Flood protection gap: Evidence for public finances and insurance premiums

Bellia, Mario; Di Girolamo, Francesca Erica; Pagano, Andrea; Petracco Giudici, Marco Working Paper Flood protection gap: Evidence for public finances and insurance premiums JRC Working Papers in Economics and Finance, No. 2023/10 Provided in Cooperation with: Joint Research Centre (JRC), European Commission Suggested Citation: Bellia, Mario; Di Girolamo, Francesca Erica; Pagano, Andrea; Petracco Giudici, Marco (2023) : Flood protection gap: Evidence for public finances and insurance premiums, JRC Working Papers in Economics and Finance, No. 2023/10, European Commission, Ispra This Version is available at: https://hdl.handle.net/10419/283110 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Flood protection gap: evidence for public finances and insurance premiums Bellia, M. Di Girolamo, F. E. Pagano, A. Petracco Giudici, M. JRC Working Papers in Economics and Finance, 2023/10 This publication is a Working Paper by the Joint Research Centre (JRC), the European Commission’s science and knowledge service. It aims to provide evidence-based scientific support to the European policymaking process. Working Papers are pre-publication versions of technical papers, academic articles, book chapters, or reviews. Authors may release working papers to share ideas or to receive feedback on their work. This is done before the author submits the final version of the paper to a peer-reviewed journal or conference for publication. Working papers can be cited by other peer-reviewed work. The contents of this publication do not necessarily reflect the position or opinion of the European Commission. Neither the European Commission nor any person acting on behalf of the Commission is responsible for the use that might be made of this publication. For information on the methodology and quality underlying the data used in this publication for which the source is neither Eurostat nor other Commission services, users should contact the referenced source. The designations employed and the presentation of material on the maps do not imply the expression of any opinion whatsoever on the part of the European Union concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. Contact information Name: Di Girolamo Francesca Email: [email protected] EU Science Hub https://joint-research-centre.ec.europa.eu JRC135372 Ispra: European Commission, 2023 © European Union The reuse policy of the European Commission documents is implemented by the Commission Decision 2011/833/EU of 12 December 2011 on the reuse of Commission documents (OJ L 330, 14.12.2011, p. 39). Unless otherwise noted, the reuse of this document is authorised under the Creative Commons Attribution 4.0 International (CC BY 4.0) licence (https://creativecommons.org/licenses/by/4.0/). This means that reuse is allowed provided appropriate credit is given and any changes are indicated. For any use or reproduction of photos or other material that is not owned by the European Union, permission must be sought directly from the copyright holders. How to cite this report: Bellia, M., Di Girolamo, F. E., Pagano A., Petracco Giudici, M. Flood protection gap: evidence for public finances and insurance premiums, JRC Working Papers in Economics and Finance, 2023/10, European Commission, Ispra, Italy, 2023, JRC135372. 3 Executive summary Natural disasters have always generated considerable losses for financial institutions, the public sector and citizens. Their impact has intensified over the last decade. Climate-related physical risks are a serious concern for both public and private finances, and it is of crucial importance to contain economic losses when physical risk materialize. Due to global warming, weather-related risks such as floods, wildfires, and droughts are projected to increase in frequency, severity and duration. According to recent figures provided by the European Commission (2021a), annual climate-related losses could amount to an estimated EUR 170 billion (1.36% of GDP) under the 3°C global warming scenario in the absence of mitigation and adaptation strategies. River floods are among the climate-related hazards most likely to intensify due to the long-term increase in temperature. Climate-related phenomena could affect GDP levels, and by extension public finances (affecting both public expenditure and revenues), and ultimately the lives of millions of people. It is therefore essential to model the impact of climate-related hazards on the economy and to estimate their overall cost. While insurance policies can help firms and households mitigate risks and withstand the economic consequences of natural disasters, public measures that enforce mitigation and adaptation policies targeted at areas and businesses affected by natural hazards are also necessary.1 This paper contributes to the debate by assessing how the landscape would change if the insurance sector covered exposures in each Member State, thereby reducing the need for public measures. It explores whether this would help curb the rise in potential government spending caused by more frequent extreme events. The methodology presented in this paper enables us to estimate: (i) the increase in insurance premiums necessary to harmonise the minimum level of protection against floods across all European countries; and (ii) the possible reduction in the amount of contingent losses for public finances,2 even if, under extreme and unlikely circumstances, some insurance contracts may not be honoured because certain insurance companies default in the aftermath of the disaster. While increasing insurance coverage is likely to be beneficial for private and public actors as it could reduce the overall cost for taxpayers, using insurance as a risk transfer mechanism could raise insurability and affordability concerns in a climate-damaged world. Increasing the insurance coverage of natural hazards could imply unrealistically high premiums that would make the product inaccessible to some policyholders. Therefore, as widely discussed in the literature, a portion of extreme risks could become uninsurable as they are not affordable for policyholders. Public and public-private insurance schemes, riskmitigation activities (i.e. preventive measures), public investment in risk reduction and prevention measures as well as targeted investments in loss prevention, could therefore be necessary. Policy context As climate change concerns society as a whole, the European Commission has announced measures to reduce the climate protection gap. The 2021 adaptation strategy aims to improve the understanding of natural disaster insurance penetration in Member States and to promote it. An additional proposal is to roll-out adaptation solutions to reduce the exposure of insurers, and more broadly of society, to climate-related risks, and to increase investment into better climate change adaptation measures. The Commission is also committed to strengthening dialogue between insurers, policymakers, and other stakeholders,3 identifying and promoting best practices in risk management funding, and exploring innovative solutions to deal with climate-induced risks, such as parametric insurances, mandatory insurance or bundling across risks, and risk transfer solutions (European Commission, 2021b). The Commission is working together