WORKING DRAFT. NOT FOR ATTRIBUTION.
Revealing Climate ChangeOpinions throughInvestmentBehavior:Evidence from Fukushima
Zhen Lei[*]
Anastasia Shcherbakova[†]
October 2014
Abstract: In this study we present a novel research approach to obtaining behavior-based evidence of regional climate change attitudes, using the 2011 Fukushima nuclear plant incident as a natural experiment. Our approach allows us to produce the first non-survey-based empirical evidence of a trans-Atlantic divide in public opinion on the environment and climate change, and an estimate of the relative monetary value that investors assign to fossil-based and renewable energy. This value is based on the perceived potential of these fuel types to substitute for nuclear generation in the aftermath of the Fukushima crisis. We carry out an event study to examine differences in abnormal returns of coal and renewable energy companies on European and American stock exchanges. We find that investors trading on U.S. markets exhibit a significantly more favorable perception of the profitability of coal stocks (generating a $4.2 billion increase in cumulative coal market capitalization), while investors trading on European exchanges display a more favorable perception about the profitability of renewable energy stocks (adding over $3 billion to cumulative renewable market capitalization).
Key words: Climate Change, Public Opinion, Event Study, Financial Markets, Stock Returns, Fukushima
1
1.Introduction
There has been an evident trans-Atlantic divide on climate policies in the United States and Europe. Europe has set an ambitious emissions reduction target under the Kyoto Protocol, implemented multiple policies and measures including cap-and-trade to achieve that target, and played an important role in post-Kyoto international negotiations. U.S. legislators, by contrast, failed to support the Kyoto Protocol, often questioned the basic science of climate change, and failed in 2010 to enact a climate bill to launch a cap-and-trade scheme to curtail carbon emissions (Carlarne, 2006; Schmidt & Haifly 2012;Skjærseth, Bang and Schreurs, 2013).
Such a trans-Atlantic divide on climate policy has been, at least in part, attributable to a reported difference in public opinion on climate change between the United States and Europe.For example, surveys conducted in 2009 by the Pew Research Center and the World Bank show that 44 percent of U.S. respondents believed at that time that global warming wasa very serious problem. A much higher percentage of respondents in France (68 percent)and Germany (60 percent)agreed with this statement. Only 46 percent of U.S. respondents thought climatechange would harm people in theircountry now or within 10 years, whereas 67 percent of respondents in France believed so (Brechin andBhandari, 2011).[‡]
Surveys and opinion polls presently constitute the primary method of obtaining individuals’ views on climate change and other environmental concerns (Brechin and Bhandari, 2011;Lachapelleet al., 2012;Lorenzoni and Pidgeon, 2006). However, the political and emotional charge that fuels the debate on climate change makes obtaining accurate and honest views on such subjects rather challenging. While useful, public opinion polls give us only a relatively simple—and often general—sense of how people feel about a certain issue. Moreover, self-reported opinions have been found to misrepresent individuals’ true beliefs and corresponding actions (Podsakoff et al., 2003). For instance, survey-based studies appear to lead to a paradox that “Americans seem concerned about global warming, yet view it as less important than nearly all other national or environmental issues” (Leiserowitz, 2006).
This paper takes a novel approach to understanding individuals’ true attitudes toward climate change by directly observing actions that people take when presented with a potential energy and environmental crisis—actions that reveal their perceptions of how grave environmental and climate change concerns are. Being able to observe what people do in response to a challenge, rather than what they say, provides more compelling evidence of one’s understanding of said challenge, especially when observed actions are tied to monetary outcomes.
More specifically, we examine behavior of investors in U.S. and European financial markets during the unfolding of the Fukushima crisis that reflects their perceptions about future profitability of various energy assets (i.e. stocks of fossil fuels and renewable energy).Our approach exploits an intuitive link between people’s attitudes toward the concept of climate change and their relative perceptions for the profitability prospects of different energy sources. Survey-basedstudieshave found that people more concerned about climate change tend to have a more optimistic and positive evaluation of renewable energy than of fossil fuels.[§] This implies that an individual’s level of environmental and climate concern likely frames his assessment of the relative future profitability of various energy types. We draw on this linkto gauge potential regional differences in climate change attitudes through actualinvesting behavior of financial market participants in the United States and Europe. These participants are mostly local investors,as established by the home country bias literature.[**]
We also address an important challenge of differentiating investor perceptions of the global profit potential of specific energy types on one hand from investor expectations about local policy response to the Fukushima crisis on the other. The former would be based on global profit expectations and thus provide evidence of investor’s environmental biases. The lattermay have nothing to do with one’s environmental attitudes, as it would impact investment decisions through only local profit expectations.
