Embedding Water‑related Risks in Financial Stability Frameworks

Annex B. Assessing economic and financial risks under different scenarios

Models assessing economic impacts of water-related risks

The NGFS currently designs climate scenarios by relying on different types of integrated assessment models (IAMs). Process-based IAMs, which are the main models supporting the climate scenarios of the NGFS, are models that describe the potential evolution of the global energy system, as well as other systems with important GHG emissions, including agriculture and land use changes. Such models allow for the study of both land use changes and the transition of the energy system, including the investment needed for such transformation of the energy matrix. They generate optimal trajectories according to the transition costs subject to a set of constraints imposed by the scenario narrative. They also allow for an estimate of both the global and regional marginal abatement costs and enable the study of the emissions trajectories under each NGFS scenario subject to the carbon budget restriction.

Econometric models

Econometric models quantify how water-related risks, including scarcity, quality degradation, and extreme weather events, impact key economic indicators such as GDP, employment rates, agricultural productivity, and industrial output. These models are versatile, capable of integrating diverse data sources such as historical records, remote sensing data, and climate model projections. While there are no standardised methods to evaluate the economic impacts of water-related risks, most researchers either use program evaluation methods (Difference-in-Difference, Synthetic control method) or panel fixed-effect models.

Limitation of econometric models is specific to each application based on the type of the model, control variables, and data quality. Moreover, compared to other types of models such as the integrated-assessment models, econometric models are not capable of simulating different water scenarios, they are rather designed to identify and understand the links between water-related risks and different economic indicators.

Table A B.1. Application of econometric models

Paper	Water-related risk	Sample	Modelling approach	Main results
(Escaleras and Register, 2011[2])	Past flood occurrences	94 countries	Panel fixed-effect model	↘FDI
(Kamber, McDonald and Price, 2013[3])	Drought	New Zealand	VAR	0.3%↘GDP 10%↗ Dairy prices 3%↘ Exchange rate
(Thiede, 2014[4])	Rainfall deficit	Ethiopia	Difference-in-Difference	↗Within-community livestock inequality
(Yamamura, 2015[5])	Flood	86 countries	Panel data, OLS	↗Income inequality
(Keiser and Shapiro, 2017[6])	Water regulation	U.S.	Difference-in-Difference	↗Home value
(Parker, 2018[7])	Flood, Drought	212 countries	Panel fixed-effect model	↗Inflation
(Kuwayama et al., 2019[8])	Drought	U.S.	Panel data, OLS	↘Crop yield No effect on Farm income
(Sheng and Xu, 2019[9])	Drought	Australia	Synthetic control method	18%↘Agricultural TFP
(Paglialunga, Coveri and Zanfei, 2022[10])	Precipitation anomalies	150 countries	Panel data, OLS	↗Inequality
(Russ et al., 2022[11])	Water pollution	17 countries	Panel fixed-effect model	↘Growth of GDP per capita
(Gröschl and Sandkamp, 2023[12])	Flood	China	Panel fixed-effect model	↘Export

Source: (Escaleras and Register, 2011[2]) (Kamber, McDonald and Price, 2013[3]) (Keiser and Shapiro, 2017[6]) (Kuwayama et al., 2019[8]) (Gröschl and Sandkamp, 2023[12]) (Paglialunga, Coveri and Zanfei, 2022[10]) (Parker, 2018[7]) (Russ et al., 2022[11]) (Sheng and Xu, 2019[9]) (Thiede, 2014[4]).

Hydro-Economic modelling (HEM)

Hydro-economic modelling is a specialised approach that combines hydrological and economic principles to understand and analyse the interactions between water resources and economic systems. Its primary objective is to offer a structured framework for evaluating the costs and benefits of various water management strategies and policies. By simulating different scenarios with varying water availability and quality conditions, hydro-economic models provide insights into the economic impacts on sectors such as agriculture, industry, energy production, and municipal water supply. Additionally, these models account for the feedback loops between water and economic systems, recognising that economic activities can affect water demand and quality, which in turn influences future water availability and costs.