with the European Central Bank and members of the European Systemic Risk Board to analyse and manage climate-related risks at EU level, and therefore contributing to the development of analytic frameworks for climate risk assessment. Main findings and key conclusions Our findings suggest that the expected losses stemming from floods in one year’s time could exceed EUR 33 billion, with only a part of them covered by insurance, and with very large variation as a share of GDP across 1Also public-private insurance schemes (PPPs) that pool risks and allow diversification are another tool. See ECB-EIOPA (2023) 2Contingent losses are amounts that, while not directly impacting public finances, might end up being covered by the public sector in an attempt to mitigate the impacts on citizens or the economy of extreme or systemic events. For the case of natural disasters, see e.g. Gamper at al. (2017) 3In this regard, see also the Climate Resilience Dialogue, as announced in the Strategy for financing the transition to a sustainable economy 4 Member States. The analysis shows that an increase of EUR 10.8 billion in written premium (+58%) is needed to level up the insurance climate penetration in the EU to at least 50%. The paper also estimates that uninsured floods, when considered together with potential insurance defaults, have the potential to generate EUR 27 billion in public finance contingent losses every year, and that reducing the climate protection gap by increasing insurance penetration could lower this impact by up to 50%. Related and future JRC work This work is related to several ongoing projects aimed at estimating the future economic consequences of climate change and climate-related physical risk, together with potential adaptation measures that would require the support of public and private funds. The outcome of selected works has been included in the annual joint ESRB/ECB reports on climate change and climate risk since 2019. JRC keeps contributing to the ESRB/ECB Climate Project Team, providing scientific analysis to support climate risk monitoring and future policies. 5 Flood protection gap: evidence for public finances and insurance premiums Bellia Mario, Di Girolamo Francesca Erica, Pagano Andrea, Petracco Giudici Marco* December 2023 Abstract Climate-related physical risks pose serious concerns for both public and private finances, and it is of utmost importance to contain economic losses when natural catastrophes occur. In this context, the paper models the potential economic impact of currently uninsured floods in the EU. It also assesses the potential reduction in economic losses by increasing the minimum level of flood insurance penetration, and the resulting increment in total premiums required to achieve this objective. First, the paper estimates the share of premiums associated with insured floods events over total premiums. Then, it investigates the extra premiums needed to close the flood protection gap by requiring all EU countries to reach a minimum level of insurance protection. Third, the paper proposes a stylised approach to quantify economic losses associated with uninsured flood events at different levels of insurance penetration, allowing to take into account that insurance protection could be partly ineffective due to defaults in the insurance sector. The model can be used to assess the size of the potential contingent loss for public finances if no preventive measures are taken to increase society’s resilience against climate and weather-related risks, and compare it with a safeguard mechanism under an “average” or ”worstcase” scenario. Results show that insurance premiums should be at least doubled to reach a harmonised level of penetration equal to 75%. Results show that average yearly uninsured losses could amount to EUR 27 billion today. Under an alternative scenario accounting for an increase in insurance penetration, losses would decrease by up to 50%. JEL Code: C15, G22, E6, Q54 Keywords: Physical risk, flood events, insurance, protection gap, insurance premiums *European Commission, Joint Research Centre (JRC) Ispra, Italy. We thank Paolo Paruolo (JRC), Tanguy De Launois (FISMA), Francesco Dottori (CIMA), Maximilian Ludwig (EIOPA), Stephan Fahr (ECB), Margherita Giuzio (ECB), members of the ECB/ESRB Climate Project Team, Silvia Fernandez Hernando (FISMA) and Markus Wintersteller (FISMA), for helpful comments and detailed discussions throughout the project. Disclaimer: The views expressed are purely those of the authors and may not in any circumstances be regarded as stating an official position of the European Commission. 6 1. Introduction and literature review Natural disasters have always generated considerable losses for financial institutions, the public sector and citizens. Their impact has intensified over the last decade. According to the European Environmental Agency (2022), weather and climate-related events caused EUR 450-520 billion in economic damage between 1980 and 2020 in the 32 European Economic Area countries. In addition, the number of reported natural disasters has almost doubled since 2010, as have the economic costs (Canova, F. and Pappa E., 2021). At global level, CRED and UNISDR (2018) calculate that climate-related disasters amounted to USD 2.45 billion between 1998 and 2017, rising by 251% during this 20-year period. Due to global warming, weather-related risks such as floods, wildfires, and droughts are projected to increase in frequency, severity and duration. According to recent figures provided by the European Commission (European Commission, 2021a), annual climate-related losses could amount to an estimated EUR 170 billion (1.36% of GDP) 'under the 3°C global warming scenario, in the absence of adaptation strategies. River floods are among the climate-related hazards most likely to intensify due to the long-term increase in temperature. For example, the recent catastrophic flooding of July 2021 in Belgium and Germany, which caused devastating damage to households and businesses, generated losses of up to EUR 32 billion (Mohr, 2022). The last three decades were among the most flood-rich periods in Europe of the last 500 years (Blöschl et al. 2020), and according to Feyen et al. (2020) river flood-related losses will reach EUR 50 billion per year (6 times more than now) under a 3°C increase in temperature in 2100 scenario. This will expose half a million people (three times more than today) to river flooding each year, and 2.2 million people to coastal inundation, and will generate losses of up to EUR 250 billion. Data analysists worldwide are gathering information on the level of losses associated with natural disasters. The JRC Risk Data Hub (JRC RDH), a platform developed by the European Commission, plays a central role in collecting information on the economic damages and human losses across Europe from hazard events. With the potentially disastrous effects of climate change in Europe, it is essential to consider the impact on the economy and potential impacts on public finances. Climate-related phenomena could affect GDP levels, and by extension public finances (affecting both public expenditure and revenues), and ultimately the lives of millions of people. It is therefore essential to model the impact of climate-related hazards on the economy and to estimate their overall