Applying the CCPI Over Multiple Years

Context: from diagnosis to construction

Two earlier notes have described why the CCPI, used at portfolio level over multiple years, can produce movements that are difficult to interpret. The numerical score is built through an annual normalisation that re-anchors the scale each year. The categorical ratings derived from that score inherit the same relativity in a more concealed form. In both cases, portfolio-level movement reflects a mixture of real-world climate change in held countries, change in non-held countries that define the scale, and shifts in the assessed universe itself.

The natural reaction is to ask whether the CCPI should be used at all in this way. In practice, the answer has been that it should (it remains one of the most methodologically transparent, externally produced sovereign climate framework currently available) but that the way it is read over time benefits from some methodological care. Several institutional teams have begun working with complementary views that sit alongside the headline portfolio-level ABC rather than replacing it. Three are particularly useful in sovereign fixed-income workflows.

Complement 1: normalized ranks

The first complement starts from a specific source of volatility in the portfolio-level CCPI score. The score is anchored at the top and bottom of each year's distribution: the best- and worst-performing countries define the scale against which all others are measured. A single strong or weak performer entering, exiting, or markedly changing position can shift the entire range and move every other country's score, even when nothing about those countries has changed.

A normalized rank replaces the score with each country's ordinal position in the global distribution, expressed on a fixed scale between 0 and 1. The strongest performer in a given year sits at one end of the scale, the weakest at the other, and every other country is positioned by its ranking rather than by the gaps in the underlying scores. For portfolio-level use, the weighted-average normalized rank of the holdings can then be tracked over time. Because the scale is anchored to ordinal position rather than to the prevailing score distribution, the resulting series is materially less sensitive to outlier countries and to universe entry or exit.

The trade-off is the loss of magnitude information. A normalized rank cannot distinguish between a country that sits narrowly ahead of the next-best performer and one that sits far ahead. For many reporting purposes this is acceptable - rank stability is itself a useful signal - but the loss should be made explicit, particularly where stakeholders are accustomed to score-based variation.

Complement 2: portfolio-to-universe ratios

The second complement reframes portfolio-level CCPI tracking as a ratio rather than a level. The weighted-average CCPI of the portfolio is compared each year to the unweighted mean CCPI of the assessed global universe, and the result is expressed as a ratio. A value above 1 indicates that the portfolio sits above the global average; a value below 1 indicates the opposite.

The mechanical advantage is that both numerator and denominator are computed on the same annual scale, and are therefore affected in similar ways by the rescaling that creates the underlying problem. If non-held countries improve and pull the scale upward, the portfolio's level falls and so does the universe mean, but the ratio between them remains comparable across years. The series captures whether the portfolio is positioned more or less favorably than the prevailing global picture, year by year, largely insulated from absolute level shifts.

A useful operational property is that the ratio can detect improvement coming from underlying climate progress in held countries, even without reallocation of capital. If sovereign issuers in the portfolio improve faster than the rest of the universe, the ratio rises; if they fall behind, it declines. This makes the ratio a natural fit for institutions that emphasize engagement and stewardship alongside capital allocation.

The principal choice the ratio forces is the definition of the universe in the denominator. A broad assessed universe gives the strongest signal but includes economies that are not investable; an investment-grade or otherwise constrained universe is more market-relevant but narrower.

Complement 3: index-relative views

The third complement extends the ratio logic from a country-mean comparator to an index-based one. Rather than comparing the portfolio's weighted-average CCPI to the unweighted mean of the assessed universe, the comparator becomes the weighted-average CCPI of a relevant global sovereign bond index, for example, a broad global government bond index or a regional equivalent.

The advantage is that the comparator now reflects the actual investable universe rather than a country-equal-weighted average that includes economies of widely differing market presence. The reference becomes one that an institutional investor could, in principle, hold; the comparison sits more naturally inside benchmark-relative reporting and tracking-error-aware portfolio frameworks. For institutions whose climate reporting is increasingly woven into mainstream performance reporting, this alignment matters.

The complement inherits useful properties from the portfolio-to-universe ratio, comparability across years, the ability to reflect underlying improvement in held countries even without reallocation, and adds investability realism. That realism is particularly relevant where alignment claims are made publicly and need to be defensible against the criticism that the comparator includes countries no portfolio could practically hold.

The operational cost is that the reference index must be chosen, documented, and its composition tracked over time; rebalancing and inclusion changes affect the comparator series and need to be reflected in any narrative around the resulting numbers.

Choosing among them

The three complements are not substitutes for each other; they address different facets of the same underlying issue. Most institutional contexts benefit from using two or three in parallel, with each providing a distinct lens on portfolio-level CCPI movement.

Normalized ranks and portfolio-to-universe ratios are natural counterparts. The rank-based view emphasizes stability: it reduces year-on-year noise from outlier movements and universe boundary shifts, and answers the question of whether the portfolio is consistently positioned in the stronger part of the global distribution. The ratio-based view emphasizes responsiveness: it picks up changes in the relative climate trajectory of held countries, including those that occur without reallocation, and expresses portfolio progress in a form that maps naturally to engagement-led narratives. Reporting both allows an institution to communicate the two movements separately and to avoid the conflation the headline portfolio-level CCPI creates.

The index-relative view is most useful where climate reporting needs to sit alongside mainstream performance reporting. Where alignment statements are published or referenced in benchmark-relative contexts, an index-relative comparator is harder to challenge than a country-equal-weighted one. Institutions that operate primarily against investable benchmarks will often find it useful to maintain an index-relative ratio as the principal external reference, with normalized ranks and the broader universe ratio serving as internal supplementary views.

None of the three complements eliminates the relative nature of the underlying CCPI. They make the relativity easier to read, easier to communicate, and easier to disentangle from genuine portfolio-level climate movement. That is the modest but practically important contribution they offer.

Closing perspective

The point of this note is not that institutions should abandon the headline portfolio-level CCPI. It remains a recognised, comparable, externally produced measure with real value in cross-sectional climate assessment. The point is that the time-series use of the metric benefits from being read through complementary lenses that respect its relative construction, and that institutional reporting credibility is well served by making those lenses explicit rather than implicit.

The three complements described here are interpretive: they refine how CCPI movement is read, not what the portfolio actually does. Whether and how portfolios should respond actively to the patterns these lenses reveal - through capital reallocation, weighting adjustments, or climate-aware index construction - is a separate question, treated elsewhere alongside the two diagnostic notes that precede this one.