Leadership performance of financial firms on climate change action

Global awareness of the urgent need to decarbonize the economy has been growing. Although legislative and regulatory actions have been lagging, some businesses have emerged as leaders in this process. In particular, financial institutions as information producers and resource allocators play an important role. In order to accelerate the global transition to a low-carbon economy, market participants need to develop the ability to identify and support firms that are leading on climate change action. Using CDP data on ten climate change action metrics for 2013, the authors apply the dichotomous Rasch model to rank the overall climate change action performance of U.S. financial firms across multiple dimensions of this effort. Simultaneously, the results identify the climate change action metrics for which success was most difficult to achieve. The authors show that investors, managers and regulators should consider ranking firms using this more comprehensive methodology rather than the CDP’s Performance Band or the CDP’s Disclosure Score alone when assessing firm leadership in this area. While this study focuses on financial firms, a similar analysis could be conducted for ranking firms in other industries as well. The authors’ results are important for investors, managers and regulators charged with firm performance evaluation and resource allocation in the face of growing pressures to decarbonize the global economy.


Introduction
Global awareness of the urgent need to decarbonize the economy has been growing (SDSN, 2014), and numerous global and local organizations have emerged to address the challenge.While politicians and government regulators have been slow to respond, some businesses have steadily increased their efforts and expanded their roles in mitigating climate change caused by human activity 1 .Corporate carbon reduction targets are currently not mandatory, but many large firms have begun to implement them nonetheless (Byrd et al., 2014)  2 .Concerns center, however, on how slowly broad-based action is unfolding (Byrd et al., 2013).While the momentum for action on climate change was strengthened at the United Nations Climate Change Conference in Paris, France in December of 2015 (United Nations, 2015), the resulting international agreement remains non-binding.This leaves the financial market as one of a few major forces capable of decarbonizing the economy.
For purposes of capital allocation, investors need to know how to measure an organization's perfor-Amy Burnett, Carolin Schellhorn, 2016.Amy Burnett, Ph.D., The Bill Munday School of Business St. Edwards University Austin, Texas, USA.Carolin Schellhorn, Ph.D. Department of Finance Saint Joseph's University, Philadelphia, Pennsylvania, USA. 1 Businesses invest in climate action.(2016, May 13).Retrieved from http://climatenexus.org/learn/private-sector/businesses-invest-climateaction. 2 Science Based Targets, a joint collaboration between CDP (formerly, the Carbon Disclosure Project), the United Nations Global Compact, the World Resources Institute, and WWF, provides guidance to companies that aim to take a leadership role in setting ambitious carbon reduction targets in line with current climate science (http://sciencebasedtargets.org).mance in the area of climate change mitigation and decarbonization so that they can direct resources to the most deserving firms.Climate-related financial and non-financial information disclosure is a critical first step in this reporting and resource allocation process, and several organizations have taken the initiative to address this need (CDSB, 2012; SASB, 2013; CDP, 2015; TCFD, 2016).Beyond climaterelated awareness and information disclosure, firms' climate-related policies and practices, as well as carbon emission reductions, matter to market participants.How is it possible to synthesize information about such a variety of efforts and, then, rank organizations by their overall performance on this important issue?Several studies address the need to engage with companies on climate change risks and ways to reduce holdings of high-carbon assets (IIGCC, 2015), but more research is needed on how firms' climate change action performance is to be assessed.This paper suggests a new, more comprehensive approach to ranking organizations' overall performance on climate change action by using data from the CDP.
We focus on financial institutions since they are one of the most important and influential sectors in the economy.These institutions occupy a central role in information production about creditworthy firms and organizations.Awareness of the need for climate change action in the financial sector is a prerequisite for the responsible allocation of funding across all sectors.In other words, the information production and monitoring by financial firms will support in meaningful and measurable ways all firms' success in formulating and meeting greenhouse gas reduction targets, as well as climate science and policy creation (IIGCC, 2014).In addition, by practicing corporate environmental responsibility and fostering it in other firms, financial institutions are enhancing their own operating performance, as the positive effects of these practices are not confined to the manufacturing sector (Jo et al., 2014).Not only will the entire economy be better off with an efficient allocation of funding consistent with environmental responsibility, but also stakeholders in financial institutions will benefit as well.
This paper assesses the overall performance of U.S. firms in the financial sector with respect to action on climate change.We use the Global 500 Climate Change 2014 Summary Data Set from the CDP (formerly Carbon Disclosure Project), which includes data reported by financial institutions for 2013.Performance rankings generated with a simple Rasch model identify the financial firms that succeed relative to their peers across a range of climate change action (CCA) metrics.In the absence of mandatory reporting and emissions reduction, these firms exhibit exceptional leadership, a remarkable ability to identify priorities and an above-average sense of corporate social responsibility.It is interesting to note that these are attributes of strong managerial talent, which has been linked to more investments in corporate social responsibility (Chatjuthamard et al., 2016).
Along with providing the firm performance rankings, the Rasch methodology simultaneously identifies which CCA metrics are the most difficult to achieve.The next section describes the Rasch model and its application.Section 2 discusses the CDP data, section 3 reports the Rasch results, and the final section concludes.

