Sustainability in Fixed Income: Does ESG Actually Help Bond Investors?
Previous: Liquidity and Trading in Fixed Income
ESG is everywhere. Every asset manager talks about it. Every pitch deck has a slide on it. But does sustainability data actually help you pick better bonds?
Chapter 10 of Systematic Fixed Income by Scott Richardson tackles this question head on. The short answer: sustainability measures have weak direct investment value for credit investors. But that does not mean they are useless. A systematic approach can improve a portfolio’s sustainability profile without hurting returns. Let us walk through the evidence.
The Rise of Sustainable Investing
Interest in sustainable investing has exploded. The United Nations Principles for Responsible Investing (PRI) launched in 2006. By 2020, around $100 trillion in assets were committed to PRI. The Global Sustainable Investment Alliance reported $35.3 trillion invested sustainably as of 2020, representing 35.9 percent of total assets under management.
Most of this attention started in equity markets. But the focus has shifted to fixed income too. For corporate bonds, the connection is natural. You are lending money to a company. Would you not want to know if that company is managing environmental, social, and governance risks well?
Does ESG Help Explain Credit Spreads?
Richardson draws heavily on research by Diep, Pomorski, and Richardson (2021). They tested whether MSCI ESG scores help explain credit spreads across four corporate bond universes: US investment grade, US high yield, European investment grade, and European high yield.
When you run the regression without controlling for credit risk, ESG scores are statistically associated with credit spreads. Companies with higher sustainability scores have lower spreads. Makes sense. Better-run companies should be less risky.
But here is the catch. The effect is small. A one standard deviation change in sustainability scores moves credit spreads by only 3 to 8 percent. And once you add a proper default probability forecast into the model, the relationship gets even weaker. Only US investment grade bonds still show a significant link.
In plain terms: ESG scores are picking up some credit risk information. But a good default model already captures most of it. The incremental value is modest.
Does ESG Predict Future Returns?
This is the question every investor really cares about. Can sustainability scores tell you which bonds will outperform?
The answer from the data is basically no. The researchers tested whether ESG scores predict future credit excess returns across all four bond universes. The results were not statistically significant in any of them. With or without controlling for expected returns from other signals, sustainability measures just do not forecast bond returns in a meaningful way.
What About Risk?
There is slightly better news on the risk side. Companies with higher sustainability scores tend to have lower future return volatility. The relationship is statistically significant for most bond universes.
But again, the economic magnitude is small. A one standard deviation improvement in ESG scores is associated with a 7 to 15 percent reduction in future idiosyncratic return volatility. Once you control for other risk measures, that shrinks to 2 to 5 percent. Helpful, but not exactly a silver bullet.
The Honest Takeaway on Direct Investment Value
Richardson is upfront about the limitations. The data only covers 2012 to 2020. Only one ESG provider (MSCI) was used. And the analysis looked at aggregate ESG scores, not tailored measures for specific industries.
Could different data, different time periods, or more targeted sustainability measures produce stronger results? Maybe. But Richardson warns against hunting for results across every possible measure and sample. For a systematic investor, that kind of data mining is a “cardinal sin.”
Incorporating Sustainability Without Sacrificing Returns
Here is where things get more interesting. Even though ESG data does not strongly predict returns or risk, you can still build a more sustainable portfolio without giving up performance.
The researchers tested a systematic global high yield portfolio with four sustainability layers:
- Static exclusions. Remove companies involved in controversial weapons, fossil fuels, and tobacco.
- Tactical exclusions. Drop the bottom 10 percent of MSCI ESG scores.
- Sustainability tilt. Push the portfolio’s overall ESG score 10 percent above the benchmark.
- Carbon reduction. Keep the portfolio’s carbon intensity at least 25 percent below the benchmark.
The combined exclusions removed about 18 percent of market capitalization in the US high yield universe and about 13 percent in European high yield. That sounds like a lot, but there were still enough liquid issuers across sectors to run a proper portfolio.
The Cost of Going Green
Every sustainability constraint moves the portfolio away from the unconstrained optimal solution. So the rational expectation is some drag on expected returns.