While many existing models effectively capture the provisioning ecosystem services of water, such as surface and groundwater provision, there remains a significant gap in modelling maintenance and regulation services. These services, including flood and storm protection, water quality management, filtration, and dilution, are critical to estimate water-related impacts on the economy (Salin, Kedward and Dunz, 2024[13]).

Box A B.1. Example of HEMs

WHAT-IF

The WHAT-IF model, short for Water, Hydropower, and Agriculture Tool for Investment and Financing, is an open-source decision support tool designed to aid in water infrastructure investment planning within the water–energy–food–climate nexus. Developed by researchers from institutions including the Technical University of Denmark and MIT, the model integrates various systems and policies to provide a holistic evaluation of infrastructure projects and their economic impacts.

Source: (Payet-Burin et al., 2019[14])

GCAM-Water

GCAM is a dynamic, long-term, multi-sectoral model that represents the interactions between the economy, energy production and use, land use, water resources, and climate change. It can simulate future scenarios up to the year 2100, providing insights into the potential impacts of different policies and environmental changes.

Source: (U.S. Environmental Protection Agency, 2019[15])

WEST

The Water Economy Simulation Tool (WEST) is designed to predict the impacts of economic and environmental shocks on regions, especially those with vulnerabilities in their food-energy-water (FEW) nexus. Developed by Jeffrey J. Reimer and colleagues at Oregon State University, WEST is a comprehensive model that integrates water, energy, food, and economic data to simulate how changes in these areas interact and affect economic and environmental outcomes.

Source: (Reimer, Babbar-Sebens and Rivera, 2020[16]).

Table A B.2. Application of HEMs

Paper	Water-related risk	Sample	Modelling approach	Main results
(Horridge, Madden and Wittwer, 2005[17])	Drought	Australia	multi-regional CGE model (TERM)	30%↘ Agricultural output 1.6%↘GDP 0.8%↗Unemployment
(Wittwer and Griffith, 2011[18])	Drought	Australia	Dynamic multi-regional CGE model (TERM-H2O)	↘GDP ↗Unemployment
(PAUW et al., 2011[19])	Drought, floods	Malawi	hydrometeorological crop-loss models with a regionalised CGE	1.7%↘GDP ↗Food import ↘Tabacco export ↗Poverty
(Carrera et al., 2013[20])	Flood	Italy	Spatial-CGE integrated model	↗Flood damages ↘GDP
(Liu et al., 2014[21])	Irrigation shortfalls	126 river basins	CGE model (GTAP-BIO-W)	USD 3.7billion↘ welfare ↘Crop output ↗Inflation ↗Agricultural import
(KOOPMAN et al., 2015[22])	Water availability for irrigation	Netherlands, France, Germany, Belgium	CGE model (GTAP-W)	↗Agricultural prices ↘Agricultural output
(Gebreegziabher et al., 2016[23])	Changes in precipitation	Ethiopia	Dynamic countrywide CGE model	↘Livestock production ↘GDP per capita
(Roson and Damania, 2017[24])	Changes in runoff	World	CGE model (GCAM GTAP)	↘World GDP (especially MENA and South Asian countries)
(Borgomeo et al., 2018[25])	Rainfall decrease	Ethiopia	CGE model	5%↘GDP 10%↘Agricultural GDP
(Taheripour et al., 2020[26])	Water scarcity	Middle-East	CGE model (GTAP-BIO-W)	↘GDP ↗Unemployment ↘Capital demand ↗Crop prices
(Briand et al., 2023[27])	Water scarcity, water insecurity	South Africa	Dynamic Water-CGE model (SAWAT)	0.44%↘GDP 0.76%↗Unemployment 0.47%↘Household consumption
(Knittel et al., 2024[28])	Flood	Europe	Multi-regional CGE model (DIFI COIN-INT)	0.5%↘GDP 0.5%↘Private consumption 2.4%↘Public consumption

Source: (Borgomeo et al., 2018[25]) (Briand et al., 2023[27]) (Carrera et al., 2013[20]) (Gebreegziabher et al., 2016[23]) (Horridge, Madden and Wittwer, 2005[17]) (Knittel et al., 2024[28]) (Liu et al., 2014[21]) (PAUW et al., 2011[19]) (Roson and Damania, 2017[24]) (Taheripour et al., 2020[26]).

