Market Effects of Voluntary Climate Action by Firms: Evidence from the Chicago Climate Exchange

Abstract

Are private voluntary environmental actions by firms a sign of mismanagement, or a profitable “win-win” replacement for regulation? Empirical evidence is decidedly mixed. In this study, we use 19 years of monthly stock price returns, from 1991 to 2009, to examine the profitability of participation in CCX, a large voluntary greenhouse gas mitigation program. After controlling for systemic market risk as well as industry-specific shocks, we find statistically significant and positive excess returns for firms that announce their decision to join CCX. In addition, the progress of proposed greenhouse gas legislation (the Waxman–Markey bill) had a positive and large impact on excess returns for CCX member firms, suggesting that a major incentive for firms to join CCX may be to prepare for future regulation. Marginal abatement costs (proxied by the carbon price), on the other hand, were unrelated to excess returns. Our results imply that voluntary approaches should play a role in combating climate change, but that relying on them alone is not enough.

Appendix: Derivation of the Regression Specification

In order to derive a theoretically consistent regression specification, we start by considering the price equation that gives rise to the CAPM while suppressing the subscripts on the coefficients:

$$\begin{aligned} \frac{P_{it}}{P_{ft}}= e^{\alpha t}\cdot \left(\frac{P_{mt}}{P_{ft}}\right)^{\beta } \end{aligned}$$

(A.1)

or, equivalently,

$$\begin{aligned} P_{it}= e^{\alpha t}\cdot P_{mt}^{\beta } \cdot P_{ft}^{-{\beta -1}} \end{aligned}$$

(A.2)

Taking logs and differentiating w.r.t time gives

$$\begin{aligned} \frac{\dot{P}_{it}}{P_{it}}= \alpha +\beta \cdot \frac{\dot{P}_{mt}}{P_{mt}} -(\beta -1) \cdot \frac{\dot{P}_{ft}}{P_{ft}} \end{aligned}$$

(A.3)

where a dot represents a time derivative. Rearranging and discretizing in order to accommodate market data results in

$$\begin{aligned} \frac{\Delta P_{it}}{P_{it-1}} - \frac{\Delta P_{ft}}{P_{ft-1}} = \alpha +\beta \cdot \left(\frac{\Delta P_{mt}}{P_{mt-1}} -\frac{\Delta P_{ft}}{P_{ft-1}}\right) \end{aligned}$$

(A.4)

where $\Delta$ is the first-difference operator. Since proportional first differences are equal to returns, () is equivalent to Eq. (2) in the main text. We now add average prices for “small” and “big” (in terms of market value), “high” and “low” (in terms of book-to-market ratio) and “winners” and “losers” (in terms of the past month’s market performance), which are the basis for the SMB, HML and MOM factors. This leads to the following equation in prices:

$$\begin{aligned} P_{it}= P_{ft}\cdot e^{\alpha t}\cdot \left(\frac{P_{mt}}{P_{ft}} \right)^{\beta _{1}} \cdot \left(\frac{P_{t}^{small}}{P_{t}^{big}} \right)^{\beta _{2}} \cdot \left(\frac{P_{t}^{high}}{P_{t}^{low}} \right)^{\beta _{3}} \cdot \left(\frac{P_{t}^{growth}}{P_{t}^{blue}} \right)^{\beta _{4}} \equiv \Gamma _{t} \end{aligned}$$

(A.5)

Following the same steps as above and using the definitions of the four factors, this becomes Eq. (3) defined in returns.