We use Japan’s Fukushima nuclear power crisis of March 2011 as a natural experiment, and base our study on stock market data from the time period encompassing this event. We apply event study methodology to these data to investigate how investors in U.S. and European financial markets, with potentially different attitudes toward climate change and the environment,envisioned the worldwide response to the nuclear crisis.[††]
As emergency workers struggled to contain the scale of the Fukushima crisis in its initial days, public support for nuclear power weakened noticeably and the prospects of nuclear power worldwide dimmed.[‡‡] A decline in nuclear generation without an offsetting reduction in energy demand implies a need for compensating generation from other energy sources.
The anticipatedshift in the global generation portfolio led investors to rebalance their energy asset holdings. We argue that the rebalancing was done in a manner consistent with investor perceptions of how the world would respond to the nuclear crisis (that is, their expectations about which energy type would become the most likely substitute for nuclear generation). The increase in demand for the products and services of these energy companies would make them more profitable and therefore more attractive to investors.[§§] Because the actions we observe have direct financial outcomes, our approach provides evidence that reflects the relative monetary value that investors assign to fossil-based and renewable energy in the immediate aftermath of the Fukushima crisis.
Our results indicate that investment behavior does indeed reflect investors’ environmental perceptions.We observe positive financial effects accruing to a portfolio of global coal companies traded on U.S. markets, on the order of 8.6 percent in excess returns or $4.2 billion in market capitalization, but not to those listed on European exchanges. By contrast, European exchange data reveal larger positive outcomes for a portfolio of global renewable energy companies than was observed on U.S. exchanges. Over the same time period, renewable stocks traded on European markets earned total excess returns of 13.8 percent (or over $3 billion in market capitalization), while renewable companies listed on U.S. exchanges gained only 3.8 percent in excess returns (which translates to about $1.7 billion in market capitalization).These results suggest that investors active on U.S. exchanges were more skeptical of climate change and perceived more optimistic prospects for coal energy, relative to investors active on European exchanges. European investors, on the other hand, were likely to be more concerned about the effects of climate change and thus formed a more favorable profitability expectation for renewable energy.
To the best of our knowledge, this is the first study to estimate the difference in monetary value placed by investors on coal and renewable energy assets in the United States and Europe, as reflected by their prospects in the aftermath of the Fukushima crisis. Having a quantitative assessment of investors’ environmental perceptions is important because it leads to creation of more realistic and effective environmental strategies. For instance, no matter how much U.S. environmental regulators favor renewable generation, they will not be able to encourage capital flows to renewable sources if investors don’t foresee high profitability potential for renewable energy and technology companies. Since investors’ capital allocation decisions over time feed into firm and industry profitability, then actions taken by owners of capital implicitly determine which government policies stand to succeed or fail.
This study is also the first to examine actual behavior, rather than opinions reported to surveyors and polling organizations, to provide empirical evidence on the trans-Atlantic difference in public environmental and climate change opinions. It is worth noting that because financial investors comprise a significant but partial share of any country’s adult population, our results reflect environmental attitudes among a subsample of the general populace.[***]However, provided that behavior-based evidence is more compelling than evidence drawn from population surveys and opinion polls, our approach presents a good complement to the survey-based approach. Our regional approach can also be scaled down and used to evaluate country-level differences in public attitudes toward climate change.[†††]
The paper is organized as follows: in the next section we provide a sketch of our research framework and hypotheses, as well as the econometric specification used in our main empirical analysis; Section 3 provides details about our data sources and stratification methods for data samples; Section 4 presents a discussion of our empirical results; and Section 5 concludes with some policy implications.
2.Research Framework
2.1 Global profitability prospects vs. local policy response
One challenge we face in this study is in distinguishingbetween investors’ perceptions of global energy profitability and their expectations about local policy response to the Fukushima crisis. On one hand, investors form profitability expectations based on their understanding of the global response to the crisis, which survey-based evidence suggests is framed by their environmental attitudes (Spence et al., 2010). On the other hand, investors’ expectations about how domestic policymakers will respond to the Fukushima crisis will also affect investment decisions by changing the relative profit potential of various energy firms locally.To make sure that we are observing activity driven by global profit perceptions, we restrict our data sample to publicly traded electricity fuel source companies, namely producers of coal and renewable energy technologies,that sell their products or have operations inmultiple global markets, and calculate outcomes based on sampled stocks’ trading and operating regions.[‡‡‡]
In contrast to companies operating solely in the United States or European Union, the overall profitability of energy firms with such global operations is determined by multiple markets (including both the United States and European Union) and is thus not fully dependent (and in some cases, not at all dependent) on the region in which the stock shares of these firms are traded. This means that implementation of local policies favoring a single type of energy source would have a muted effect on profit potential of firms dealing in that specific energy type. This allows us to rule out local policy expectations as the main cause of any differences observed in local stock market outcomes. We hypothesize that such differences, if observed,would instead be driven by investors’ perceptions about global profitability prospects of certain energy sources, which reveal their intrinsic attitudes toward climate change.