Input-output model

Input-Output (I-O) tables offer a valuable tool for understanding the interconnectedness of economic activities and assessing the transmission of risks. By tracking the flow of goods and services within an economy, I-O tables provide insights into direct and indirect dependencies. Multiregional Input-Output (MRIO) tables have gained popularity thanks to their ability to analyse the cascading effects of shocks, particularly in contexts with limited substitutability of natural resources. MRIO tables provide a snapshot of the global economy, mapping sectoral linkages and production networks (NGFS, 2023[29]).

By incorporating water-related data into I-O models, supervisors can gain a more comprehensive understanding of the economic impacts of water-related risks. MRIO models can provide an insight on how water-related shocks in one sector can propagate through the economy, as well as the macroeconomic impacts of this materialised risk.

While there is a large variety of MRIOs tables displaying flows between industries and products in monetary terms, water-focused databases are relatively scarce. The Eora Global Supply Chain provides high-resolution I-O tables of environmental and social data, including water footprint, for 190 countries (Lenzen et al., 2013[30]). Exiobase is a time series of environmentally extended multiregional input-output tables ranging from 1995 to 2011 for 44 countries. Its key water-related indicators encompass freshwater ecotoxicity, green water consumption and withdrawal, freshwater aquatic and sedimental ecotoxicity, and eutrophication (Stadler et al., 2018[31]).

Main limitations of the MRIO models derives from the uncertainties in the MRIO tables. This can be an issue of data source, imputation, temporal gaps, proportionality and homogeneity assumptions, and others (Caggiani, Ottomanelli and Dell’Orco, 2014[32]). Another key issue in MRIO models is the static nature of the technical coefficients of production. Most models assume outputs to always require the same share of inputs, therefore MRIO models should be improved by endogenously updating the technical coefficients of production (NGFS, 2023[29]). For example, (Ye et al., 2021[33]) developed a MRIO framework with endogenous capital, in which they try to quantify environmental pressures associated with China’s capital formation. Although their application is not dedicated to the effects of nature risks on the economy, their results show that blue water stress in China is driven by the real estate sector.

Box A B.2. Exiobase3

EXIOBASE is a comprehensive global dataset comprising environmentally extended multiregional supply-use and input-output tables (MR-SUT and MR-IOT). These tables allow detailed assessment of both direct and indirect input requirements needed to produce specific outputs within various sectors across countries. By integrating 417 emissions and 662 material and resource categories, EXIOBASE significantly enhances the understanding of economic flows. Covering 44 countries, five rest-of-world regions, and 163 industries from 1995 to 2021, it facilitates analysis of environmental impacts, including ecosystem services, biodiversity, and broader environmental footprints associated with production and consumption activities. Key water-related indicators encompass Freshwater Ecotoxicity, green water consumption and withdrawal, Freshwater aquatic and sedimental ecotoxicity, and eutrophication.

Source : https://www.exiobase.eu/.