The RHS of () does not include any firm-specific information but only depends on aggregate market data. We now introduce firm-specific information. Specifically, we include industry-level stock prices $P_{it}^{sic}$, the CCX carbon price $P_{t}^{C}$, and dummies indicating CCX membership $M_{it}$ (1 for CCX members, 0 otherwise) and the passing of the Waxman–Markey bill $W_{t}$ (1 for June 2009, 0 otherwise). Since the effect of the carbon price and Waxman–Markey can be expected to differ between members and nonmembers, we also include interaction terms. In order to be consistent (both in terms of theory and math) with the four-factor model, we add these terms multiplicatively to the price equation:

$$\begin{aligned} P_{it} =\Gamma _{t} \cdot e^{\gamma _{1} M_{it}} \cdot e^{\gamma _{2} W_{t}} \cdot e^{\gamma _{3} M_{it} W_{t}} \cdot \left(\frac{P_{it}^{sic}}{P_{ft}}\right)^{\delta _{1}}\cdot \left(P_{t}^{C}\right)^{\delta _{2}}\cdot \left(P_{t}^{C}\right)^{\delta _{3}M_{it}} \end{aligned}$$

(A.6)

Taking logs and differentiating w.r.t time:

$$\begin{aligned} \begin{aligned}\frac{\dot{P}_{it}}{P_{it}}&= \frac{\dot{\Gamma }_{t}}{\Gamma _{t}} + \gamma _{1} \dot{M}_{it} + \gamma _{2} \dot{W}_{t} + \gamma _{3}\left( \dot{M}_{it}W_{t} +{M}_{it}\dot{W}_{t} \right)\nonumber \\&\quad +\,\delta _{1}\left(\frac{\dot{P}_{it}^{sic}}{{P}_{it}^{sic}} -\frac{\dot{P}_{ft}}{P_{ft}}\right) +\delta _{2}\frac{\dot{P}_{t}^{C}}{{P}_{t}^{C}} +\delta _{3}\left(M_{it}\frac{\dot{P}_{t}^{C}}{{P}_{t}^{C}}+\dot{M}_{it} ln \left(P_{t}^{C}\right)\right) \end{aligned} \end{aligned}$$

(A.7)

Substituting for $\Gamma _{t}$, discretizing and using the fact that all member firms joined CCX before June 2009 (implying that $\Delta M_{it}W_{t-1}=0\,\forall t$) leads to the following specification:

$$\begin{aligned} \begin{aligned}&\frac{\Delta P_{it}}{P_{it-1}}- \frac{\Delta P_{ft}}{P_{ft-1}} \nonumber \\&\quad =\alpha +{\beta _{1}} \left(\frac{\Delta P_{mt}}{P_{mt-1}} -\frac{\Delta P_{ft}}{P_{ft-1}}\right) +{\beta _{2}} \left(\frac{\Delta P_{t}^{small}}{P_{t-1}^{small}} -\frac{\Delta P_{t}^{big}}{P_{t-1}^{big}}\right) \nonumber \\&\quad \quad +{\beta _{3}} \left(\frac{\Delta P_{t}^{high}}{P_{t-1}^{high}} -\frac{\Delta P_{t}^{low}}{P_{t-1}^{low}}\right)\, +{\beta _{4}} \left(\frac{\Delta P_{t}^{growth}}{P_{t-1}^{growth}} -\frac{\Delta P_{t}^{blue}}{P_{t-1}^{blue}}\right) \nonumber \\&\quad \quad + \gamma _{1}\Delta M_{it}+ \gamma _{2}\Delta W_{t} + \gamma _{3}\Delta M_{it-1}\Delta W_{t} \nonumber \\&\quad \quad + \delta _{1}\left(\frac{\Delta {P}_{it}^{sic}}{{P}_{it-1}^{sic}} -\frac{\Delta {P}_{ft}}{P_{ft-1}}\right) +\delta _{2}\frac{\Delta {P}_{t}^{C}}{{P}_{t-1}^{C}} +\delta _{3}\left(M_{it-1}\frac{\Delta {P}_{t}^{C}}{{P}_{t-1}^{C}}+\Delta {M}_{it} ln \left(P_{t-1}^{C}\right)\right)\qquad \end{aligned} \end{aligned}$$

(A.8)

Writing the proportional first differences as returns and substituting the definitions of the SMB, HML and MOM factors, this becomes Eq. (4).