cost. In 2007, Hallegatte et al. (2007) suggested that changes in the distribution of extremes could result in significant GDP losses in the absence of specific adaptation plans. Therefore, more accurate estimates of economic damages from climate-related events should consider the distribution of extremes, instead of their average cost, and make explicit assumptions on the organisation of future economies. On the same line of research, Prettenthaler et al. (2017) and P. Jindrová et al. (2019) use advanced extreme value theory and fit heavy-tailed distributions to quantify the size of flood-related losses in Europe. The World Bank and the European Commission have proposed a risk management model for fluvial and surface water floods coupled with a macro-fiscal analysis (see Solon-Swan economic growth model) to evaluate the impact of damages to assets - caused by disasters – on GDP and government spending. Recently, Gagliardi (2022) presented a stylised stress test to evaluate the fiscal impact of extreme weather and climate events. The authors quantify the deviation from the Commission’s 10-year baseline debt-to-GDP projections should a past extreme event reoccur in the medium term. The paper notes that such an event may pose risks to fiscal (debt) sustainability in some countries, namely Spain and Czechia, but remain manageable under standard global warming scenarios. Results point to a debt increasing effect of up to 5 percentage points of GDP. A complementary strand of research focuses on the financial risks associated with weather-related events. Mandel et al. (2021) quantify the risks of floods by modelling the propagation of climate-related shocks through financial networks. They show that both a country’s exposure to climate-related natural hazards and its financial leverage have an impact on the magnitude of global risks. Morana and Sbrana (2019) show that the increase in climate-related risks has a direct impact on the catastrophe bonds market, resulting in a decrease in the returns. Lending institutions may also be severely affected by climate-related catastrophic events, as they could create adverse economic conditions leading to an increase in insolvency rates in specific economic sectors. This could increase the number of corporate defaults, and consequently the risk to lenders. The European Central Bank (2022) published a preliminary climate stress test to assess the exposure of the banking sector to the impact of losses due to drought, heatwaves, and flood risks, by making use of data on the geographic location 7 of their lending activities. Results show that the combined credit and market risks losses for a sample of 41 European institutions would amount to EUR 70 billion, under a three-year disorderly transition scenario4. The increasing frequency and/or severity of extreme events may also affect the affordability and availability of insurance in the future. According to the European Insurance and Occupational Pensions Authority (EIOPA 2022a), property insurance is the business line that was the most affected by these climate-related risks. This latest publication discusses the potential impact of both extreme weather events and gradual global warming by assessing the potential negative consequences on the insurance sector. Tesselaar et al. (2020) study EU river flood insurance systems’ vulnerability to climate change. They apply a dynamic integrated flood insurance model and conclude that the rise in premiums causes problems of affordability, leading to a decline in the demand for flood insurance products. This, in turn, increases the financial vulnerability of households to flooding. The authors claim that government reinsurance for flood risk can be a suitable solution. Finally, the Commission is working together with the European Central Bank and members of the European Systemic Risk Board to analyse and manage climate-related risks at EU level, and therefore contributing to the development of analytic frameworks for climate risk assessment5. 2. Scope of the paper Climate-related events affect multiple stakeholders, from firms and households to the insurance and financial sectors, and could eventually impact public finances in case the state decides to intervene to cover losses following extreme events. The insurance sector can play a central role in managing the overall costs of climaterelated disasters by reducing costs that could potentially impact public finances, and by incentivising the development of good practices to reduce vulnerability through adaptation and mitigation measures. In other words, the insurance sector has a role to play in closing the protection gap by providing new insurance solutions, enhancing risk awareness6, developing new risk transfer solutions and creating the right incentives. Along this line of research, Holzheu and Turner (2018) address the discussion on the protection gap for extreme events and set out a framework to quantify the protection, by geography and risk type, in historical and expected terms. Following an empirical analysis of the key drivers of the protection gap, the authors propose several measures to narrow it. EIOPA has developed a pilot European dashboard that illustrates the insurance climate protection gap at Member State-level (EIOPA 2022b, NGFS 2019) for natural catastrophes and selected climate risks. The data show that protection gaps vary significantly between Member States, as well as between different perils. While the lowest protection gap is observed for windstorms, flood is the peril with the highest number of countries showing a high protection gap, specifically the Netherlands, Germany and Croatia. There are notable differences between insurance products in terms of accessibility, coverage, risk pricing and options across the EU, as well as differences in the share of disposable income to afford insurance premiums (Tesselaar, 2020). If insurance uptake does not increase to a minimum level of potential damage in every Member States, there may be withdrawals from the EU’s main solidarity instrument, the Solidarity Fund, which was designed to respond to ‘exceptional’ and ‘uninsurable’ disasters. Commission staff working document (2021c) outlines the current state of knowledge in that respect. The European Commission has therefore announced measures to reduce the climate protection gap. The 2021 adaptation strategy (European Commission, 2021a)7 aims to improve the understanding of natural disaster insurance penetration in Member States and promote it. An additional proposal is to roll-out adaptation solutions to reduce the exposure of insurers, and more broadly of society, to climate-related risks, and to increase investment into better climate change adaptation measures. The Commission is also committed to strengthening dialogue between insurers, policymakers and other stakeholders, identifying and promoting best 4 A disorderly transition scenario assumes delays in the implementation of climate policies to limit warming. 