Table A B.3. Application of input-output model

	Water-related risk	Sample	I-O table	Model	Results
(White et al., 2017[34])	Drought, degradation of water reserves, floods.	UK	UK input-output analytical table	Natural Capital Stress Test (NCST)	↘GVA ↗Economic losses
(Eamen, Brouwer and Razavi, 2020[35])	Water supply restriction	Saskatchewan River Basin in Canada	Canadian supply and use tables	Inter-regional Supply-side IO (SIO) model	↘GDP
(Jenkins et al., 2021[36])	Drought, water scarcity	England and Wales	Exiobase	Static economic supply-side input-output (SIO) model	↘GVA
(Garcia-Hernandez and Brouwer, 2021[37])	Water supply restriction	Great Lakes Basin (Canada)	Canadian supply and use tables	Multi-regional input-output (MRIO) model	↘GDP
(Ortuzar, Serrano and Xabadia, 2023[38])	Drought	World	Exiobase3	Multiregional Input-Output (MRIO) model	↘GVA
(Almazán-Gómez et al., 2023[39])	Drought	Ebro river basin (Spain)	MRIO table	Mix of a partial economic equilibrium and a MRIO model	↘VA ↘Employment
(Wang et al., 2024[40])	Water stress	China	Chinese MRIO tables	Multi-regional input-output (MRIO) model	↘Economic output Impact on domestic trade

Source: (Almazán-Gómez et al., 2023[39]) (Eamen, Brouwer and Razavi, 2020[35]) (Garcia-Hernandez and Brouwer, 2021[37]) (Jenkins et al., 2021[36]) (Ortuzar, Serrano and Xabadia, 2023[38]) (Wang et al., 2024[40]) (White et al., 2017[34])w.

References

[39] Almazán-Gómez, M. et al. (2023), “A Multiregional Input–Output Hydro-Economic Modeling Framework: An Application to the Ebro River Basin”, Water Economics and Policy, Vol. 09/01, https://doi.org/10.1142/S2382624X21400026.

[25] Borgomeo, E. et al. (2018), “The Distributional and Multi-Sectoral Impacts of Rainfall Shocks: Evidence From Computable General Equilibrium Modelling for the Awash Basin, Ethiopia”, Ecological Economics, Vol. 146, pp. 621-632, https://doi.org/10.1016/j.ecolecon.2017.11.038.

[27] Briand, A. et al. (2023), “Assessing the Macroeconomic Effects of Water Scarcity in South Africa using a CGE Model”, Environmental Modeling & Assessment, Vol. 28/2, pp. 259-272, https://doi.org/10.1007/s10666-023-09883-4.

[32] Caggiani, L., M. Ottomanelli and M. Dell’Orco (2014), “Handling uncertainty in Multi Regional Input-Output models by entropy maximization and fuzzy programming”, Transportation Research Part E: Logistics and Transportation Review, Vol. 71, pp. 159-172, https://doi.org/10.1016/j.tre.2014.09.006.

[20] Carrera, L. et al. (2013), “Assessing direct and indirect economic impacts of a flood event through the integration of spatial and computable general equilibrium modelling”.

[35] Eamen, L., R. Brouwer and S. Razavi (2020), “The economic impacts of water supply restrictions due to climate and policy change: A transboundary river basin supply-side input-output analysis”, Ecological Economics, Vol. 172, p. 106532, https://doi.org/10.1016/j.ecolecon.2019.106532.

[2] Escaleras, M. and C. Register (2011), “Natural Disasters and Foreign Direct Investment”, Land Economics, Vol. 87/2, pp. 346-363, https://doi.org/10.3368/le.87.2.346.

[37] Garcia-Hernandez, J. and R. Brouwer (2021), “A multiregional input–output optimization model to assess impacts of water supply disruptions under climate change on the Great Lakes economy”, Economic Systems Research, Vol. 33/4, pp. 509-535, https://doi.org/10.1080/09535314.2020.1805414.

[23] Gebreegziabher, Z. et al. (2016), “Climate change and the Ethiopian economy: a CGE analysis”, Environment and Development Economics, Vol. 21/2, pp. 205-225, https://doi.org/10.1017/S1355770X15000170.

[12] Gröschl, J. and A. Sandkamp (2023), “Flood Events and Plant Level Trade: A Chinese Experience”, No. 389, IFO.

