The Pollution Premium, Hsu et al. (2023)

ABSTRACT

• This paper studies the asset pricing implications of industrial pollution. A long-short portfolio constructed from firms with high versus low toxic emission intensity within an industry generates an average annual return of 4.42%, which remains significant after controlling for risk factors. This pollution premium cannot be explained by existing systematic risks, investor preferences, market sentiment, political connections, or corporate governance. We propose and model a new systematic risk related to environmental policy uncertainty. We use the growth in environmental litigation penalties to measure regime change risk and find that it helps price the cross section of emission portfolios' returns.

Research target

• Investigate the empirical relation between toxic emissions and expected stock returns at the firm level

Methodology

• Pollution & stock relationship: Cross-sectional regression/Fama-MacBeth

• Pollution premium estimates: Sorted-portfolio/Fama-French factor models

Data

• EPA: Pollution data

• Compustat: Finance data

Findings

• Theoretically develop a general equilibrium asset pricing model where it is optimal for the government to replace a weak regulation regime by a strong one if the environmental cost is perceived to be sufficiently high.

• Theory implication: High emission firms’ profitability are affected more than low emission firms, thus display a larger decline in stock prices upon a regime shift and therefore earn higher average excess returns ex-ante.

• Empirical findings: Firms producing more pollution are associated with higher subsequent stock returns.

• FF-factor models: The high-minus-low portfolio strategy based on simple emissions (toxicity-adjusted emissions) yields statistically significant average returns of 5.52% (5.87%) per annum.

• Regime switch: Firms with higher emissions experience more declines in future profits and market value when the policy regime is more likely to shift.