5 The latest contribution, ‘The macroprudential challenge of climate change’ includes several analysis and data provided by the JRC. 6 See for instance the report of SwissRe (2021), where ‘no action is not an option’ available here. 7 https://ec.europa.eu/clima/eu-action/adaptation-climate-change/eu-adaptation-strategy_fr 8 practices in risk management funding, and exploring innovative solutions to deal with climate-induced risks, such as parametric insurances, mandatory insurance or bundling across risks, and risk transfer solutions (European Commission, 2021b). This working paper contributes to the debate by assessing how the landscape would change if the insurance sector covered exposures in all EU Member States, thereby reducing the need for public measures. This course of action could minimise the rise in public costs and provide the financial capacity to rebuild infrastructures in the aftermath of extreme natural, thereby keeping the economy stable. The paper focuses on coastal and river floods specifically. The methodology presented in this paper enables us to quantify: (i) the increase in insurance premiums necessary to harmonise the level of protection against flood events across all European countries; and (ii) the possible reduction in the amount of public finance losses even under a ‘worst-case’ scenario where the insurance mechanism is only partially effective due to defaults in the insurance sector. On the former, we need to estimate the share of premiums for fire and other damages to property insurance pertaining to floods. The flood-related expected losses are assumed to be the insured share of climate-related losses associated with floods (coastal and river), which are calculated using the Risk Data Hub’s figures on the number of people exposed. Starting from this assumption, we assess by how much written premiums would increase if the insurance sector were called to reduce the climate protection gap in silos (i.e. without considering adaptation or mitigation measures or the involvement of other actors in reducing economic losses). The second part of the analysis uses a stylised model to assess the maximum loss for public finances under a worst-case scenario where some insurance companies default in the aftermath of flood events, rendering the insurance protection only partly effective. The framework does not address the issue of changes in risk unit prices, and assumes that neither preventive measures to increase the resilience of society against climate and weather-related risks, nor increases in the frequency or severity of extreme events due to climate change are considered.8 Preliminary findings suggest that the expected losses stemming from floods in one year could exceed EUR 33 billion. Only a quarter of climate-related losses are covered by insurance, and there is significant variation across Member States9. Premiums for floods account for 12.5% of premiums for fire and other damages to properties and businesses, and an increase of EUR 10.8 billion (+58%) is needed to level up the insurance climate penetration in the EU to at least 50%. Finally, the paper shows that floods, together with possible insurance company defaults, have the potential to generate EUR 27 billion in public finance losses every year. Reducing the climate protection gap could lower the impact by 50%, even when considering the possibility of insurance defaults. While increasing insurance coverage is likely to be beneficial for private and public actors as it could reduce the overall cost for taxpayers, using insurance as a risk transfer mechanism could raise insurability and affordability concerns in a climatechanged world. Increasing the insurance coverage of natural hazards could lead to an increase in premiums for certain risks, which could make the product inaccessible to some policyholders. On this, the literature consensus is that a portion of extreme risks is not insurable as it may not be financially sustainable for policyholders. The results seem to support the need to develop and roll-out adaptation measures to increase climate resilience, as envisaged by the climate adaptation strategy. Risk-mitigation activities (i.e., preventive measures), public investment in risk reduction and prevention measures, as well as targeted investments in loss prevention are necessary. Once future disaster-related spending decreases, the insurance market will be able to provide additional coverage against these disasters. The paper is structured as follows. Section 3 describes the database and the methodology used to estimate the impact of insurance premiums. Section 4 presents methodology and results to quantify the impact on public finances. Finally, Section 5 presents the conclusions. 8 Based on the latest projections from climate change modelling it would also be possible to include future impacts. 9 https://www.eiopa.europa.eu/what-do-about-europes-climate-insurance-gap-2023-04-24_en 15 3.3. Results According to EIOPA (2022a), non-life and all property-related premiums are likely to increase in the absence of mitigation and adaptation measures, given the risk-based calculation of the insurance premiums. Nevertheless, the European Commission’s strong commitment to strengthening the EU’s resilience to climate change, through initiatives such as the European Green Deal, the strategy on adaptation to climate change and the Climate Resilience Dialogue should mitigate the potential damages of climate risks and extreme weather events. In our empirical analysis, we therefore assume that the actual premiums reflect the short-term riskiness typical of non-life insurance, as opposed to long-term life insurance. To quantify the increase in insurance premiums required to achieve a minimum level of penetration in all Member States against flood events, we apply the previously presented methodology. For each Member State, Table 1 presents the estimations of the pure premiums (𝐸𝐸𝑃𝑃𝑃𝑃𝑖𝑖) and gross written premium (𝐸𝐸𝐺𝐺𝑃𝑃𝑖𝑖). Table 1 also reports the actual insurance penetration for floods IPflood(i) for each MS and the value of the OIRFi.22 The total amounts of pure premiums needed to increase the penetration rate for floods are roughly EUR 20 billion and EUR 26 billion for a minimum penetration of 50% and 75%, respectively. Our estimation of the pure premium at around EUR 12.6 billion (see Table 1) is relatively close to the actual amount of premiums for flood events according to EIOPA (2022a), which is around EUR 10 billion23. However, when including insurance margins, the two amounts diverges substantially; the total 𝐸𝐸𝐺𝐺𝑃𝑃 yields EUR 18.6 billion of premiums using our proxies. The final estimated gross written premiums needed to reach a minimum harmonised penetration of 50% (75%) are around EUR 30 (38) billion. It is worth noting that the additional amounts that comes from the shock via the OIRF, based on the observed relationship in the data, are rather small (the average is about 23%). One potential explanation is that we impose fixed long-run coefficients, constraining the variables to comove in the long-run and diverging only in the short-run. Another explanation is that the modelling framework is quite suitable for small, marginal increases in the insurance coverage (which are the ones that we observe in the data) but it might not be adequate for large increases in the insurance penetration. According to our model, total premiums written for flood events should therefore be increased by more than 58% to reach a minimum 50% penetration across the EU. The additional premiums amount to EUR 10.8 billion to reach a minimum of 50% penetration, and EUR 19.7 billion to achieve a minimum 75% penetration