[17] Horridge, M., J. Madden and G. Wittwer (2005), “The impact of the 2002–2003 drought on Australia”, Journal of Policy Modeling, Vol. 27/3, pp. 285-308, https://doi.org/10.1016/J.JPOLMOD.2005.01.008.

[36] Jenkins, K. et al. (2021), “An Integrated Framework for Risk‐Based Analysis of Economic Impacts of Drought and Water Scarcity in England and Wales”, Water Resources Research, Vol. 57/8, https://doi.org/10.1029/2020WR027715.

[3] Kamber, G., C. McDonald and G. Price (2013), Drying out: Investigating the economic effects of drought in New Zealand, Reserve Bank of New Zealand.

[6] Keiser, D. and J. Shapiro (2017), Consequences of the Clean Water Act and the Demand for Water Quality, National Bureau of Economic Research, Cambridge, MA, https://doi.org/10.3386/w23070.

[28] Knittel, N. et al. (2024), “Who bears the indirect costs of flood risk? An economy-wide assessment of different insurance systems in Europe under climate change”, Economic Systems Research, Vol. 36/1, pp. 131-160, https://doi.org/10.1080/09535314.2023.2272211.

[22] KOOPMAN, J. et al. (2015), “WATER SCARCITY FROM CLIMATE CHANGE AND ADAPTATION RESPONSE IN AN INTERNATIONAL RIVER BASIN CONTEXT”, Climate Change Economics, Vol. 06/01, p. 1550004, https://doi.org/10.1142/S2010007815500049.

[8] Kuwayama, Y. et al. (2019), “Estimating the Impact of Drought on Agriculture Using the U.S. Drought Monitor”, American Journal of Agricultural Economics, Vol. 101/1, pp. 193-210, https://doi.org/10.1093/ajae/aay037.

[30] Lenzen, M. et al. (2013), “Building Eora: A Global Multi-Region Input–Output Database At High Country And Sector Resolution”, Economic Systems Research, Vol. 25/1, pp. 20-49, https://doi.org/10.1080/09535314.2013.769938.

[21] Liu, J. et al. (2014), “International trade buffers the impact of future irrigation shortfalls”, Global Environmental Change, Vol. 29, pp. 22-31, https://doi.org/10.1016/j.gloenvcha.2014.07.010.

[29] NGFS (2023), Recommendations toward the development of scenarios for assessing nature-related economic and financial risks, Network for Greening the Financial System, Paris.

[38] Ortuzar, I., A. Serrano and À. Xabadia (2023), “Macroeconomic impacts of water allocation under droughts. Accounting for global supply chains in a multiregional context”, Ecological Economics, Vol. 211, p. 107904, https://doi.org/10.1016/j.ecolecon.2023.107904.

[10] Paglialunga, E., A. Coveri and A. Zanfei (2022), “Climate change and within-country inequality: New evidence from a global perspective”, World Development, Vol. 159, p. 106030, https://doi.org/10.1016/j.worlddev.2022.106030.

[7] Parker, M. (2018), “The Impact of Disasters on Inflation”, Economics of Disasters and Climate Change, Vol. 2/1, pp. 21-48, https://doi.org/10.1007/s41885-017-0017-y.

[19] PAUW, K. et al. (2011), “The economic costs of extreme weather events: a hydrometeorological CGE analysis for Malawi”, Environment and Development Economics, Vol. 16/2, pp. 177-198, https://doi.org/10.1017/S1355770X10000471.

[14] Payet-Burin, R. et al. (2019), “WHAT-IF: an open-source decision support tool for water infrastructure investment planning within the water–energy–food–climate nexus”, Hydrology and Earth System Sciences, Vol. 23/10, pp. 4129-4152, https://doi.org/10.5194/hess-23-4129-2019.

