Abstract
The aim of this paper is to study the performance of carbon-based portfolios when all emissions scopes are accounted for. We formalize low-carbon mean-variance portfolio strategies by integrating a carbon penalty to a constrained mean-variance optimization framework. We resort to direct and indirect emissions, split between Scopes 1–2 and Scope 3, across geographical zones (Europe and US) and data providers (Refinitiv and Carbon4 Finance). Our results show that it is possible to cut emission intensities in half at least with virtually no loss in Sharpe ratio for reasonable levels of the carbon constraint. These results are valid across various choices of risk aversions, and irrespective of emissions data provider. For a sustainability-aware investor, these low-carbon portfolios are associated to a higher level of welfare. We find that the corresponding allocations are shifted towards assets with higher returns while keeping the portfolio's volatility unchanged. Our results add to the literature contending that sustainable investing is not costly.
