A new Constrained IPCA Model and its Application in Climate Finance

On the application of the constrained IPCA model in climate finance: We study whether the US stock market is pricing exposures to climate risks through an extension of the instrumented principal component analysis (IPCA) of Kelly, Pruitt, and Su (2019). 
In our specification, the factor loadings are allowed to be functions of both “financial” and “environmental” company-specific characteristics. By using a constrained IPCA model, we allow for the presence of different sets of orthogonal factors whose loadings are driven by only one of the two types of characteristics. We identify one “environmental” factor which is important for the out-of-sample pricing of stocks in the Oil and Utilities sectors, above and beyond “financial” factors, which suffice to explain the cross section of stock returns of the companies in the other sectors.
 

Time-varying Environmental Alphas, Betas, and Latent Factors in Corporate Bonds: We study the impact of environmental risks on a large panel of individual US corporate bonds through the lenses of a latent factor model where the different alphas, pricing factors, and risk exposures (betas) of returns can have either a purely environmental or financial connotation. We use an extension of the Instrumented Principal Component Analysis (IPCA) model of Kelly, Pruitt and Su (2019) that allows for two generic groups of metrics to shape the dynamics of alphas and betas. Our model reveals that widely used measures for the sustainability footprint of firms, such as emissions and environmental scores, are weakly related to systematic factors of their bond returns. 
Nonetheless, the same sustainability metrics can be used to form an “anomaly” portfolio producing abnormal returns (alphas) both in- and out-of-sample. We show that this portfolio significantly improves the risk-return profile of the investment opportunity set when added not only to traditionalbond pricing factors, but also to the Fama and French equity factors.