Essays on Empirical Asset Pricing

Regulation, Rents, and Returns
Regulation of the US economy has increased significantly over the past decades. Differ-ent economic theories offer testable predictions on how regulation affects firms, raising the question of which theory best aligns with the data. Existing empirical studies focus either on a regulatory change affecting a particular industry or firm-specific regulatory exposure. This chapter complements the existing literature by analyzing the economic effects following changes in all federal regulations across all industries from 1985 to 2021. Focusing on the entire stock of regulations allows me to address questions not captured in prior studies.
I find that incumbent firms in industries subject to more (less) regulation subsequently have significantly positive (negative) risk-adjusted equity returns. In an attempt to give a more causal interpretation, I show that the result remains significant in a triple-difference analysis when hedging this long-short regulatory portfolio with a similar portfolio of non-US industries. Increases in regulation are associated with higher future markups and returns on invested capital. These findings can be explained by higher entry barriers and lower levels of competition, as firm net entry declines after increases in regulation.
To determine who benefits within industries, I show that firms with higher asset intensity outperform even more so around regulatory increases than those with lower asset intensity. This result aligns with the view that incumbent assets are exempt from new regulations, such as environmental requirements, which are the main source of regulatory changes in the data. Overall, these findings support increasing entry costs as being the dominant impact of regulation and reject both the public interest theory of regulation or that regulatory impact is a result of random unintended consequences.

In Search of the True Greenium
ESG investors and sustainable finance regulators seek to improve the environment by lowering the cost of capital for green firms while raising it for brown firms. The success of this mechanism depends on the size of the resulting “greenium”, defined as the difference in the cost of capital between green and brown firms. However, greenium estimates varytremendously across papers and, while most academic papers report a negative greenium, many practitioners expect a positive one. In this chapter, we search for the true greenium.
In the first part of the paper, we replicate and extend existing papers that estimate the equity greenium using realized returns. We collect data for 23 greenness measures from the literature and estimate the corresponding greeniums based on a unified methodology and data samples that are extended over time and across the world. Our main finding is that all these greenium estimates are insignificant when we account for multiple testing. In fact, we show that estimating the greenium using realized returns requires centuries of data, while the literature often uses little over a decade.
In the second part of the paper, we seek to estimate the greenium with less noise and study its properties. Guided by a theoretical model, we construct an “aggregate green score.” To reduce noise in the dependent variable, we consider forward-looking expected returns instead of realized returns. Using these two components, we estimate a statistically significant annualized equity greenium of −30 basis points (bps) per standard deviation increase in the aggregate green score. This corresponds to an expected return differential of 56 bps per year for a green-minus-brown portfolio. We also show that the equity greenium has become more negative over time and is more negative in greener countries. Finally, we find significantly negative greenium estimates for corporate and sovereign bonds, as well as the weighted-average cost of capital.

Carbon Home Bias
Firms are under increasing pressure to decarbonize their operations, giving rise to carbon transition risk. Given their large assets under management, institutional investors play a key role in hedging and pricing this risk. They also influence the transition to a low-carbon economy through their investment allocation decisions and shareholder engagement. In this chapter, we use granular data at the institutional holding level to analyze how these investors tilt their portfolios with respect to carbon emissions. One important dimension we examine is whether the location of a firm relative to the institution matters.
From a portfolio diversification or carbon transition risk perspective, the location of a company alone should not matter; ignoring frictions and preferences, only the exposure to carbon-transition risk should be relevant. Countries may have different carbon policies—both current and expected—that influence local carbon transition risk, but these should impact domestic and foreign institutions alike.
We find that institutional investors in aggregate significantly reduce their holdings of firms with higher carbon emissions, both in terms of carbon intensity and the level of emissions. When we separate portfolios into local and foreign holdings, we find that the tilting effect is mostly driven by underweighting or divestment of foreign stocks. If anything, domestic stocks with high carbon emissions are overweighted. These results reveal the presence of a carbon home bias, above the well-documented home bias effect in investor portfolios. The bias is stronger in countries with weaker climate policies and increased significantly among US institutions following the unexpected 2016 election of Donald J. Trump.