But the actual cost turned out to be surprisingly small. The expected return frontier for tactical exclusions and sustainability tilts was downward sloping but nearly flat. You could apply meaningful sustainability improvements with minimal return sacrifice.
Why? Two reasons. First, there were still enough non-excluded issuers to do security selection. Second, the original systematic portfolio was not actively seeking out the worst ESG offenders anyway. If your starting portfolio already avoids the bottom of the barrel, excluding them formally does not cost you much.
For carbon intensity reduction, the story was similar up to a point. You could reduce carbon by up to 70 percent with little impact on returns. Beyond that threshold, the cost climbed sharply. That is because carbon emissions are concentrated in a small number of industries. Remove those and you are fine. Try to push further and you start running out of room.
The Proof: Sustainable vs. Standard Portfolio
When the researchers plotted cumulative returns for the sustainable systematic portfolio against the standard systematic portfolio, the two lines were nearly identical. The sustainable version achieved meaningfully better ESG scores and lower carbon intensity while delivering almost the same returns.
The portfolio characteristics (number of bonds, rating breakdown, sector mix, maturity profile) were all very similar too. The sustainability constraints did not distort the portfolio in any obvious way.
Governance Measures That Actually Work
Not all ESG data is created equal. Richardson highlights that the “G” in ESG tends to have the most investment relevance.
One example: aggregate accruals. This measures how aggressively a company’s management uses accounting discretion. Companies with high accruals tend to have lower future profitability, more negative analyst revisions, and higher risk of earnings restatements. Research shows this holds for bond returns too. Corporate issuers with high accruals (poor governance) experience lower future credit excess returns.
So while broad ESG scores may not predict much, specific governance measures can add real value to a bond portfolio.
A Tricky Problem: Flows vs. Fundamentals
Richardson raises an important point. As investors pile into high-ESG securities, those securities experience positive returns from buying pressure alone. But once that flow stops, those same securities may earn lower future returns because they are now expensive.
This makes it hard to draw conclusions from recent data. Did high-ESG bonds outperform because sustainability matters, or because everyone was buying them at the same time? Separating flows from fundamentals is a challenge for every investor.
Sustainability for Government Bonds
The chapter also covers sustainable investing for government bonds, and this is much harder. Government bond benchmarks are very concentrated. The top 10 sovereign issuers make up 95 percent of typical developed market indices. The US and Japan alone account for 63 percent. If your sustainability framework excluded either country, your tracking error would go through the roof. So tilting is far more practical than outright exclusion.
How do you even measure sustainability at the country level? Robeco published a framework that breaks it into environmental, social, and governance dimensions with multiple sub-measures. But there is no industry consensus. Different asset managers use different data, different weights, and different definitions.
Bridgewater Associates anchored their sustainable strategy to the United Nations Sustainable Development Goals (SDGs). There are 17 SDGs covering everything from clean energy to reduced inequality. The idea is to identify which SDGs matter most to the asset owner and then estimate each country’s sensitivity to those goals. APG Asset Management in the Netherlands had about 80 billion euros invested consistent with SDGs as of early 2021.
The Bottom Line
ESG data has limited direct investment value for bond investors today. It does not strongly predict credit spreads, future returns, or risk once you already have a good credit model in place.
But that is not the end of the story. A systematic investment approach can integrate multiple sustainability constraints (exclusions, tilts, carbon targets) without meaningful sacrifice in risk-adjusted returns. The key insight is that the cost of sustainability is not zero, but it is far lower than most people assume. For asset owners who care about sustainability objectives alongside financial returns, that is genuinely good news.
The governance component of ESG shows the most promise as an actual investment signal. And as measurement improves and data coverage expands, the direct investment relevance of sustainability data may grow over time.
Next: Putting It All Together: Systematic Fixed Income in Action
Book: Systematic Fixed Income: An Investor’s Guide Author: Scott A. Richardson, Ph.D. ISBN: 9781119900139 Publisher: John Wiley & Sons, 2022