[16] Reimer, J., M. Babbar-Sebens and S. Rivera (2020), “WEST: Water economy simulation tool to predict impacts of economic and environmental shocks”, Advances in Water Resources, Vol. 142, p. 103648, https://doi.org/10.1016/j.advwatres.2020.103648.

[24] Roson, R. and R. Damania (2017), “The macroeconomic impact of future water scarcity”, Journal of Policy Modeling, Vol. 39/6, pp. 1141-1162, https://doi.org/10.1016/j.jpolmod.2017.10.003.

[11] Russ, J. et al. (2022), “The impact of water quality on GDP growth: Evidence from around the world”, Water Security, Vol. 17, p. 100130, https://doi.org/10.1016/j.wasec.2022.100130.

[13] Salin, M., K. Kedward and N. Dunz (2024), Assessing Integrated Assessment Models for Building Global Nature-Economy Scenarios, Banque de France, Paris, https://www.banque-france.fr/en/publications-and-statistics/publications/assessing-integrated-assessment-models-building-global-nature-economy-scenarios (accessed on 21 July 2025).

[9] Sheng, Y. and X. Xu (2019), “The productivity impact of climate change: Evidence from Australia’s Millennium drought”, Economic Modelling, Vol. 76, pp. 182-191, https://doi.org/10.1016/j.econmod.2018.07.031.

[31] Stadler, K. et al. (2018), “EXIOBASE 3: Developing a Time Series of Detailed Environmentally Extended Multi‐Regional Input‐Output Tables”, Journal of Industrial Ecology, Vol. 22/3, pp. 502-515, https://doi.org/10.1111/jiec.12715.

[26] Taheripour, F. et al. (2020), Water in the Balance The Economic Impacts of Climate Change and Water Scarcity in the Middle East, http://www.worldbank.org/gwsp.

[4] Thiede, B. (2014), “Rainfall Shocks and Within-Community Wealth Inequality: Evidence from Rural Ethiopia”, World Development, Vol. 64, pp. 181-193, https://doi.org/10.1016/j.worlddev.2014.05.028.

[1] TNFD (2022), Discussion paper: A Landscape Assessment of Nature-related Data and Analytics Availability, https://tnfd.global/wp-content/uploads/2022/03/TNFD_DataDiscussionPaper.pdf.

[15] U.S. Environmental Protection Agency (2019), Global Change Assessment Model (GCAM), https://cfpub.epa.gov/si/si_public_record_Report.cfm?Lab=OAP&dirEntryId=212503 (accessed on 10 July 2024).

[40] Wang, X. et al. (2024), “Impacts of water constraints on economic outputs and trade: A multi-regional input-output analysis in China”, Journal of Cleaner Production, Vol. 434, p. 140345, https://doi.org/10.1016/j.jclepro.2023.140345.

[34] White, C. et al. (2017), Developing and piloting a UK Natural Capital Stress Test: Final Report, https://www.camecon.com/wp-content/uploads/2019/01/Developing-Piloting-a-Natural-Capital-Stress-Test_AECOM-WWF-Sept2017.pdf#page=24.10 (accessed on 17 September 2024).

[18] Wittwer, G. and M. Griffith (2011), “Modelling drought and recovery in the southern Murray-Darling basin*”, Australian Journal of Agricultural and Resource Economics, Vol. 55/3, pp. 342-359, https://doi.org/10.1111/j.1467-8489.2011.00541.x.

[5] Yamamura, E. (2015), “The Impact of Natural Disasters on Income Inequality: Analysis using Panel Data during the Period 1970 to 2004”, International Economic Journal, Vol. 29/3, pp. 359-374, https://doi.org/10.1080/10168737.2015.1020323.

[33] Ye, Q. et al. (2021), “Linking the Environmental Pressures of China’s Capital Development to Global Final Consumption of the Past Decades and into the Future”, Environmental Science & Technology, Vol. 55/9, pp. 6421-6429, https://doi.org/10.1021/acs.est.0c07263.