The Rasch model and application
Schellhorn and Sharma (2013) have previously used a range of financial metrics and a Rasch model to rank the multi-dimensional financial performance of firms in individual industries.In this paper, we take a similar approach to measure the CCA performance of firms in the financial sector.In the human sciences, a dichotomous Rasch model predicts the probability of a person's success on a test that consists of several items by simultaneously measuring person ability and item difficulty.The probability of a correct response increases with higher measures of person ability and with lower measures of item difficulty.Similarly, financial firms being evaluated on CCA are tested on a number of metrics covering various CCA performance dimensions.Managers' ability to move their firms into CCA leadership positions, then, corresponds to person ability, and the difficulty of earning a favorable reading on a particular CCA me-tric corresponds to the difficulty of finding the correct answer to a test item.The dichotomous Rasch model developed by Georg Rasch in 1960 3 , the simplest in the family of Rasch models, defines the conditional probability P ni of a correct answer with score x = 1 (as opposed to x = 0) by person n to a test item i as a function of the difference between the estimated ability of the person (B n ) and the estimated difficulty of the item (D i ): The calculation of probabilities involves an iterative procedure, which estimates person ability and item difficulty on a logit scale with the average logit set to equal zero 4 .The model implies that when person ability is exactly matched by item difficulty (B n D i = 0), the person has a 50% chance of correctly answering the item.When item difficulty exceeds person ability, the probability of success is less than 50%, while it exceeds 50% when person ability is greater than item difficulty.
Any application of the Rasch model requires a certain relationship among the data.In this application, the probability of a given firm's success is a logistic function of the difference between the estimated ability of the firm to lead on CCA performance and the estimated CCA metric difficulty.The data for a set of CCA metrics for firms in the financial sector at a particular time either fit this model, or they do not.Measures of fit, specifically the infit t, the outfit t and the Rasch Metric Reliability Index, provide information about how well any given data set meets the requirements of the Rasch model.The following section explains the CCA metrics data that were provided by financial firms to CDP and are used in our analysis of the firms' leadership performance on CCA.

The use of CDP data
Currently, the disclosure of climate change risks and CCA metrics in the U.S. is not mandatory.Yet, on behalf of hundreds of institutional investors, CDP periodically requests that companies voluntarily disclose a comprehensive set of climate change policy and implementation data.An increasing number of firms have complied over the years likely because of positive valuation effects from both carbon disclosure and reduced carbon emissions as documented by Matsumura et al. (2014).
We use data from the Global 500 Climate Change Summary Data Set provided by the CDP in December 2014 5 to rank U.S. financial firms by their CCA per-formance in 2013.We focus on U.S. firms only because firms' CCA performance is likely to vary across countries with different levels of sensitivity to the climate change issue and differences in the regulatory environment.Our overall CCA performance ranking considers performance along three dimensions: awareness and willingness to participate, carbon disclosure, and carbon performance.We apply ten metrics across the three dimensions.The first of these, the awareness and willingness to participate dimension, includes the two measures of whether a firm answered the CDP questionnaire (Response Status), and whether the firm gave permission to make the response publicly available (Permission Status).The carbon disclosure dimension is measured with three items: whether or not the firm disclosed its Scope 1 and Scope 2 emissions (Scope 1 and 2 Disclosure) 6 , its CDP Disclosure Score for 2013 7  6 Scope 1 emissions are defined as all direct greenhouse gas emissions reported by the firm including, for example, emissions associated with the firm's facilities and vehicles.Scope 2 emissions are all indirect greenhouse gas emissions reported by the firm including, for example, emissions associated with the consumption of purchased electricity, heat or steam. 7The CDP Disclosure Score measures the comprehensiveness of the firm's response, the quality and transparency of its internal data management, and its understanding of climate change issues as evident in the detail provided. 8The CDP Performance Band recognizes actions considered to advance climate change mitigation and adaptation.
values, there is some room for varying the difficulty of achieving success.Given that our study includes some metrics for which favorable readings are easy to achieve, such as Several caveats are important to keep in mind.In the notes that accompany the Global 500 Climate Change Summary Data Set, the CDP emphasizes the need for detailed scrutiny of a company's choice of carbon accounting methodology.A uniform standard currently does not exist, but it is useful to ascertain whether a firm is accounting for its carbon emissions and, if it is, the direction of emission change.Another issue with the raw data submitted to the CDP is that it is selfreported, that is, the CDP scores are based only on the information it receives directly from the companies.No attempt is made to verify the information that is provided 9 .Nonetheless, this information is valuable to market participants as it sheds some light on the extent of a firm's commitment to CCA and environmental responsibility.

