Abstract
There is clear scientific evidence of the shift in the probability distribution of climate-related disasters in recent decades. Is this shift reflected in the behavior of forward-looking measures of economic activity such as real exchange rates? I evaluate the role of different belief formation assumptions on the ability of the model to predict the response of real exchange rates to climate-related disasters. I consider Bayesian and backward-looking belief updates as well as static beliefs with no update or a one-time update. To do so, I construct a version of the Farhi-Gabaix (2015) framework augmented with explicit belief formation. I use two approaches to model calibration and simulate the model for 47 countries for 1964–2019 using actual data for climate-related disasters. I find that in general differences in belief formation do not have much effect on the model fit because the productivity loss component dominates the predicted response. Specifically, I find that even in recent years there is no evidence of Bayesian beliefs being a better fit for the data.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
I do not study the effects of transition risks that may arise from climate mitigation policies, greening technologies, or greening consumer and investor preferences. For a study of transition risk effects on commodity currencies, see Kapfhammer et al. (2020).
Such effect is found, for example, in Strobl and Kablan (2017).
It could be argued that a more rational belief model will include a deterministically trending prior for climate disasters arrival rate. A calibration with such a trend produces even less precise predictions than the Bayesian update model considered. Because the trend depends on beliefs about future climate scenarios, this would add another layer of complexity to the model, and therefore this possibility is not considered in the paper. The results with deterministic trends are available upon request.
Similar pictures for big climate disasters and non-climate disasters are available upon request.
Estimating a negative binomial model that allows for overdispersion produces nearly identical results, indicating that Poisson regression is a good fit.
The estimates are based on a balanced panel with missing data replaced with zeros.
The results are not specific to the functional form of the utility function.
In FG setup, B varies over time but not across countries.
In FG setup, F varies across countries and over time.
In FG setup, p only varies over time. I allow it to vary over time and across countries.
In FG setup, \(\phi\) only varies by country. I allow it to slowly change over time.
FG show that in the absence of disasters interest rate is \(r-\delta\).
Moreover, the literature links climate beliefs to personal experience with extreme weather (Myers et al., 2013; Rudman et al., 2013; Dai et al., 2015; Howe et al., 2014; Linden, 2015; Frondel et al., 2017) and shows that economic agents act on perception of local effects of climate change (Spence et al., 2011; Blennow et al., 2012). For a recent review of the literature on this topic, see Taylor (2014).
Simulated disaster probabilities do not depend on calibration of model parameters. Simulations are based on disaster arrival rate and realizations taken directly from the data.
Pre-sample is longer for countries with data availability starting later than 1964.
For countries where no disasters are observed or reported prior to 1960 this value is set to zero.
Other parameters in the pre-sample are the same for all countries and are identical to FG calibration: \(\widehat{\omega _{i0}} =1\), \(F=1\), \(B=0.66\).
Ibarrarán et al. (2007) argue that there are important cumulative macroeconomic effects of natural disasters, while Kalkuhl and Wenz (2020) estimate the substantial decline in productivity resulting from climate change even in the absence of extreme weather events. Felbermayr and Gröschl (2014) find significant negative effects of natural disasters on economic growth. At the firm level, Pankratz et al. (2021) show that extreme heat and floods affect global customer–supplier relationships and Kruttli et al. (2021) find a large and long-lasting increase in implied volatility of the stock options of firms exposed to hurricanes.
In a robustness analysis, I use a 0/1 indicator of whether there was at least 1 disaster instead of the count of disasters. The results are qualitatively the same even though simulated probabilities are affected by this change.
The optimization is conducted via simple grid search due to the complexity and non-linearity of the model. The allowable ranges for the grid search were informed by FG calibration as well as data calibration and were as follows: \(F\in [0.9;1]\), \(B\in [0.6;1]\), \(\mu \in [0;1]\).
For fit calibration, I need an alternative approach when fitting the model to non-climate shocks, because B and F are not available. For this part of the exercise, I compute the slow-moving component of H as a moving average of its past history: \({\overline{H}}_{it} = 1/t * \sum _{s=0}^t H_{is}\). Then, \({\widehat{H}}_{it} = H_{it} - {\overline{H}}_{it}\). Since non-climate events frequency does not trend over time, the results produce stable estimates for each country over time, approximating time-invariant component of H well.
