Introduction

The ongoing debate about the merits of ESG (Environment, Social, Governance) investing in financial markets requires careful measurement of its effect on portfolio performance. Investors may choose to integrate ESG tilts in their portfolios for different reasons, based on sustainability considerations and/or because they believe that ESG ratings reflect material risks and corresponding performance opportunities. These considerations may be reflected in the investment policy in different ways, ranging from strict exclusion of companies and sectors involved in non-compliant activities to a more nuanced best-in-class approach that selects the companies with the best ESG rankings within each peer group.

A simple comparison between the returns of a sustainability index and the standard underlying index, whether in equities or in credit, can result in a distorted view of the ESG effect on performance. Two such indices could differ in sector allocations, average issue size, and credit ratings-all sources of performance with risk premia of their own. How should we measure the effect of ESG investing on portfolio performance? Do traditional risk factors in both equity and credit markets retain their properties when subjected to ESG constraints? Do measures taken by corporate issuers to improve their ESG profile help their subsequent ratings and the performance of their debt and equity securities? How should investors handle the lack of uniformity in ESG definitions? Addressing all these issues requires a quantitative framework aligned with the systematic approach to investing.

We pursue a consistent parallel analysis of the ESG effect on systematic strategies in equity and bond markets. Applied to security selection these strategies involve the systematic use of financial models for all securities within the investment universe, and the construction of highly diversified portfolios that reflect a number of investment themes, or factors, in a risk-efficient manner. While systematic investing has been in the mainstream of equity investing for decades, it has recently started gaining popularity among bond investors as well. There are several reasons for these past differences and for the recent convergence in acceptance of algorithmic investing between the two markets. Most equities are exchange traded and more liquid than bonds. Equity market data have been broadly available to researchers in academia and the financial industry for many years. As a result, all aspects of quantitative investing in equities-from definition of the factors driving stock returns, to selection signals predictive of future security or sector performance, to portfolio optimization methodologies-have been well researched, exploited by investors, and widely accepted alongside the traditional fundamental, discretionary investment style. In the past few years fixed-income investors also saw increased availability of bond market data from vendors, improved price transparency, increased liquidity due to regulatory reporting requirements to shared databases such as TRACE, and a rise in e-trading, ETFs, and portfolio trading. All of these developments, coupled with the influence of established quantitative insights from the equity markets, enabled the expansion of systematic investing to fixed income, as we discussed in our book, Systematic Investing in Credit (Wiley, 2021). In the current volume, we focus on the intersection of systematic investing with the trend towards ESG integration, particularly on the impact of an ESG ratings tilt ('positive screening') or of ESG-related exclusions ('negative screening') on the performance of systematic strategies in credit and equities and on the valuation of securities. Our objectives are to offer consistent methodologies for measuring the effects of ESG on the performance of equity and fixed income portfolios, to document the historical magnitude of these effects and the related valuation trends, to quantify the impact of ESG constraints on the performance of systematic strategies and style factors, and to measure the efficacy of corporate actions in the sustainability area.

The book is purely methodological and relies on historical analysis of market data,1 offering no subjective views on the merits of ESG investing. This is in line with the long-standing mandate of our research group. The authors are members of the Quantitative Portfolio Strategy (QPS) group, which has been a part of Barclays (and previously Lehman Brothers) Research for over three decades. The group has a unique focus on working with major institutional investors across the globe on any issues of portfolio management that are quantitative in nature. As a result of this focus, research produced by the group tends to be practical and implementable. The group's publications target portfolio managers and other investment practitioners, as well as research analysts and academics. The group's past involvement in the creation of fixed-income indices and expertise in quantitative research in both equities and bonds further helped it develop consistent methodologies across the two markets. To enable parallel analysis in equity and bond markets, we rely on a proprietary issuer-level historical mapping (that accounts for corporate events) between corporate bonds and equity of a given company. The approach taken in this book is fully objective and free of any views or opinions. Rather, we 'let the data speak'.

The conventional definition of systematic strategies includes fully rule-based algorithmic methodologies aimed at improving portfolio performance by generating alpha. Some of them fall into the 'smart beta' category and take advantage of inefficiencies in the design of traditional market indices. Others harvest risk premia associated with risk factors, both traditional and new. In this book, we take a more expansive view of systematic investing to include any aspects of portfolio construction that are quantitative in nature. For example, we will include in this expanded definition methodologies for isolating the ESG risk premium from other unrelated systematic exposures. In the language of systematic investing, a risk factor is a source of portfolio risk independent of other established risk factors, which is priced in the market and is expected to be compensated by extra portfolio return-the risk premium. Is ESG a risk factor? Do bonds issued by firms that have strong ESG ratings have fundamentally different risk profiles than those with low ESG ratings? On the one hand, many proponents of ESG investing hold the view that stronger governance is associated with management quality, and hence corporate decisions that lead to higher investor cash flows. Stronger credentials on the Environmental and Social dimensions may reduce exposure to adverse corporate developments such as litigation, changes in regulation, or changes in customer acceptance. On the other hand, there has been insufficient empirical evidence so far that ESG ratings are indeed associated with systematic risk. In this book, although we use the term 'ESG risk premium' to refer to the isolated ESG-related return (free of any other risk factor exposures and idiosyncratic risk), we are not taking a view on whether ESG exposure is a risk factor that should be expected to carry a risk premium. (In fact, in Chapter 4 we show that for sovereign bonds the ESG-related return is subsumed by the credit rating.) We hope that our work to document the relationships between ESG characteristics will help inform this discussion going forward.

All the materials included in the book reflect original QPS research as it was first published. With few exceptions where an update was essential, we decided against going back and updating the data analysis in individual chapters to avoid any possibility of hindsight tainting the results.

This book is structured in four parts.

In Part I, we address the seemingly simple question of how to measure the returns associated with an ESG tilt in a portfolio or an index. Most sustainable versions of broad market indices in both equity and credit are defined by exclusion of non-compliant issuers or industries. However, the difference in performance between these sustainable indices and the original index cannot be interpreted as return due to ESG, as the two indices differ in sector allocations, credit quality, issuer size, and a number of other characteristics that also affect security returns. Even if sector allocations are constrained to match the broad index, tilting a portfolio within sectors towards high ESG issuers will simultaneously tilt it towards higher rated, large-cap companies, which tend to be more compliant. We propose a methodology for isolating the performance effect of ESG while matching the underlying index in all other risk dimensions, and we document the behaviour of this premium in equities and bonds over time. The ESG risk premium obtained in this exposure-matched way, free of all systematic biases, can differ from the simple performance differential between a sustainable and standard index not only in magnitude, but also in sign. Separately we study the ESG effect on the pricing and performance of sovereign bond portfolios. In addition to our methodology for measuring the performance of 'best-in-class' ESG investing, we also study the effect of the exclusionary approach of Socially Responsible Investing (SRI) on credit portfolio performance. The negative screening of entire industry groups makes it difficult to exactly match index risk characteristics; we therefore introduce a new technique for measuring the performance effect of such...