Results
The results of our analysis for 2013 reveal a difficulty hierarchy of CCA metrics for the financial firms in our sample along with a ranking of firms by their overall CCA performance (see Tables 1 through 3).The standardized fit statistics for the Rasch methodology, infit t and outfit t, for the difficulty hierarchy all lie between +2 and -2, indicating that the ten CCA metrics fit the requirements of the Rasch model for this sample 10 .In addition, the Rasch Metric Reliability Index is high at 0.86 (the range is 0 to 1), thus, indicating repeatability of the metric difficulty rankings for a similar sample of firms.
Of the ten metrics we analyzed, the CDP Performance Band was the one metric for which a favorable reading (A or A-) was most difficult to achieve with the highest metric difficulty measure of 2.91 logits.Next were the CDP Disclosure Score and Scope 1 and 2 Reduction at 1.62 logits each.These were followed by Scope 1 Reduction at 1.15 logits and Scope 2 Reduction at 0.44 logits.Not surprisingly, responding to the CDP questionnaire was the least difficult accomplishment with a metric difficulty measure of -3.51 logits.In other words, the CCA performance rankings provided by the Rasch method simultaneously consider ten CCA metrics across three different CCA performance dimensions and are, therefore, more comprehensive and better at identifying corporate CCA leaders than any one metric alone.Investors, as well as other stakeholders, attempting to ascertain which financial firms are most committed to CCA and environmental responsibility will likely find that the Rasch analysis gives them a more complete picture.Better performance measurement is likely to improve capital allocation and financial security valuation.

Conclusion
Using a simple Rasch model, we present composite rankings of climate change action performance for U.S. financial firms that reported to the CDP in 2013, and an estimated difficulty hierarchy of the ten climate change action metrics that inform the analysis.Our results suggest that Morgan Stanley and Wells Fargo emerged as the strongest performers across this particular range of metrics assessing climate change action for that year.
Our results synthesize information about a range of variables that are available to institutional investors who value climate change action and environmental responsibility with important implications for security valuation and resource allocation.
As the need to contain rising global average temperatures becomes increasingly urgent, it is critically important that all firms act to mitigate climate change.Taking a leadership role in the absence of mandatory climate change disclosure and performance is remarkable and potentially signals superior managerial ability.Future research may develop increasingly informative Rasch climate change action performance rankings not just for financial firms, but for businesses in other industries as well.Synthesizing information about climate change action metrics from more than one data source to produce the Rasch performance rankings may offer additional insights.Future research might also explore a possible link between leadership in the area of climate change action, investments in other areas of corporate social and environmental responsibility, and long-term financial performance.
Response Status and Permission Status, we chose to restrict the definition of success for both the CDP Disclosure Score and the CDP Performance Band to those of the highest performers, those with scores at 90 or above and bands of A or A-, re-

Table 1 .
"CDP" and capitalization should be consistentDefinition of Climate Change Action metrics "CDP" sample of 32 U.S. financial firms 2013

Table 2 .
11DP" and capitalization should be consistent Difficulty Hierarchy of Climate Change Action metrics "CDP" sample of 32 U.S. financial firms 201311

Table 2 (
cont.)."CDP" and capitalization should be consistent Difficulty Hierarchy of Climate Change Action metrics "CDP" sample of 32 U.S. financial firms 2013 Note: Mean metric difficulty measures for non-extreme metric readings are set to zero by the model by default.More positive (less negative) measures indicate greater difficulty.Infit t and outfit t statistics between +2 and -2 indicate that the sample data meet the requirements of the Rasch model.The reliability index is reported on a 0 to 1 scale with 1 being maximum reliability.

Table 3 .
"CDP" and capitalization should be consistent Climate Change Action Performance Ranking "CDP" sample of 32 U.S. financial firms 2013 Note: More positive (less negative) performance measures indicate better performance.Infit t and outfit t statistics between +2 and -2 indicate that the sample data meet the requirements of the Rasch model.The reliability indices are reported on a 0 to 1 scale with 1 being maximum reliability.Both the Rasch Real Reliability Index and the Cronbach-Alpha Reliability Index are based on raw scores rather than logit measures.Big absolute differences between the indices likely result from differences in the treatment of extreme scores.

Table 4 .
Climate Change Action (CCA) Performance Ranking Versus Ranking by CDP Performance Band & CDP Disclosure Score 2013

Table 4 ,
we present the top twelve financial firms in the 2013 CDP sample sorted first by the CDP Performance Band and then by the CDP Disclosure Score, two of the most restrictive stand-alone CCA metrics used in this study.Five companies earned a CDP Perfomance Band of A, including Morgan Stanley and Wells Fargo (the top two, according to the results of the Rasch analysis), but three of those companies (BNY Mellon, Bank of America and Goldman Sachs Group Inc.) had higher CDP Disclosure Scores than Morgan Stanley and Wells Fargo.If investors naively used just those two CDP metrics, they might incorrectly judge Morgan Stanley and Wells Fargo, ranked fifth and fourth by those metrics, as not being the top leaders for CCA when, in fact, they appear to be according to the Rasch analysis.