The repetition is useful due to random draws of disaster realization in the pre-sample and, in case of Rational belief update, due to random draws from the Gamma distribution for the belief update.
While the model produces a balanced panel, real exchange rate data is not available for all countries going back to 1964. However, I verified that the distribution of model-generated parameters is not materially different for the unbalanced panel for which the real exchange rate data are available. Comparison tables are available upon request.
Export share and TFP growth are from the Penn World Table, share of fuel exports is from the World Bank, exchange rate regime classification is from Harms and Knaze (2021).
In a robustness test, I add Exports/GDP in the regression of the observed real appreciation, to allow for larger response in countries more open to trade, but such control does not make a difference.
Alternatively, I interact the number of disasters with an indicator of whether the country is classified as risky or safe, which produces very similar results. Note that the response of simulated exchange rates is not mechanical, because in the simulation a country can be classified as safe in some years and risky in other years. Here I split countries based on their average classification into safe and risky over the entire sample period.
For this reason, in the analysis of differences across beliefs, as a robustness test, I compare the differences between the resilience component only and the data as opposed to model prediction and the data, across belief types.
Backward beliefs model produces a slight bias – the model predicts on average more appreciation than observed in the data.
The results for data calibration are available upon request and continue to produce little to no difference across belief types.
References
Adam, Klaus, and Stefan Nagel. 2023. Chapter 16—Expectations data in asset pricing. In Handbook of Economic Expectations, ed. Rüdiger. Bachmann, Giorgio Topa, and Wilbert van der Klaauw, 477–506. Academic Press.
Alok, Shashwat, Nitin Kumar, and Russ Wermers. 2020. Do fund managers misestimate climatic disaster risk. The Review of Financial Studies 33 (3): 1146–1183.
Batten, Sandra, Rhiannon Sowerbutts, and Misa Tanaka. 2020. Climate change: Macroeconomic impact and implications for monetary policy. In Ecological, Societal, and Technological Risks and the Financial Sector, ed. Thomas Walker, Dieter Gramlich, Mohammad Bitar, and Pedram Fardnia, 13–38. Cham: Springer International Publishing.
Blennow, K., J. Persson, M. Tomé, and M. Hanewinkel. 2012. Climate change: Believing and seeing implies adapting. PLoS ONE 7 (11): e50182.
Burke, M., S. Hsiang, and E. Miguel. 2015. Global non-linear effect of temperature on economic production. Nature 527: 235–239.
Cheema-Fox, Alexander, George Serafeim, and Hui Wang. 2022. Climate change vulnerability and currency returns. Financial Analysts Journal 78 (4): 37–58.
Cortés, Kristle Romero, and Philip E. Strahan. 2017. Tracing out capital flows: How financially integrated banks respond to natural disasters. Journal of Financial Economics 125 ((1): 182–199.
Dai, Jing, Martin Kesternich, Andreas Löschel, and Andreas Ziegler. 2015. Extreme weather experiences and climate change beliefs in China: An econometric analysis. Ecological Economics 116: 310–321.
Dell, Melissa, Benjamin F. Jones, and Benjamin A. Olken. 2012. Temperature shocks and economic growth: Evidence from the last half century. American Economic Journal: Macroeconomics 4 (3): 66–95.
Engel, Charles. 2016. Exchange rates, interest rates, and the risk premium. American Economic Review 106 (2): 436–74.
Farhi, Emmanuel, and Xavier Gabaix. 2015. Rare disasters and exchange rates. The Quarterly Journal of Economics 131 (1): 1–52.
Felbermayr, Gabriel, and Jasmin Gröschl. 2014. Naturally negative: The growth effects of natural disasters. Journal of Development Economics 111: 92–106.
Frondel, Manuel, Michael Simora, and Stephan Sommer. 2017. Risk perception of climate change: Empirical evidence for Germany. Ecological Economics 137: 173–183.
Gabaix, Xavier. 2008. Variable rare disasters: A tractable theory of ten puzzles in macro-finance. American Economic Review 98 (2): 64–67.
Gourinchas, Pierre-Olivier., and Aaron Tornell. 2004. Exchange rate puzzles and distorted beliefs. Journal of International Economics 64 (2): 303–333.
