Sovereign Exposures in Transition Plans: Measured, Not Targeted

Context: how the pattern emerged

Over the past three to four years, several institutional climate frameworks have moved from voluntary aspiration toward something resembling expected practice. The Partnership for Carbon Accounting Financials (PCAF) Standard has become the default approach to measuring financed emissions across asset classes, including sovereign debt. The Net Zero Investment Framework (NZIF), updated to version 2.0 in 2024, now provides explicit guidance for the alignment classification of sovereign holdings. Transition plan reporting, codified through the UK Transition Plan Taskforce, the European Sustainability Reporting Standards, and the IFRS S2 climate disclosures, has moved from a small set of leading institutions to a much broader expectation.

For sovereign fixed-income teams in large universal banks, asset managers and insurers, this convergence has produced a specific decision point. Sovereign bonds are typically held for liquidity management, regulatory liquidity coverage, asset-liability matching, or general portfolio diversification. They are not normally held with the institutional intention of influencing the climate trajectory of the issuing country. At the same time, sovereign exposures are often material in size, and the methodological infrastructure now exists to measure their associated emissions, classify their alignment with net-zero pathways, and locate them within transition plan reporting.

The question institutions have had to answer is not whether sovereign exposures can be integrated into transition plans, but whether they should be. A specific institutional answer has emerged, and it is the subject of this note.

What the pattern looks like in practice

The pattern can be described in five connected elements.

First, sovereign emissions are measured. PCAF's sovereign methodology is used to attribute production emissions of issuing countries to the holding institution, typically on a weighted basis. Estimates are produced, disclosed, and updated annually. The measurement work happens.

Second, sovereign alignment is classified. Where the institution uses NZIF, sovereign holdings are assigned to alignment categories, typically a hierarchy ranging from "not aligning" to "achieving net zero," with intermediate categories such as "committed to aligning," "aligning," and "aligned." The share of sovereign holdings in the more advanced categories enters the institution's reported portfolio alignment score.

Third, sovereign exposures are then explicitly carved out of the institution's net-zero targets and transition plan trajectories. Where corporate financed emissions are subject to interim decarbonisation targets, sovereign financed emissions are typically reported separately, without an associated target. Where corporate alignment is integrated into the trajectory toward 2030 and 2050 commitments, sovereign alignment - despite being scored - sits outside the target architecture.

Fourth, the justification is articulated. The standard institutional language refers to the limited influence the institution has over national emissions, the primary holding purpose of sovereign debt (liquidity, regulatory requirements, asset-liability matching), and the inappropriateness of treating sovereign issuers in the same target framework as corporate issuers, over which engagement and capital reallocation have direct effects.

Fifth, an alternative posture is sometimes positioned in place of targets, transparency in disclosure, dialogue with sovereign issuers and other investors, contributions to sector-level discussion forums.

The institutional logic: where it holds

The institutional logic behind this pattern is not difficult to articulate, and it is not weak.

Sovereign issuers are categorically different from corporate issuers in their relationship to the institutions that hold their debt. A large investor in a corporate bond can, in principle, engage management on transition strategy, file resolutions, divest, or reallocate within an industry to favor better-positioned issuers. The set of available instruments is well-defined, and the link between capital allocation and corporate behavior, while contested in detail, is structurally direct.

Sovereign issuers have no equivalent governance interface. National climate policy is set by elected governments accountable to electorates, not to bondholders. Engagement with finance ministries and debt management offices, where it exists, occurs through a fundamentally different channel. Divestment from a sovereign issuer's bonds rarely has a discernible effect on its borrowing costs in normal market conditions, and even where it does, the policy transmission is indirect. For state-owned investors -public pension funds, sovereign wealth funds, central bank reserve managers - the engagement channel narrows further, since direct engagement with government on broad climate policy can sit uncomfortably with the investor's own public-sector status.

The holding purpose is also different. Sovereign government bonds are typically held for high-quality liquid asset requirements under banking liquidity regulation, for matching long-dated liabilities, for portfolio diversification against credit risk, and for general liquidity management. These purposes are largely insulated from climate considerations in the way the holdings of an actively managed equity portfolio are not.

PCAF, the framework that drives the measurement, is explicit that its standard does not prescribe target-setting; it provides a methodology for attribution. NZIF, in its sovereign-specific guidance, similarly stops short of requiring institutions to set sovereign-specific reduction targets. The frameworks themselves leave room for the carve-out.