rate. These estimates are clearly a lower bound, given that our framework does not consider several factors that could substantially increase premiums. However, the increases vary widely between Member States, depending on their exposure to flood events and, more importantly, their starting insurance penetration rates. The Netherlands alone accounts for more than half of the additional premiums written required for a 75% penetration rate, due to a high risk of flooding, low insurance penetration, and lack of coverage for potential losses from floods (see the earlier discussion regarding the Netherlands). Other Member States requiring a substantial increase in premiums written include Germany and Italy. Under these conditions, insurance companies are likely to be willing to cover risks not currently insured only at higher premiums. Therefore, this estimate is clearly on the conservative side. Other economic factors not related to the willingness to buy additional coverage could explain why insurance coverage is lower in certain countries and regions. From the insurers’ perspective, increasing the riskiness of the portfolio demands capital, which is costly. Furthermore, as risk increases, so do the prices of reinsurance, potentially deterring both investors and potential new policyholders. 22 The estimated plots of the OIRF for each Member State are reported in Appendix 4. 23 According to EIOPA(2022a) the overall gross written premium for extreme climate events amount to EUR 19.3 billion in the EIOPA sample. The same source reports that the exposure to flood risk represents around 27% of the total exposures to climate. Using these values and considering that the EIOPA sample covers around 51.76% of the total non-life gross written premium, a rough estimation points to an actual total amount of premiums for flood events of around EUR 10.06 billion. This estimation is only a proxy since generally for the non-life insurances, multiple risks are bundled together, and the coverage for natural catastrophes is part of the fire or property insurance. 16 Table 1: Estimation of the additional expected premiums. Panel A Panel B Member State 𝑰𝑰𝑷𝑷𝒇𝒇𝒇𝒇𝒇𝒇𝒇𝒇𝒇𝒇 (𝒊𝒊) (𝟏𝟏+𝑶𝑶𝑰𝑰𝑶𝑶𝑶𝑶𝒊𝒊) 𝑬𝑬𝑷𝑷𝑷𝑷𝒊𝒊 (EUR Mn) 𝑬𝑬𝑬𝑬𝑷𝑷𝒊𝒊 (EUR Mn)) 𝑬𝑬𝑷𝑷𝑷𝑷𝒊𝒊𝟓𝟓𝟓𝟓 (EUR Mn) 𝑬𝑬𝑷𝑷𝑷𝑷𝒊𝒊𝟕𝟕𝟓𝟓 (EUR Mn) 𝑬𝑬𝑬𝑬𝑷𝑷𝒊𝒊𝟓𝟓𝟓𝟓 (EUR Mn) 𝑬𝑬𝑬𝑬𝑷𝑷𝒊𝒊𝟕𝟕𝟓𝟓 (EUR Mn) AT 70% 1.006 575.08 849.81 578.64 618.14 849.81 913.43 BE 78% 1.014 827.15 1 222.29 839.04 839.04 1 222.29 1 222.29 BG 25% 1.036 27.43 40.53 56.83 85.24 83.98 125.97 CY 28% 1.003 0.31 0.46 0.55 0.83 0.82 1.22 CZ 91% 1.033 401.12 592.74 414.45 414.45 592.74 592.74 DE 43% 1.003 2 724.51 4 026.07 3 177.37 4 766.06 4 695.27 7 042.91 DK 75% 1.000 171.62 253.60 171.64 171.64 253.60 253.60 EE 46% 1.026 6.13 9.06 6.81 10.21 10.06 15.09 EL 8% 1.002 10.16 15.01 62.19 93.28 91.89 137.84 ES 75% 1.011 736.34 1 088.10 744.15 744.15 1 088.10 1 088.10 FI 90% 1.047 603.93 892.44 632.37 632.37 892.44 892.44 FR 75% 1.008 3 666.91 5 418.68 3 694.64 3 694.64 5 418.68 5 418.68 HR 21% 1.021 29.18 43.12 70.39 105.58 104.01 156.02 HU 75% 1.021 387.19 572.16 395.50 395.50 572.16 572.16 IE 75% 1.007 57.56 85.06 57.94 57.94 85.06 85.06 IT 23% 1.002 874.12 1 291.71 1 899.78 2 849.66 2 807.34 4 211.01 LT 33% 1.027 13.28 19.63 20.52 30.78 30.33 45.49 LU 75% 1.079 29.28 43.27 31.60 31.60 43.27 43.27 LV 40% 1.025 33.54 49.57 42.70 64.05 63.10 94.65 MT 25% 1.023 0.00 0.00 0.00 0.01 0.01 0.01 NL 3% 1.039 295.88 437.23 5 810.21 8 715.32 8 585.88 12 878.82 PL 60% 1.008 604.68 893.55 609.30 761.63 893.55 1 125.47 PT 26% 1.011 11.91 17.60 23.51 35.26 34.74 52.11 RO 26% 1.030 137.40 203.04 269.84 404.77 398.75 598.13 SE 96% 1.001 235.14 347.46 235.46 235.46 347.46 347.46 SI 85% 1.019 75.45 111.50 76.91 76.91 111.50 111.50 SK 25% 1.029 63.81 94.30 131.30 196.95 194.03 291.04 Total Premiums 12 599.12 18 618.00 19 943.45 25 929.45 29 470.87 38 316.52 Tot al Premiums (excluding NL) 12 303.24 18 180.77 14 133.23 17 214.13 20 884.99 25 437.70 Additional Premiums 7 344.33 13 330.33 10 852.88 19 698.52 Note: premiums in grey-italics refers to MS that already reach the penetration level of 50% or 75%. EPP stands for Expected Pure Premiums (without margins), while EGP stands for Expected Gross Premiums (with margins). Source: JRC elaboration using data from EIOPA insurance statistics. Figure 5 illustrates the distribution of the potential increase in gross premium needed to harmonise the insurance penetration up to 50% or 75% at EU level. Considering a harmonised level of 50% for insurance penetration for floods, on average, Member States that do not reach this threshold (see also Figure 3) should increase their penetration by 76.15% (with a standard deviation of 41%)24. Instead, considering a harmonised level of 75%, the average increase is 143% (with a standard deviation of 78%). 24 Relative change of the penetration rate. 17 Figure 5: Percentage increase in gross written premium Note: the boxplot represents the distribution, across Member States, of the percentage increase of gross written premium in order to harmonise the insurance penetration for river and coastal floods up to 50% (blue boxplot) and 75% (red boxplot). The Netherlands and Greece (outliers) and Member states that already reach the 50% (75%) penetration are excluded from the plot. Source: EIOPA insurance statistics, JRC elaboration. 4. Economic losses to public finances The aim of the framework outlined in this section is to estimate the changes in impact on public finances when flood-related damages occur under different insurance penetration rates, while considering that insurance might be partly ineffective due to insurance sector defaults. To address this point, we use a stylised stress test model to assess the maximum loss to which public finances are exposed. If an insurance company defaults due to unexpected events that exceed its repayment capacity, it cannot provide appropriate coverage to its policyholders, and not all claims can necessarily be covered. In such cases, public finances may be subject to financial losses. Our stylised framework quantifies the maximum loss to public finance in a worst-case scenario where flood-related losses, eventually under an increased penetration rate, are accompanied by insurance defaults. 25 4.1. Methodology and data We assume that the insurance sector can be regarded as a portfolio of counterparty risks. Within the portfolio, each insurer has a small, but non-zero probability of causing a liability to policyholders upon default. Upon default of an insurance undertaking, the exposure at default (𝐸𝐸𝐸𝐸𝐺𝐺𝑐𝑐) is the maximum amount of the company’s liabilities to claimants, beneficiaries and insured. The loss given default (LGD) is the percentage loss that will effectively be incurred on the exposure once the defaulted company's recovery rate is considered. With the oneyear probability of default of the company given by 𝑃𝑃𝐺𝐺𝑐𝑐, the expected liability (𝐸𝐸𝐸𝐸𝑐𝑐) for a single company ‘c’ over the period of 1 year, is given by: 𝐸𝐸𝐸𝐸𝐶𝐶= 𝐸𝐸𝐺𝐺𝐺𝐺×𝐸𝐸𝐸𝐸𝐺𝐺𝐶𝐶×𝑃𝑃𝐺𝐺𝐶𝐶. (10) Since we are not interested in a single insurance undertaking, but in all insurance companies at individual country level (or even at the aggregate EU-27 level), we can make some simplifying assumptions to estimate the loss distribution of the insurance sector in each country, without the need to estimate the loss distributions of individual insurance undertakings (see European Commission, 2010 and European Commission, 2021c for a fuller discussion). As different insurers may have different loss rates, information on the distribution of losses from insurance defaults is necessary to assess the effective risk the public is exposed to. The loss-rate distribution can be seen as the loss rate on a portfolio of exposures to several insurance undertakings. Specifically, we use the Vasicek model (Vasicek 2002) to define the event of default as occurring when the 25 In this analysis we are not considering explicitly the role of reinsures, though the use of reinsurance will affect reserves and provisions. Also, when considering the potential impact of insurance default we are considering the whole portfolio, and not just flood risks, and we do not take into account the mitigating impact of Insurance Guarantee Schemes. The impact of Solvency II regulation in minimizing the default rate of insurers is implicitly taken into account in the choice of maximum probability of default. 