Guo, Kai. 2007. Exchange rates and asset prices in an open economy with rare disasters. Federal Reserve Bank of Dallas Working Paper.
Gupta, Rangan, Tahir Suleman, and Mark E. Wohar. 2018a. Exchange rate returns and volatility: The role of time-varying rare disaster risks. The European Journal of Finance 25 (2): 190–203.
Gupta, Rangan, Tahir Suleman, and Mark E. Wohar. 2018b. The role of time-varying rare disaster risks in predicting bond returns and volatility. Review of Financial Economics 37 (3): 327–340.
Hadri, Hajare El, Daniel Mirza, and Isabelle Rabaud. 2019. Natural disasters and countries’ exports: New insights from a new (and an old) database. The World Economy 42 (9): 2668–2683.
Harms, Philipp and Jakub Knaze. 2021. Effective exchange rate regimes and inflation. Working Paper 2102, Gutenberg School of Management and Economics, Johannes Gutenberg-Universität Mainz.
Heinen, Andréas, Jeetendra Khadan, and Eric Strobl. 2022. The price impact of extreme weather in developing countries. The Economic Journal 129 (619): 1327–1342.
Howe, Peter D., Hilary Boudet, Anthony Leiserowitz, and Edward W. Maibach. 2014. Mapping the shadow of experience of extreme weather events. Climatic Change 127 (2): 381–389.
Huang, H.H., J. Kerstein, and C. Wang. 2018. The impact of climate risk on firm performance and financing choices: An international comparison. Journal of International Business Studies 49: 633–656.
Ibarrarán, María Eugenia., Matthias Ruth, Sanjana Ahmad, and Marisa London. 2007. Climate change and natural disasters: Macroeconomic performance and distributional impacts. Environment, Development and Sustainability 11 (3): 549–569.
Ivanov, Ivan T., Marco Macchiavelli, and João. A.. C.. Santos. 2022. Bank lending networks and the propagation of natural disasters. Financial Management 51 (3): 903–927.
Javadi, Siamak, and Abdullah Al Masum. 2021. The impact of climate change on the cost of bank loans. Journal of Corporate Finance 69: 102019.
Jones, Benjamin F., and Benjamin A. Olken. 2010. Climate shocks and exports. American Economic Review 100 (2): 454–59.
Kalkuhl, Matthias, and Leonie Wenz. 2020. The impact of climate conditions on economic production. Evidence from a global panel of regions. Journal of Environmental Economics and Management 103: 102360.
Kapfhammer, Felix, Vegard H. Larsen, and Leif Anders Thorsrud. 2020. Climate risk and commodity currencies. Norges Bank Working Paper 18/2020.
Klomp, Jeroen. 2017. Flooded with debt. Journal of International Money and Finance 73: 93–103.
Krueger, Philipp, Zacharias Sautner, and Laura T. Starks. 2020. The importance of climate risks for institutional investors. The Review of Financial Studies 33 (3): 1067–1111.
Kruttli, Mathias, Brigitte Roth Tran, and Sumudu Watugala. 2021. Pricing poseidon: Extreme weather uncertainty and firm return dynamics. Federal Reserve Bank of San Francisco Working Paper 2021–23.
Lee, Sinyoung O., Nelson C. Mark, Jonas Nauerz, Jonathan Rawls, and Zhiyi Wei. 2022. Global temperature shocks and real exchange rates. Journal of Climate Finance 1: 100004.
Lewis, Karen K., and Edith X. Liu. 2017. Disaster risk and asset returns: An international perspective. Journal of International Economics 108: S42–S58.
Mallucci, Enrico. 2022. Natural disasters, climate change, and sovereign risk. Journal of International Economics 139: 103672.
Myers, T., E. Maibach, C. Roser-Renouf, et al. 2013. The relationship between personal experience and belief in the reality of global warming. Nature Climate Change 3: 343–347.
Osberghaus, Daniel. 2019. The effects of natural disasters and weather variations on international trade and financial flows: A review of the empirical literature. Economics of Disasters and Climate Change 3 (3): 305–325.
Pankratz, Nora M. C. and Christoph Schiller. 2021. Climate change and adaptation in global supply-chain networks. European Corporate Governance Institute — Finance Working Paper 775/2021 June 2021.
