New on FinoLib · Volume III: Finance Encyclopedia
New on FinoLib, my finance education website: I have published an article called Green Bonds Explained and Why ESG Ratings Diverge, from Volume III: Finance Encyclopedia. Here is what the article explains, written as a short expert briefing; the full piece, with every worked example, is one click away.
7.3 The Problem of Ratings Divergence
The idea in plain words
ESG ratings for the same company agree far less than credit ratings do. Across the major providers they correlate only about 0.38 to 0.71, against roughly 0.99 for credit ratings. Most of the disagreement comes from measuring the same issues with different indicators, not from giving the issues different weights.
Key points
■Berg, Kölbel and Rigobon attribute 56% of divergence to measurement, 38% to scope and 6% to weights.
■A rater effect means disagreement is systematic, so averaging two ratings does not cancel it.
■In the worked example, a different emissions measure explains 69% of a 1.45-point gap between two raters.
A closer look
Unlike traditional credit ratings (Capital Markets Part 3), where the major agencies usually land on the same grade for a given bond, ESG ratings for the same company often disagree. The best-known measurement is Berg, Kölbel and Rigobon, “Aggregate Confusion: The Divergence of ESG Ratings” (Review of Finance, 2022). Across six major providers, pairwise correlations of ESG ratings ranged from 0.38 to 0.71; in the working-paper version the authors set this against a 0.994 correlation between Moody’s and S&P credit ratings. A correlation near 0.5 means that a company in one provider’s top quintile is often in another provider’s middle of the pack.
Decision rule
Treat a single ESG rating as one opinion, never as a fact. If two providers place a company more than one quintile apart, do not average them: find the pillar where they disagree and check the underlying data yourself, usually emissions, controversies or labor indicators. If you are building an index or screen, test whether your holdings change when you swap providers; if more than about a quarter of names change, the screen is measuring the provider’s method, not the companies.
The costliest mistake
Assuming disagreement is random noise that averages away. The study’s rater effect shows it is systematic, so averaging two biased raters gives a biased average. In the example above, averaging Rater A (5.8) and Rater B (7.25) gives 6.5, a score that rests on neither the disclosed nor the estimated emissions figure. An investor who sets a cut-off of 6.0 on the average buys the company; one who checks the data and finds the disclosed emissions behind Rater A’s 4.0 does not. The fix is to resolve the measurement question, not to blend the outputs.
Recap
ESG ratings for the same company diverge far more than credit ratings, mainly because providers measure the same issues with different indicators, and a rater effect makes the disagreement systematic. Treat any rating as one opinion, resolve disagreements by checking the underlying data, and note that since July 2026 the EU supervises rating providers.
7.4 Green Bonds
The idea in plain words
A green bond is an ordinary bond whose issuer promises to spend an amount equal to the proceeds on eligible environmental projects, and to report on it. The credit risk is the issuer’s, as with any bond. The green label changes the disclosure, and sometimes the price by a few basis points, but not the repayment.
Key points
■The ICMA Green Bond Principles have four core components: use of proceeds, project selection, management of proceeds and reporting.
■Taxonomies define what counts as green; the EU Taxonomy requires substantial contribution, no significant harm and minimum safeguards.
■The European Green Bond Standard requires at least 85% of proceeds in taxonomy-aligned activities.
Decision rule
As an investor, price a green bond off the issuer’s conventional curve: if the greenium is wider than about 5 bps, you are paying for the label, so check that you need it for a mandate. Then read three documents before calling it green: the framework (which projects, judged against which taxonomy or criteria), the external review, and the latest allocation report (how much is spent, how much is refinancing of old assets).
The costliest mistake
Treating “green” as a credit feature. A green bond’s default risk and recovery are the issuer’s; the label changes neither. An investor who accepts a 6 bp greenium on a $50 million holding gives up $50,000,000 × 0.0006 = $30,000 a year, $231,652 in present value over 10 years at 5% (30,000 × 7.7217), for no reduction in credit risk, and a green bond of a weak issuer will fall with that issuer’s other debt. Price the credit first, then decide whether the label is worth paying for.
What the full article covers
In the full article I take you through these sections:
■7.3 The Problem of Ratings Divergence
■7.4 Green Bonds
It includes worked examples, a look under the hood at the mechanics, a clear decision rule, the one mistake that costs people the most, a recap, a reveal-on-click quiz and the official sources behind every figure.
Volume III: Finance Encyclopedia
Read the full article on FinoLib
The complete article, Green Bonds Explained and Why ESG Ratings Diverge, with every worked example, diagram, quiz and source, is free to read in Volume III, Finance Encyclopedia on finolib.com.
Explore the Finance Education Series: Vol. II — Capital Markets · Vol. I — Global Economy · Vol. III — Finance Encyclopedia · Vol. IV — Finance Careers · Vol. V — Personal Finance