0 .5 1 1.5 2 2.5 75% 50% Increase in gross written premium River and Coastal flood 18 insurer’s asset value falls below a predetermined threshold. The value of 𝐸𝐸𝑖𝑖 for country 𝑝𝑝 represents the maximum loss that should not be expected to exceed in 1 year with a probability level α is given by:26 𝐸𝐸𝑖𝑖=𝐸𝐸𝐸𝐸𝐺𝐺𝑖𝑖×𝐸𝐸𝐺𝐺𝐺𝐺×𝑁𝑁��𝜌𝜌+𝛿𝛿(1−𝜌𝜌) 𝑁𝑁−1(1−𝛼𝛼)+𝑁𝑁−1(𝑃𝑃𝐷𝐷) �1− 𝜌𝜌−𝛿𝛿(1−𝜌𝜌)�. (11) Some notes on the parameters used in our analysis: • The LGD is set equal to 15% as in European Commission (2021c). • PD is fixed at 0.5% for simplicity, this value being the maximum probability of default, which should be attained in the Solvency II framework and therefore marks an upper bound for the probability distribution of defaults. • 𝜌𝜌 is the correlation among defaults and has been set at 20%, consistent with European Commission (2021c). • 𝛿𝛿 is the concentration exposure term, tackling the fact that a portfolio of insurers consists of a discrete number of relatively large exposures. This correction term is calculated on the basis of the companies’ market share, as a proxy for the relative size of individual exposures in the portfolio27, by summing up the squares of the relative sizes of the markets shares. We estimate 𝛿𝛿 separately for each country based on information from EIOPA on the market share of the top 1, top 3, top 5, top 10, and top 15 insurance undertakings. We refer to European Commission (2021c) for more details and we report the estimated values of δ per country for the total insurance sector in Figure 10 of Appendix 3. • EADi is the total exposure of the portfolio. In our case it is estimated as the sum of TPi, our best estimate of liabilities and risk margin, and SRCi the total amount of funds that an insurer is required to hold to ensure that the company will be able to meet its obligations with a probability of at least 99.5%(Table 2).28 We assume that EADi increases together with the increase in insurance penetration. We calculate the additional exposures at default to be equivalent to the extra losses that would be covered by the insurance sector. Table 2: Exposure at default, EUR million (as of 2021) Member State Exposure at default AT 82 219 BE 268 167 BG 3 680 HR 4 226 CY 2 524 CZ 12 519 DK 257 384 EE 1 668 FI 69 404 FR 1 300 185 DE 767 461 EL 15 599 HU 5 867 IE 124 731 IT 599 635 LV 1 136 LT 1 114 LU 56 426 MT 4 911 26 It is one of the most widely applied tools for quantitative financial risk management and it is mostly used to assess default portfolio risk across a variety of business sectors, including the insurance sector. The framework of Vasicek (2002) hinges on the asymptotic behaviour of an extended Merton model (Merton, 1974) when the number of exposures in the portfolio of insurers goes to infinity. This model was also initially proposed for counterparty default risk module in QIS3 and QIS4. 27 The calculation methodology is the same as that of the calculation of the Herfindahl–Hirschman Concentration Index (HHI), used widely in competition literature. 28 Liabilities at the time of default for individual insurers can be much larger and could deviate substantially from the sum of TP and SCR. In addition, there might be additional capital buffers on top of the current minimum capital requirements. Thus, the estimation of the EAD provides a conservative lower bound for the exposures and the subsequent calculation of losses. 19 Member State Exposure at default NL 83 161 PL 16 383 PT 19 988 RO 2 724 SK 4 041 SI 6 261 ES 208 083 SE 211 736 Source: EIOPA insurance statistics and JRC elaboration We apply this modelling framework under two scenarios. In the baseline scenario, we consider the expected economic loss (𝐸𝐸𝐸𝐸𝐸𝐸𝑖𝑖) estimated using Risk Data Hub values and the one-year expected liability from insurer’s defaults at country level 𝐸𝐸𝐸𝐸𝑖𝑖. Specifically, we compare the situation with the actual insurance penetration rate to a situation with a harmonised 75% insurance penetration rate for flood events across all Member States. The baseline expected losses (𝐵𝐵𝐸𝐸𝑖𝑖) are therefore calculated as follows: 𝐵𝐵𝐸𝐸𝑖𝑖=𝐸𝐸𝐸𝐸𝑖𝑖+�1−IPflood,i�×𝐸𝐸𝐸𝐸𝐸𝐸𝑖𝑖 (12) where IPflood,i represent the actual penetration rate, which will be increased up to 75% for Member States that do not reach this threshold. This amount represent the potential expected losses to public finances in 1 year for flood-related events. In a second, more severe, worst-case scenario, we look at what would happen in the case of a compound event. We do so by considering uninsured flood-related losses together with losses stemming from defaults in the insurance sector in a tail scenario. Under this scenario, we consider a set of very rare events that occur once every 200 years (i.e. with a probability of 0.5%) and therefore, we evaluate the losses 𝐸𝐸𝑖𝑖 with a confidence level 𝛼𝛼= 0.5. Similarly, we consider only losses from flood events with a return period of 200 years (𝐸𝐸𝐸𝐸𝐸𝐸𝑖𝑖,200). Specifically, we begin by calculating the share of the population affected by floods events with a return period of 200 years (𝐸𝐸𝐸𝐸𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑇𝑇𝑇𝑇𝑇𝑇𝑖𝑖𝑝𝑝𝑝𝑝,200) as a proportion of the total population. We than apply the 2020 GDP at current market prices and the vulnerability index for each country to the formula, as follows: 𝐸𝐸𝐸𝐸𝐸𝐸𝑖𝑖,200= 𝐺𝐺𝐺𝐺𝑃𝑃𝑖𝑖 × 𝐸𝐸𝐸𝐸𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑖𝑖𝑝𝑝𝑝𝑝,200 𝑇𝑇𝑝𝑝𝑇𝑇𝑇𝑇𝑇𝑇 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑇𝑇𝑇𝑇𝑇𝑇𝑖𝑖𝑝𝑝𝑝𝑝𝑖𝑖×𝑉𝑉𝑖𝑖. (13) Finally, worst-case scenario losses on public finances 𝑊𝑊𝑊𝑊𝐸𝐸𝑖𝑖 are calculated as the sum of uninsured flood losses and leftover losses from insurance sector defaults: 𝑊𝑊𝑊𝑊𝐸𝐸𝑖𝑖=𝐸𝐸𝑖𝑖+�1−IPflood,i�×𝐸𝐸𝐸𝐸𝐸𝐸𝑖𝑖(200𝑦𝑦) (14) where 𝐸𝐸𝑖𝑖 represents the maximum loss that should not be expected to be exceeded in 1 year with a probability level αf 0.5%, and 𝐸𝐸𝐸𝐸𝐸𝐸𝑖𝑖(200𝑦𝑦)represents the expected economic losses from a flood event with a return period of 200 years. In line with the baseline, we compare the situation with the actual insurance penetration rate to a situation where there is a harmonised 75% insurance penetration rate for flood events across all Member States29. 