Let us provide some support for the above hypothesis through a simple example. China MingYang Wind Power Group and XinJiangGoldwind Science & TechnologyCompany are two prominent Chinese wind turbine manufacturers. MingYang Wind Power is traded on NASDAQ in the United States, while XinJiangGoldwindis traded on Germany’s FrankfurtStock Exchange. Both companies sell and install wind turbines in China, United States, Europe, and other global markets. If, in response to the Fukushima crisis, European governments decided to boost support for renewable energy projects but the U.S. government did not, the impact on the economic prospects of MingYangwould not necessarily be worse than on XinJiang, because both firms have sales in the U.S and in Europe. Therefore, assuming that investors in the United States and Europe are sophisticated, have equal access to information and make rational assessments about firms’future profit potential,the demand for and prices of these two companies’ stocks, determined by their global potential for future profitability, should not vary from one regional exchange to another. However, if investors in the United Statesare in general more skeptical about climate change, they mayon average have a more optimistic view about the prospect of coal-fired electricityas a viable substitute for nuclear energy in both the United States and Europe, relative totheir European counterparts. Hence, investors’ demand for MingYang stock in U.S. markets might be lower than that for XinJiang stock in European markets, consistent with investors’ translation of expected global response into profitability expectations.
One might also be concerned about the presence of naïve investors. If investors do not understand well the global nature of the sampled companies, they may simply assume that companies whose stocks are listed on their regional exchanges mainly operate within their region. Such investors would then base profitability expectations for various energy stocks on their anticipated domestic policy response to the Fukushima crisis, and rebalance their energy investments accordingly. However, our data sample contains energy stocks that have global operations and are traded on multiple exchanges around the world. The naïveassumption would then lead an investor to give equal treatment to all stocks within the same energy category and thereby assign inappropriate profitability expectations to stocks of companies that have financial and physical presence in several regions around the world. To account for this possibility, we carry out separate analyses on securities that are listed on financial markets of only one geographic region, and those cross-listed in both regions. If naïve investors dominate and both cross-listed and region-specific stocks are given equal treatment, we should observe no difference in returns outcomes amongthese two categories of stocks.
Finally, there is also the possibility that a global company may nevertheless do the majority of its business in a single region, which may drive its decision to list its shares on the local market. In this case, stock market responses to the Fukushima crisis in different regions could still reflect investors’ expectations about domestic policies in the aftermath of the crisis, rather than their profitability expectations for certain fuel types that are influenced by environmental opinions. To check this, we refine our sample further by focusing only on those companies that have major operations outside of their listing region (see footnote 18). Focusing only on companies with major operations outside of their listing region would lend further support to the hypothesis that investors’ responses to the crisis reflect their attitudes toward climate change and their global profit expectations for certain energy types.
2.2 Event study methodology
The importance of correctly identifying the event day and length of the event window has been highlighted and addressed in event study literature.[§§§] Our task of defining the event day is simplified by the fact that the earthquake that struck northern Japan, being a natural disaster, was in all likelihood unexpected. We confirm the accuracy of this event day by referencing the Fukushima Nuclear Accident Update Log, in which the International Atomic Energy Agency (IAEA) recorded all Fukushima developments. The first IAEA Fukushima Nuclear Log update concerning problems at the nuclear plant was issued at 21:10 GMT on Friday, March 11, 2011. This corresponds to 16:10 Eastern Standard Time (10 minutes after the closing bell at NYSE and NASDAQ) and 22:10 Central European Time (well after trading stopped at all European markets).[****] Wedefine the event day (day 0) as March 11, 2011 for both U.S.- and European-listed securities and include one trading week of pre-earthquake data in our event window to illustrate any diversions in financial trends in the earthquake’s aftermath.
We then proceed to define an appropriate event window—the relevant time frame during which the Fukushima crisis affected stock market outcomes. Our data includes three trading weeks of post-earthquake observations, since we are interested in short-term market response to the Fukushima crisis, for which investor perceptions and instincts play important roles.[††††] We use post-earthquake data to construct two measures of stock responses, abnormal stock turnover and abnormal stock returns, to determine the timing and duration of investor activities following the earthquake.