Recent floods in major cities around Southeast Asia and other parts of the world have reopened the conversation on flood coverage in insurance products and the means by which they will be financed. In just the past year, Malaysia, Pakistan and South Korea have all witnessed the worst floods to hit their shores in decades. Economic losses from such floods have grown steadily, with losses estimated to increase drastically in the future due to increased urbanisation, climate change and poor flood prevention measures. As it stands, it is estimated that only 18% of all economic losses¹ from floods in the past decade were insured. The recent flood in Kuala Lumpur had only 36% of the total economic losses insured.² With the lack of flood coverage in fire, motor, business interruption and crop insurances, these costs will either be borne by the government or the affected societies themselves, raising the question of why flood coverage is so severely underinsured despite the emerging threat.

Why is there underinsurance?

The most apparent problem is the consumers’ lack of awareness of possible increasing flood risks. Flood coverage benefits are often understated among consumers who have not yet been threatened with flood losses. Insurance practitioners we interviewed indicated that this is prevalent amongst commercial property owners who generally opt out of the flood add-on cover to reduce their insurance expenses. Residential properties, except those protected by an all-risk cover required by their financing, are generally uninsured as well. In Malaysia’s case, the awareness for flood cover among vehicle owners had just started to increase in the aftermath of the 2021 floods. Beyond a lack of awareness, there have also been complaints from policyholders that the coverage for floods is often confusing to understand, especially for commercial cover. Flood cover could also be tedious to underwrite, such as in the case of flood cover for business interruption where a projection of estimated financial statements might be required by insurers to offer coverage. As such, many enterprises choose to remain uninsured despite the significant financial risks; commercial losses, in particular consequential losses, contribute the majority of uninsured losses.

The second reason for flood underinsurance is the high cost of cover. Once affected by flood, a location or region could be identified as being vulnerable and susceptible to future floods, causing premiums to increase for policies based in the area. In such cases of flood-prone areas, insurers rely on passing the exceptionally large flood losses to reinsurers in the event of catastrophes, helping them to control their claim costs. However, after a catastrophe event reinsurers themselves are forced to increase their premium, an additional cost which will be passed on to policyholders residing in flood-prone areas eventually. Flood policy premiums in Thailand had more than doubled³ in the aftermath of the 2011 floods, which subsequently required government intervention. Without proper investments into flood mitigation systems, these areas could be subjected to increasing premiums and dwindling demand for flood cover.

Flood risks are also notoriously complex to model for insurers such as estimating the severity of damage and losses arising from floods. The range of risk factors, such as urbanisation, climate change and inadequate infrastructure against flooding, continues to expand and evolve, which requires any models used to be sufficiently dynamic to capture the changing environment. Insurers are also required to improve on any existing data collection on flood exposure. A lack of data and inadequate models causes a lack of conviction in the rates produced and additional margins built into the estimates leading to unnecessarily high premiums.

Solutions: Not without their own challenges

Each country has attempted to solve the causes of underinsurance in its own way, involving a mix of innovative and interventionist solutions, albeit with their own shortcomings.

One approach would be for governments to enforce mandatory flood coverage for properties, motor and crops. This will reduce the risk of anti-selection and allow the loss from floods to be spread over more policies. Mandatory cover would also ensure that insurers are able to gather adequate amounts of data gradually to help them price risks with more credibility. This should work to reduce the level of prudence embedded into the flood insurance premium rates. On the flipside, increased granularity and credibility in pricing could lead to greater differentiation in premium rates for homes in flood-prone areas. Furthermore, it is not straightforward, as governments will be concerned about such measures being politically unpopular.

Parametric insurance has been gaining traction as the ”go-to” solution for affordable flood cover. A fixed payment in the event a prespecified parameter exceeds a predetermined threshold presents more certainty to the insurers on the scale of losses as compared to insurance written on an indemnity basis, allowing premiums to be more affordable. However, the need for precise measuring equipment and the possibility of the fixed claim payment being insufficient to cover the losses incurred prevent parametric insurance from being a comprehensive solution. Insurers also tend to be wary of possible cash flow issues from needing to settle large numbers of claims in a short period of time as payment is guaranteed once the event has been triggered. Certain parameters, for example amount of rainfall, might also go by an average measure within a specified postcode, ignoring the possibility of very different landscapes within each postcode if the area of coverage is large.

Regulators are increasingly requiring insurers to model possible climate stress scenarios, which often involves modelling losses arising specifically from floods. Although insurers are still struggling at this stage to accurately quantify the impact of climate change over the long-term projection horizon, this encourages insurers to be aware of climate-related risks and develop internal models to estimate flood-related losses. Eventually, this could lead to greater emphasis on risk mitigation and pricing flood losses more reliably.

