Climate risks and U.S. stock‐market tail risks: A forecasting experiment using over a century of data

• Published date: 01 June 2023
• Author: Afees A. Salisu,Christian Pierdzioch,Rangan Gupta,Reneé Eyden
• Date: 01 June 2023
• DOI: http://doi.org/10.1111/irfi.12397

ORIGINAL ARTICLE

Climate risks and U.S. stock-market tail risks:

A forecasting experiment using over

a century of data

Afees A. Salisu

1,2

| Christian Pierdzioch

3

| Rangan Gupta

2

|

Reneé van Eyden

2

1

Centre for Econometrics & Applied Research,

Ibadan, Nigeria

2

Department of Economics, University of

Pretoria, Hatfield, South Africa

3

Department of Economics, Helmut Schmidt

University, Hamburg, Germany

Correspondence

Christian Pierdzioch, Department of

Economics, Helmut Schmidt University,

Holsten hofweg 85, P.O.B. 700822, 22008

Hamburg, Germany.

Email: macroeconmoics@hsuhh.de

Abstract

We examine the predictive value of the uncertainty associ-

ated with growth in temperature for stock-market tail risk

in the United States using monthly data that cover the sam-

ple period from 1895:02 to 2021:08. To this end, we mea-

sure stock-market tail risk by means of the popular

Conditional Autoregressive Value at Risk (CAViaR) model.

Our results show that accounting for the predictive value of

the uncertainty associated with growth in temperature, as

measured either by means of standard generalized auto-

regressive conditional heteroskedasticity (GARCH) models

or a stochastic-volatility (SV) model, mainly is beneficial for

a forecaster who suffers a sufficiently higher loss from an

underestimation of tail risk than from a comparable

overestimation.

KEYWORDS

asymmetric loss, climate risks, forecasting, stock market, tail risks

JEL CLASSIFICATION

C22, C53, G10

1|INTRODUCTION

Climate change, associated with increased temperature as well as its volatility, poses a large aggregate risk to the

economy and the financial system of the United States (U.S.) due to the occurrences of rare disasters (Giglio

Received: 20 September 2021 Revised: 9 May 2022 Accepted: 22 September 2022

DOI: 10.1111/irfi.12397

© 2022 International Review of Finance Ltd.

228 International Review of Finance. 2023;23:228–244.

wileyonlinelibrary.com/journal/irfi

et al., 2021). In essence, disasters abound when temperature reaches or crosses a threshold level and capture the

idea of tail-risk related to global warming (Pindyck, 2012). At the same time, macroeconomic developments in the

U.S. play a key role for the world economy. Despite a declining share in world GDP, U.S.-based transnational corpo-

rations, and U.S. ownership of such corporations, play a leading or, depending on the sector that is being studied,

even a dominant role, implying that the extent of globalization of the U.S. economy cannot be downplayed

(Starrs, 2013). It is, therefore, not surprising that developments in the U.S market mirror the dynamics of the global

economy, and that any shock that hits the U.S. economy, such as the financial crisis of 2007, will unfold a reverberat-

ing impact on the global economy. Meanwhile, following the rise in the demand for water and cooling in the U.S. in

the wake of severe mega-droughts (Stahle et al., 2000,2007), significant efforts have been undertaken to shed light

on the effects of global climate change on various sectors of the U.S. economy (see, e.g., Karl et al., 2009; Knowlton

et al., 2011). In particular, the nexus between climate change and exposure to extreme heat (temperature) is well

established (Luber & McGeehin, 2008). In this process, the U.S. financial markets play a key role because, as also

withnessed by the global financial crisis of 2007, they have the potential to pose systemic threats to global financial

stability (Alvarez et al., 2020).

Against this background, quite a few recent studies (see for example, Balvers et al., 2017; Bansal et al., 2021;

Bansal et al., 2016; Donadelli et al., 2017; Donadelli, Jüppner, Paradiso, & Schlag, 2021; Donadelli, Jüppner, &

Vergalli, 2021; Engle et al., 2020) have extended general equilibrium models of rare disaster risks to incorporate cli-

mate risks driven by first- and second-moment shocks of growth in temperature to formalize the theoretical channels

via which the economy and the stock market is impacted. For example, Balvers et al. (2017) argue that temperature

shocks arising due to uncertainty of climate change raise the cost of equity, and subsequently, result in a loss of

wealth. In the same vein, Donadelli, Jüppner, Paradiso, and Schlag (2021) argue that temperature-volatility shocks

undermine equity valuations and produce non-negligible welfare costs. Peillex et al. (2021) show that extreme heat

above 300C depresses stock-market activity and, thereby, market returns (Hou et al., 2019). In general, the results

of these significant research efforts indicate that climate risk tends to reduce productivity and/or the stochastic

depreciation rate of capital, and also affects investors' propensity to trade via its implications for distraction, mood,

etc., to produce an adverse impact on macroeconomic variables and equity valuations.

1

According to this line of

research in behavioral finance, the nexus between temperature, mood, and investor behavior works, on the one-

hand side, through the temperature-aggression channel, where lower temperature leads to aggression and propen-

sity to engage in a more risky venture. On the other-hand side, higher temperature leads to apathy, which depresses

risk-taking behavior (Cao & Wei, 2005). The foregoing suggests a negative relationship between temperature and

stock returns. This has equally been validated in some other empirical literature (Floros, 2011; He & Ma, 2021;

Piard, 2013). In contrast to the foregoing, another line of literature establishes a positive nexus (Makkonen

et al., 2021) while others submit a no-relationship between the two variables (Chandra, 2021; Jacobsen &

Marquering, 2008; Lu & Chou, 2012).

The key assumption underlying rare-disaster models is that the entire universe of assets in an economy is

exposed to an aggregate jump-risk factor. It follows that, even though in the cross section, some assets are more

exposed to such a tail event than others, such a jump-risk factor should be an important driver of the time-series var-

iation in the tails of individual asset returns (Barro, 2006,2009; Rietz, 1988). In other words, we can hypothesize that

the jump-risk factor associated with the dynamics of the mean and the volatility of the growth of temperature, that

is, climate risks, has predictive power for movements in the tail risks of the aggregate stock market.

2

Tail risks have

been shown to lead stock returns and real-economic-activity-related variables (Almeida et al., 2017; Chevapatrakul

et al., 2019; Hollstein et al., 2019; Kelly & Jiang, 2014; Salisu, Gupta, & Ogbonna, 2022d).

3

Understandably, deducing

the future path of tail risks in real-time is an important question for both investors and policymakers (Gkillas

et al., 2021; Gupta et al., 2021).

Against this backdrop, given that out-of-sample forecasting is considered to be a more robust test of predictabil-

ity compared to an in-sample analysis in terms of the predictors and econometric model being used

(Campbell, 2008), the objective of our empirical research is to analyze the role of the growth and conditional

SALISU ET AL.229