Print
E-MailGeorge Serafeim is Professor of Business Administration at Harvard Business School. This post is based on a recent paper authored by Professor Serafeim; Malcolm Baker, Professor of Finance at Harvard Business School; Daniel Bergstresser, Associate Professor of Finance in the Brandeis International Business School; and Jeffrey Wurgler, Professor of Finance at New York University.
Climate change is accelerating. Since recordkeeping began in 1880, the six warmest years on record for the planet have all occurred since 2010. One estimate suggests that keeping the world below the 2-degree Celsius scenario, a threshold viewed as limiting the likelihood of devastating consequences, will require $12 trillion over the next 25 years.
In the absence of a global carbon pricing scheme, bond markets will be central to financing these interventions. In this paper, we study the U.S. market for “green bonds,” which we and others define as bonds whose proceeds are used for an environmentally friendly purpose. Examples include renewable energy, clean transportation, sustainable agriculture and forestry, energy efficiency, and biodiversity conservation. Since the first green bond was issued in 2007 by the European Investment Bank (EIB), the market has expanded to include a variety of issuers, including supranationals, sovereigns, corporations, and U.S. and international municipalities. It is a small but increasingly well-defined area of the fixed income markets. Yet in spite of the general acceptance of the notion of a “green” bond, there is not yet a single universally-recognized system for determining the green status of a bond. Green bonds may be labeled and promoted as such by the issuer, such as the 2007 EIB issue; formally certified by a third party according to a set of guidelines; or, labeled green by a data provider, for example Bloomberg. We review the origins of the market and standards for identifying green bonds in the next section.
Our sample includes 2,083 green U.S. municipal bonds issued between 2010 and 2016 and 19 green U.S. corporate bonds issued between 2014 and 2016. Our analysis of pricing and ownership patterns is organized by a framework featuring a subset of investors with a nonpecuniary component of utility, such as a sense of social responsibility from holding green bonds, in addition to standard portfolio mean and variance. In this framework, expected returns include the usual CAPM beta term plus a second term, reflecting demand for a security’s environmental attributes, which illustrates that securities with higher scores—such as green bonds—are priced at a premium. We confirm that green municipal bonds are indeed priced at a premium. After-tax yields at issue for green bonds versus ordinary bonds are roughly 6 basis points below yields paid by otherwise equivalent bonds. Depending on specification, the estimates control for ratings, maturity, the yield curve, tax status, other time-varying characteristics, other bond-specific characteristics, ratings-maturity-yield curve interactions, and even issuer fixed effects. On a bond with a 10-year duration, a yield difference of 6 basis points corresponds to a plausible and economically meaningful 0.60 percentage-point difference in value. Interestingly, this premium doubles or triples for bonds that are not only self-labeled as green (and confirmed by Bloomberg) but also externally certified as green by a third party, according to industry guidelines, and publicly registered with the Climate Bonds Initiative (CBI).
Our framework also makes predictions for ownership concentration of green bonds. Green bonds should be held disproportionately by concerned investors, who must be willing to accept their equilibrium lower returns. This concentration will be particularly strong for small bonds, where tilting away from market weights is less consequential, and when the bond is almost riskless, since risk aversion limits the extent to which concerned investors are willing to pursue a nonpecuniary benefit. Using institutional bond ownership data, we find supportive evidence for these predictions. The Herfindahl-Hirschman index is indeed higher for green bonds, especially relatively small bonds and those rated AAA. Also, echoing the pricing effect, concentration is particularly elevated for CBI certified green bonds.
In brief, our contributions relative to this prior work are to provide: an academic introduction to the U.S. market for green bonds; a consistent framework to study both pricing and ownership patterns; a consistent set of empirical results in a comprehensive sample; and, evidence that certification is important in this emerging market. Nonetheless, the green bond market is just a first step toward addressing enormous problems. There is a commensurate need for additional research on green bonds and other areas of climate finance.
A draft of the paper can be found here.
