The Most Common Mistake in NZIF Sovereign Alignment

Context: how alignment shares became a portfolio KPI

Across the institutional sovereign fixed-income community, the Net Zero Investment Framework (NZIF) developed by the IIGCC has become a common reference point for portfolio-level climate alignment. NZIF asks investors to classify holdings into alignment categories (typically labels such as "Achieving Net Zero," "Aligned," "Aligning," "Committed to Aligning," and "Not Aligning") and then to report the share of the portfolio that sits in the more advanced categories. That share has become a headline KPI in many institutions, appearing in TCFD-reporting, sustainability committee dashboards, ESG working group discussions, and increasingly in public Net Zero commitments where a target is expressed as a future alignment share.

For sovereign holdings, classifying countries into NZIF categories is not a trivial exercise, and several institutions have addressed it by anchoring the classification in an external sovereign climate rating. The CCPI is a common choice. Its ratings are familiar, externally produced, peer-reviewed, and structured into five categories: from "very high" down to "very low." A natural mapping has emerged: countries rated "very high" or "high" by the CCPI are treated as "Aligned" or "Aligning" under NZIF, and the share of the portfolio invested in those countries becomes the institution's sovereign alignment share.

This is a reasonable design choice. It uses an established, transparent external metric to populate a framework that asks for categorical judgments. But the design has a structural feature worth understanding before the resulting alignment share is read as a measure of climate progress.

What sits behind the rating categories

The CCPI's overall rating is not constructed against absolute thresholds. A country is not classified as "high" because it has crossed a fixed climate-performance line in the way a sovereign credit rating uses fixed leverage or debt-service thresholds. Instead, the CCPI overall rating is derived from the CCPI overall score, which is itself produced by an annual normalization across the assessed country universe, and then sorted into the five rating categories using thresholds that, in effect, set a roughly stable share of countries in each band every year.

The practical consequence is that the rating categories are, by design, relative bands. The share of countries that fall into "very high" or "high" stays broadly constant from year to year, because the bands themselves are defined relative to the prevailing distribution of country performance. A country can move from "medium" into "high" not because its underlying climate performance has crossed a fixed line, but because it has improved its position relative to others. The same logic applies in reverse: a country can drop out of "high" even when its underlying climate trajectory has not deteriorated, simply because other countries have improved more quickly.

This relativity is not a flaw. The CCPI is, by design, a comparative assessment of country performance, and a comparative assessment naturally produces comparative categories. The friction only arises when these comparative categories are used as building blocks for a portfolio-level KPI that is read as a measure of absolute climate progress over time.

Why a categorical view can be harder to read than a numerical one

The same underlying issue can show up either as a moving numerical score or as a moving categorical classification. The numerical form is, in some ways, easier to handle. A score is visibly continuous, and a reader confronted with a small fluctuation is usually willing to ask what it means.

A categorical classification feels different. When a country is described as "Aligning" rather than "Committed to Aligning," or when a portfolio's reported alignment share moves from one year to the next, the language carries the suggestion of a clear, fixed judgment, the kind of judgment one expects from credit ratings, regulatory classifications, or compliance categories. The natural reading is that the country has either met the criteria or it has not, and that the share has either grown or shrunk for a real reason.

That reading is well-founded in many institutional contexts. It is not well-founded here. When the categories are derived from relative bands, the same intuitive language describes a different kind of movement: a country can change category without any change in its real-world climate position, and a portfolio's alignment share can move without anything in the portfolio's holdings having changed in a climate-relevant way. The categorical form does not change the underlying mathematics; it changes how the underlying mathematics is read.

What this means for alignment-share tracking

Three patterns recur in practice when alignment shares are tracked over multiple reporting cycles.

First, the alignment share can decline even when no held country has deteriorated. If countries outside the portfolio improve faster, they push held countries down within the rating distribution. A country previously rated "high" can drop into "medium" without changing anything about its emissions, energy mix, or climate policy in absolute terms, and the portfolio's alignment share contracts as a result.

Second, the alignment share can stay flat for long periods even when held countries are genuinely improving. Because the share of countries in the upper categories is structurally constrained, a country can improve substantially in absolute climate terms and remain in the same rating category for several consecutive years.

Third - and most consequentially - multi-year alignment trajectories embedded in public commitments can become difficult to interpret. An institution that has publicly committed to a 2030 interim alignment share and a 2050 full alignment share has implicitly assumed that the categorical structure is stable enough over those horizons for the share to move primarily as a function of portfolio composition and underlying country progress. If a meaningful part of the share's movement comes instead from the periodic recalibration of the bands, headline progress and underlying progress can drift apart over the commitment period without that drift being visible in the reported number.

Reading alignment shares in practice

For sovereign fixed-income portfolio managers, the practical reading is methodological rather than evaluative. NZIF remains a coherent and broadly adopted framework. The CCPI remains a credible and methodologically transparent sovereign climate rating. The point is however: when CCPI ratings are used to populate NZIF alignment categories, the resulting portfolio-level alignment share is a relative-positioning measure dressed in categorical language, and it can be read more accurately if treated as such.

Three orientations help. First, treat alignment-share movement as a positioning signal rather than as a direct measure of climate progress. An alignment share rises or falls because the portfolio is more or less concentrated in the upper bands of a relative distribution, not because climate progress has occurred or reversed in any absolute sense. Second, where year-on-year progress assessment is required, supplement the alignment share with views that look at the underlying climate indicators directly, country by country, so that real-world movement in the holdings can be seen even when the categorical layer is silent. Third, where alignment targets are publicly committed, make explicit (internally at minimum, and ideally in disclosures) that the share is constructed from comparative categories, and consider how the target architecture itself should accommodate the relative nature of the inputs.

Closing perspective

The point of this note is not that NZIF-alignment shares are unsuited to sovereign portfolios, and not that the CCPI's rating system is misconceived. Both frameworks are coherent in what they were designed to do. The friction lies at the interface between them: a comparative rating system feeding a KPI that is widely read as an absolute progress measure. Recognizing that the alignment share is, in this configuration, a positioning signal rather than a direct reading of climate progress recovers most of the interpretive clarity that the categorical form can otherwise obscure.

A related question arises in the numerical version of the same use case, where the CCPI's overall score is used directly for portfolio-level year-on-year tracking; that question is treated in a separate note.