4.2. Results The results of the baseline scenario are of a similar order of magnitude as the previous analysis. Notably, when considering the baseline expected losses (𝐵𝐵𝐸𝐸𝑖𝑖) with the current protection rate, expected losses are estimated to be around EUR 27 billion. This amount represents the average losses that could occur in 1 year that would need to be covered by the private or the public sector, owing to the potential defaults of insurance companies and uninsured flood-related losses. When considering a harmonised minimum level of protection of 75% across Member States, this amount drops substantially. Since losses due to the default of insurance companies represent only a small fraction of the total, the final overall reduction is directly due to an increase in insurance protection. At EU level, the reduction in expected losses, increasing the penetration rate to 75%, comes to around EUR 14 billion, an amount smaller with respect to the increase in gross written premium of EUR 19.6 billion presented before. When excluding the Netherlands, the baseline figure stands at EUR 15.7 billion. This drops to 29 Note that we do not take into consideration the correlation between the different events, assuming instead that a very large flood will correspond to an extreme fragility situation in the insurance sector. The actual probability of the compound event could therefore be lower than 0.5% 20 EUR 10.6 billion when penetration increases to 75%. Figure 6 Panel A (left boxplot) shows that the reduction in public losses when increasing the penetration level to 75% is substantial, amounting to 40-60% for most Member States. However, some caveats need to be considered when reading the results. Firstly, flood impact estimates are affected by a higher degree of uncertainty, since they cannot be comprehensive and they cannot cover all potential consequences of global warming. Secondly, in this scenario we do not consider adaptation and mitigation measures or the effect of actual and future flood defences. Furthermore, it is challenging to model adaptation, as this course of action also requires strong commitment by public and private stakeholders, which is difficult to include in the analysis. Finally, we are not considering the role of re-insurance. Nevertheless, the estimations in our analysis provide an additional assessment that could help inform the policy debate on reducing the insurance protection gap or increasing adaptation measures. Our model also reflects a worst-case scenario where a very extreme weather-related event take place (a catastrophic event that occurs once every 200 years) everywhere in the EU, and insurers’ defaults make them unable to fulfil their contractual commitments. In this very extreme (and unlikely) event, aggregated public finance losses can be relevant and impactful. The results show that total losses for the EU would amount to EUR 1 576 billion in 1 year. The confidence level for this projection is 99.5%. When excluding the Netherlands, the projected losses amount to EUR 1 194 billion. Increasing the insurance penetration to 75% would reduce losses by around 50-70% for most Member States (Figure 6 Panel A, right boxplot). This scenario results in a 10% decrease in GDP on average, with considerable differences between countries depending on exposure to river and coastal floods and the actual level of insurance protection (Figure 6 Panel B)30. Based on our model, harmonising insurance coverage to 75% across all EU countries could potentially cut public finance losses in half. The confidence level for this is 99.5%. Moreover, numbers suggest that harmonising the insurance penetration rate might reduce losses by up to 80% in countries with a low penetration rate. Figure 7 shows that the reduction will be 40-60% for six Member States and 60-80% for another six Member States. 30 This extreme scenario must be considered as very unlikely, as extreme floods usually affect only a limited area. For example, the flood event in July 2021 had approximately a 1-in-200 year probability of occurrence, but affected only part of Belgium, Germany, and the Netherlands. Results are available under different confidence levels, upon request. 21 Buckets Figure 6: Reduction in public finance losses in the baseline and worst-case scenario, where the insurance penetration rate is harmonised at 75% (Panel A – percentage reduction, note ‘inverted’ scale). Distribution of public finance losses in EU in the baseline and worst-case scenario, under the actual insurance penetration rate and under a harmonised level of insurance penetration of 75% (% GDP, Panel B, 𝜶𝜶 = 0.5%. Outliers are excluded from the plot). Panel A Panel B Source: EIOPA insurance statistics, EIOPA dashboard, Risk Data Hub, and JRC elaboration Figure 7: Frequency of the reduction of public finance losses under an insurance penetration level of 75% (Worst-case scenario, 𝜶𝜶 = 0.5%). Source: EIOPA insurance statistics, EIOPA dashboard, Risk Data Hub, and JRC elaboration 22 5. Conclusion Due to climate change, weather-related risks are projected to increase in frequency, severity and duration, and to affect financial stability. Natural disasters can be devastating, generating significant losses for financial institutions, the public sector and citizens alike. Against this background, the paper offers a stylised modelling approach to quantify the increase in insurance premiums necessary to harmonise the level of protection against floods across all European countries. It explores the scale of public finance losses in a worst-case scenario where floods and increases in the insurance premiums are accompanied by defaults in the insurance sector. Findings suggest that the expected losses stemming from floods could exceed EUR 33 billion (EUR 22.5 billion when excluding the Netherlands) in 1 year. Only a fraction of potential losses are covered by insurance, with significant variation between Member States. Our estimations show that an increase of EUR 10.8 billion (+58%) would be needed to harmonise the penetration rate in Europe to a minimum of 50%. Finally, the paper shows that floods, together with possible insurances defaults, have the potential to generate EUR 27 billion in annual public finance losses, and increasing insurance penetration for floods to up to 75% could lower the impact by up to 50%. In a worst-case scenario (a rare event that could occur once every 200 years), losses can be substantial. We show that for some Member States losses can be reduced by 80% when insurance penetration is harmonised at 75%. Given the high uncertainty of flood impact estimates, the results of our models are highly sensitive to the initial loss data and the underlying assumptions, including no mitigation effects, and do not consider the effect of actual and future flood defences. In addition, we are not explicitly modelling the effect of re-insurance. Although increasing insurance coverage would seemingly be beneficial for both private and public actors, even in a worst-case scenario, reducing overall costs for taxpayers, using insurance as a risk transfer mechanism could raise insurability and affordability concerns in a climate-damaged world. Moreover, increasing the insurance coverage of natural hazards could result in unrealistically high premiums that would be unaffordable for policyholders. The paper therefore supports the consensus that a portion of extreme risks is not insurable as it may not be financially bearable for policyholders. Risk-mitigation activities (i.e. preventive measures), public investment in risk reduction and prevention measures as well as targeted investments in loss prevention, are necessary. Once future disaster-related expenditures are reduced, the insurance market will be able to provide additional coverage against these disasters. Future research could therefore explore the issues of increased risk and unit risk prices as the penetration rate increases. The scientific evidence resulting from this research would presumably demonstrate the need to develop and roll out adaptation measures to increase climate resilience, as envisaged in the climate adaptation strategy. 