Rudman, Laurie A., Meghan C. McLean, and Martin Bunzl. 2013. When truth is personally inconvenient, attitudes change: The impact of extreme weather on implicit support for green politicians and explicit climate-change beliefs. Psychological Science 24 (11): 2290–2296 (PMID: 24058064).
Spence, A., W. Poortinga, C. Butler, et al. 2011. Perceptions of climate change and willingness to save energy related to flood experience. Nature Climate Change 1: 46–49.
Stott, Peter. 2016. How climate change affects extreme weather events. Science 352 (6293): 1517–1518.
Strobl, Eric, and Sandrine Kablan. 2017. How do natural disasters impact the exchange rate: An investigation through small island developing states (SIDS)? Economics Bulletin 37 (3): 2274–2281.
Taylor, Andrea L., Suraje Dessai, and Wändi Bruine. de Bruin. 2014. Public perception of climate risk and adaptation in the UK: A review of the literature. Climate Risk Management 4–5: 1–16.
Valente, Joao Paulo, Kaushik Vasudevan, and Tianhao Wu. 2022. The role of beliefs in asset prices: Evidence from exchange rates, April 2022. https://doi.org/10.2139/ssrn.3872077.
van der Linden, Sander. 2015. The social-psychological determinants of climate change risk perceptions: Towards a comprehensive model. Journal of Environmental Psychology 41: 112–124.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This paper is prepared for the Annual Research Conference (ARC) of the IMF, 2022. I thank Ted Liu and Anirban Sanyal for their expert research assistance. For insightful comments, I thank two anonymous referees, Oya Celasun, Andrei Levchenko, participants of the 2022 IMF Annual Research Conference as well as seminar participants at the U.S. Treasury Office of Financial Research, Federal Reserve Bank of San Francisco Virtual Seminar on Climate Economics, Norges Bank, Bank of England, University of Luxembourg, and UC Santa Cruz. All errors are mine.
Appendix
Appendix
1.1 Effect of Disaster on Productivity
To proxy for the disaster productivity loss factor (\(F_i\)) I estimate, for each country, a time series regression of a TFP change on an indicator of a disaster in the previous year.
The estimates \(\beta _{i,\textrm{TFP}}\) and 95% confidence intervals are reported in Fig. 9. For countries where the estimates are positive, \(F_i\) is set to be equal to 1. For those with negative estimates, \(F_i = 1 - \beta _{i,\textrm{TFP}}\).
Estimates of \(\beta _{i,\textrm{TFP}}\)
1.2 Model Calibration and Simulated Parameters
1.3 Safe and Risky Countries: Classification and Model Fit
See Tables 6, 7, and 8 and Figs. 10, 11, and 12.
Difference between data and resilience component across beliefs. Plotted is the distribution of difference between actual percent real appreciation and the resilience component of percent real appreciation predicted by the model for the same country and the same year. Horizontal line indicates the median of the difference. “Bayesian” is calibration based on rational Bayesian belief update, “Backward” is based on historical adaptive belief formation, “Step” is based on average disaster frequency in 1965–1990 and 1990–2019 in the data, constant within each of the two periods, “No update” keeps disaster frequency belief at 1965–1990 level
Difference between data and resilience component across beliefs and time periods. Fit calibration. Plotted is the distribution of difference between actual percent real appreciation and the resilience component of percent real appreciation predicted by the model for the same country and the same year. Horizontal lines indicates the median of the difference. Dashed lines indicate interquartile range. “Bayesian” is calibration based on rational Bayesian belief update, “Backward” is based on historical adaptive belief formation, “Step” is based on average disaster frequency in 1965–1990 and 1990–2019 in the data, constant within each of the two periods, “No update” keeps disaster frequency belief at 1965–1990 level
Difference between data and resilience component across beliefs and time periods: safe versus risky countries. Fit calibration. Plotted is the distribution of difference between actual percent real appreciation and the resilience component of percent real appreciation predicted by the model for the same country and the same year. Horizontal lines indicates the median of the difference. Dashed lines indicate interquartile range. “Bayesian” is calibration based on rational Bayesian belief update, “Backward” is based on historical adaptive belief formation, “Step” is based on average disaster frequency in 1965–1990 and 1990–2019 in the data, constant within each of the two periods, “No update” keeps disaster frequency belief at 1965–1990 level
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
