes/

6. WMO (World Meteorological Organization) (2024). “State of the Climate in Asia 2023.” This report details the impacts of extreme weather, including floods and storms, across the continent. Available at: https://wmo.int/media/news/asia-was-worlds-most-disaster-hit-region-2023-due-weather-climate-and-water-related-hazards

Luxi Zhou,

Vice President,

Peak Re 
  XiaoMing Shi

Professor,

HKUST
Partner Content