23 References Antofie, T., Luoni, S., Eklund, L. G., & Montserrat, M. F. (2020). Update of Risk Data Hub software and data architecture. Ispra, Italy: JRC technical report. Balbás, A., Garrido, J., & Mayoral, S. (2009). Properties of distortion risk measures. Methodology and Computing in Applied Probability, 11(3), 385-399. Blöschl, G., Kiss, A., Viglione, A., Barriendos, M., Böhm, O., Brázdil, R., Coeur, D., Demarée, G., Llasat, M. C., Macdonald, N., Retsö, D., Roald, L., Schmocker-Fackel, P., Amorim, I., Bˇelínová, M., Benito, G., Bertolin, C., Camuffo, D., Cornel, D., Doktor, R., Elleder, L., Enzi, S., Garcia, J. C., Glaser, R., Hall, J., Haslinger, K., Hofstätter, M., Komma, J., Limanówka, D., Lun, D., Panin, A., Parajka, J., Petri´c, H., Rodrigo, F. S., Rohr, C., Schönbein, J., Schulte, L., Silva, L. P., Toonen, W. H. J., Valent, P., Waser, J., and Wetter, O. (2020). Current European flood-rich period exceptional compared with past 500 years. Nature, 583(7817):560–566. Canova, F. and Pappa E., (2021). Costly Disasters & the role of fiscal policy: evidence from US states. European Economy Fellowship Initiative discussion paper. DNB - De Nederlandsche Bank (2017) Waterproof? An exploration of climate-related risks for the Dutch financial sector. Engle, R. F., and C. W. J. Granger. 1987. Co-integration and error correction: Representation, estimation, and testing. Econometrica 55: 251–276. EIOPA (2022a). European insurers’ exposure to physical climate changes risk. Potential implications for non-life business. EIOPA-22/278, 20 May 2022. EIOPA (2022b). The dashboard on insurance protection gap for natural catastrophes in a nutshell. EIOPA-22/507, 05 December 2022. European Central Bank (2022). Climate risk stress test – Methodology, scenarios and quality assurance. ECB – EIOPA (2023). Policy options to reduce the climate insurance protection gap Discussion Paper European Commission (2010). Insurance Guarantee Schemes: derivation of loss distributions and funding needs. Ispra (Italy): Methodological report for internal use by the European Commission. European Commission (2021a), Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions. Forging a climate-resilient Europe - the new EU Strategy on Adaptation to Climate Change. European Commission (2021b). Closing the climate protection gap - Scoping policy and data gaps. Commission staff working document. European Commission (2021c). Insurance Guarantee Schemes: quantitative impact of different policy options. Joint Research Centre. JRC technical report European Environmental Agency (2022). Economic losses and fatalities from weatherand climate-related events in Europe. Briefing no. 21/2021. Feyen, L., Ciscar Martinez, J. C., Gosling, S., Ibarreta Ruiz, D., Soria Ramirez, A., Dosio, A., Naumann, G., Russo, S., Formetta, G., Forzieri, G., Girardello, M., Spinoni, J., Mentaschi, L., Bisselink, B., Bernhard, J., Gelati, E., Adamovic, M., Guenther, S., De Roo, A., Cammalleri, C., Dottori, F., Bianchi, A., Alfieri, L., Vousdoukas, M., Mongelli, I., Hinkel, J., Ward, P. J., Gomes Da Costa, H., De Rigo, D., Liberta’, G., Durrant, T., San-Miguel-Ayanz, J., Barredo Cano, J. I., Mauri, A., Caudullo, G., Ceccherini, G., Beck, P., Cescatti, A., Hristov, J., Toreti, A., Perez Dominguez, I., Dentener, F., Fellmann, T., Elleby, C., Ceglar, A., Fumagalli, D., Niemeyer, S., Cerrani, I., Panarello, L., Bratu, M., Després, J., Szewczyk, W., Matei, N., Mulholland, E., and OlariagaGuardiola, M. (2020). Climate change impacts and adaptation in Europe: JRC PESETA IV final report. Technical report, Publications Office, Luxembourg. Gamper, C., Signer B., Alton L. and Petrie M. (2017), Managing disaster-related contingent liabilities in public finance frameworks, OECD Working Papers on Public Governance, No. 27, OECD Publishing, Paris, https://doi.org/10.1787/a6e0265a-en. Gagliardi, N., Arévalo, P., and Pamies S. (2022). The Fiscal Impact of Extreme Weather and Climate Events: Evidence for EU Countries. European economy discussion papers, 168, July 2022. Hallegatte, S., Hourcade, J. C., Dumas P. (2007). Why economic dynamics matter in assessing climate change damages: illustration on extreme events. Ecological Economics, Elsevier, 62 (2), pp.330-340. 24 Jindrová, P., & Pacáková, V. (2019). Natural catastrophe models for insurance risk management. Wseas Transactions on Business and Economics, 16, 9-17. Johansen, S. (1995). Likelihood-based inference in cointegrated vector autoregressive models. OUP Oxford. Kuik, O., Scussolini, P., Mechler, R., Mochizuki, J., Hunt, A., & Wellman, J. (2016). Assessing the economic case for adaptation to extreme events at different scales. ECONADAPT Deliverable, 5. Mandel A., Tiggeloven T., Lincke, D.,· Koks E., Ward P., Hinkel J. (2022). Risks on global financial stability induced by climate change: the case of flood risks. Climatic Change. 166: 4. Merton, R. (1974). On the Pricing of Corporate Debt: the Risk Structure of Interest Rates. Journal of Finance, 29(2), 449-470. Mohr, S., Ehret, U., Kunz, M., Ludwig, P., Caldas-Alvarez, A., Daniell, J. E., Ehmele, F., Feldmann, H., Franca, M. J., Gattke, C., Hundhausen, M., Knippertz, P., Küpfer, K., Mühr, B., Pinto, J. G., Quinting, J., Schäfer, A. M., Scheibel, M., Seidel, F., and Wisotzky, C. 2022: A multi-disciplinary analysis of the exceptional flood event of July 2021 in central Europe. Part 1: Event description and analysis. Nat. Hazards Earth Syst. Sci. Discuss. Morana, C., and Sbrana, G. (2019). Climate change implications for the catastrophe bonds market: An empirical analysis. Economic modelling, 81, 274-294. NGFS (2019). A call for action Climate change as a source of financial risk. Prettenthaler, F., Albrecher, H., Asadi P., and Köberl, J. 2017. On flood risk pooling in Europe, Natural Hazards, Journal of the International Society for the Prevention and Mitigation of Natural Hazards, ISSN 0921-030X Vol 88 (1). Scussolini, P., Aerts, J. C., Jongman, B., Bouwer, L. M., Winsemius, H. C., de Moel, H., & Ward, P. J. (2016). FLOPROS: an evolving global database of flood protection standards. Natural Hazards and Earth System Sciences, 16(5), 1049-1061. Tesselaar M., Botzen W.J.W., Haer T., Hudson P., Tiggeloven T., Aerts J. C. J. H. (2020) Regional Inequalities in Flood Insurance Affordability and Uptake under Climate Change. Sustainability 2020, 12, 8734; doi:10.3390/su12208734. Tesselaar, M.; Botzen, W.J.W.; Aerts, J.C.J.H (2020). Impacts of Climate Change and Remote Natural Catastrophes on EU Flood Insurance Markets: An Analysis of Soft and Hard Reinsurance Markets for Flood Coverage. Atmosphere 2020, 11, 146. https://doi.org/10.3390/atmos11020146 Vasicek, O. (2002). The Distribution of Loan Portfolio Value. Risk (December). 31