However, flood-prone risks will continue to face high and increasing premiums in the future without active flood management measures. To make matters worse, we may see more and more flood-prone areas emerge with increased urbanisation and climate change.

What about the flood-prone risks?

It seems inevitable that intervention would be required to ensure flood-prone risks are insured. We explore three unique solutions currently being implemented in three different parts of the world.

Flood Re was founded in the United Kingdom at a time when insurers refused to provide flood cover for high-risk properties as flood reinsurance was costly.⁴ It is effectively an industry pool providing reinsurance cover against large losses or large aggregate losses, funded by levies on all home insurance policies on top of a premium imposed on policies included in the scheme. The drawback is that Flood Re might become an unsustainable solution if flood losses continue to rise. Without a gradual improvement in flood risk management, Flood Re will not be able to maintain its current premium structure, with any increases in its premiums likely to be passed down to the policyholders. Flood Re is also only available for residential properties built before 2009, with commercial properties (with a more heterogeneous risk profile) being excluded. Although this was to encourage developers to avoid flood-prone areas in future developments, this has not proved to be effective in reducing new developments with high flood risks.

National Flood Insurance Programme (NFIP) in the US is a flood insurance programme offered directly by the government to those from high-risk or special flood hazard areas (SFHAs).⁵ In order for homeowners or small businesses to be able to purchase cover, their local communities will be required to implement floodplain management standards to control flood losses. However, the premium rates implemented in the past have proven to be insufficient, resulting in the NFIP being debt-ridden and requiring bailouts by the federal government. In light of this, the regulation enabling NFIP was amended to allow NFIP to secure reinsurance from private reinsurance and capital markets besides introducing Risk Rating 2.0 to incorporate more dynamic factors such as flood frequency and distance from water source into their pricing engine.

The Caribbean Catastrophe Risk Insurance Facility (CCRIF) acts as a multi-country risk pool.⁶ Although not specific to flood losses, the concept could be emulated between neighboring countries facing similar levels of flood threat. Acting as a pooling facility between member countries, the CCRIF offers parametric insurance for perils occurring in the Caribbean, such as catastrophic tropical cyclones, earthquakes and excess rainfall events, with pre-agreed payments made to member countries once a certain parameter threshold, such a flood level, is breached. Being a parametric insurance allows the CCRIF to provide natural catastrophe coverage at prices below non-pooled arrangements. The cover provides short-term liquidity to any affected members when the policy conditions are triggered but the inherent weaknesses of parametric insurance discussed above are present within the CCRIF as well.

The way forward

Every jurisdiction will require its own unique solution, which may take the form of preventive measures, damage control protocols in the event of flooding and financing mechanisms to help flood-prone areas deal with the economic consequences of flooding. It will therefore require all stakeholders to come together and contribute towards a sustainable solution. However, we believe insurance provides a platform for scrutinising the issues facing the flood-prone communities, i.e., the increasing frequency and intensity of floods, flood management infrastructure and the financial losses (insured and uninsured), and crystallising them into a risk score which translates to the premium rate. If flood insurance remains unaffordable despite concerted efforts to mitigate the impact of floods, measuring financing and spreading the risk and cost of floods, then perhaps it is time to question the location of those homes and businesses. We believe the insurance industry is best positioned to lead the stakeholders in discussion.

¹ Clifford, C. (1 September 2022). Flood losses to insured property are jumping drastically and only a small fraction of what’s damaged by floods is insured. CNBC. Retrieved 27 September 2022 from https://www.cnbc.com/2022/09/01/swiss-re-flood-losses-covered-by-insurance-are-rising-drastically.html.

² Ong, S. (30 March 2022). BNM: December 2021 floods highlight significant underinsurance of flood risks. The Edge Markets. Retrieved 27 September 2022 from https://www.theedgemarkets.com/article/bnm-december-2021-floods-highlight-significant-underinsurance-flood-risks.

³ A.M. Best (9 February 2012). Flood Losses Prompt Key Changes in Thai Insurance Industry. Best’s Briefing. Retrieved 27 September 2022 from https://www.ambest.com/press/021001thaifloodbriefing.pdf.

⁴ Flood Re. What Is Flood Re? Retrieved 27 September 2022 from https://www.floodre.co.uk/about-us/.

⁵ Metz, J. (17 May 2022). How Does The National Flood Insurance Program Work. Forbes. Retrieved 27 September 2022 from https://www.forbes.com/advisor/homeowners-insurance/national-flood-insurance-program/.

⁶ See the CCRIF website at https://www.ccrif.org/.