Treated together, these considerations make the institutional logic coherent on its own terms. It is not a methodological evasion; it is a position with its own intellectual structure.

The institutional logic: where it breaks

That logic, however, does not eliminate the structural tensions the pattern creates.

The first tension is internal coherence. An institution that measures sovereign financed emissions is implicitly affirming that those emissions are materially attributable to its activities - that is the point of PCAF attribution. An institution that classifies its sovereign holdings under NZIF alignment categories is affirming that their alignment is part of its portfolio's climate quality. Yet the same institution, when constructing its forward-looking target architecture, separates sovereign exposures out and assigns them no trajectory. The same holdings are included in two backward-looking and present-state measures and excluded from the forward-looking commitment those measures would normally support.

The second tension is reporting interpretation. Where an institution publishes a portfolio-level NZIF alignment score that aggregates corporate and sovereign holdings, the resulting number invites a particular reading: that the portfolio as a whole is on a path consistent with net zero to the extent indicated by the score. If sovereign exposures contribute to that score but are not anchored in the target architecture, the headline number conveys a stability of expectation that the underlying construction does not actually support. Movement in the sovereign portion of the score is decoupled from anything the institution intends to act on.

The third tension is selectivity. The limited-influence argument is defensible on its own merits. Its institutional reach is sometimes uneven, however. Materiality and influence are connected concepts, and a consistent framework would either acknowledge influence limits across all relevant asset classes or articulate why sovereign exposures are uniquely positioned. In practice, this articulation is rare.

The fourth tension is materiality. If sovereign exposures are large enough to require PCAF measurement and NZIF classification, the institutional case that they are too marginal to warrant target-setting becomes harder to sustain on materiality grounds alone.

The interface problem: PCAF, NZIF, and transition plans

At the methodological level, the difficulty of integrating sovereign exposures into transition plans reflects how the underlying frameworks fit together when applied to sovereigns specifically.

PCAF provides the financed-emissions accounting standard. For sovereigns, it attributes production emissions of the issuing country to the institution holding the bonds. NZIF integrates PCAF as the accounting basis for its emissions-based sovereign objective, while adding its own alignment-classification system as a separate objective, categorical labels indicating where issuers sit relative to net-zero pathways. Transition plan frameworks then sit above both, requiring institutions to articulate forward trajectories and intermediate targets, typically expressed in absolute or intensity-based emission terms.

For corporate issuers, the three layers articulate cleanly: PCAF measures present emissions, NZIF assesses present alignment, the transition plan commits to forward reduction. The corporate transition planning toolkit - engagement, voting, allocation tilting, divestment - provides instruments that link the three layers operationally.

For sovereign issuers, the layers are present but the instruments are not. The institution can measure sovereign financed emissions through PCAF and classify sovereign alignment through NZIF, but it lacks the operational toolkit to commit to forward trajectories that it controls. Institutional investors sometimes make a related point when the question is put to them directly: that even if formal sovereign targets were in place, the concrete portfolio actions those targets would imply are not always obvious. This is the structural source of the "measured but not targeted" pattern: the methodological frameworks for measurement and classification have outpaced the operational frameworks for sovereign-specific action.

The carve-out is, in this sense, an honest reflection of the operational asymmetry. It is also, however, an unresolved methodological gap that institutional sovereign frameworks have not yet fully addressed.

Methodological options

Institutions that wish to reduce the asymmetry - without abandoning the institutional logic that gives rise to it - have several methodological options available. None resolves the structural issue completely. Each addresses it in a specific way, with specific trade-offs.

A first option is separate sovereign sub-targets within the transition plan, distinct from the corporate trajectory but explicit about the commitment. These could be expressed as alignment-share trajectories for the sovereign portion of the portfolio, as indicators drawn from sovereign climate frameworks, or as the climate performance of underlying issuing countries weighted by holdings. The trade-off is methodological: alignment-share targets inherit the relative-scoring questions that come with NZIF-categorization, and indicator-based targets require choices about which indicators are tracked and on what basis.

A second option is an engagement-based framework for sovereign holdings that specifies what engagement consists of in this context, what its operational instruments are, and how its results are tracked. The challenge is that sovereign engagement has no widely accepted methodology, and tracking its results requires defining indicators that are themselves contested.

A third option is integrating sovereign climate performance into allocation logic. Some portion of the sovereign portfolio is held for liquidity and regulatory purposes, but climate considerations can influence the remaining allocation decisions. This separates the regulatory and discretionary portions and allows the latter to participate in the transition plan more like the corporate side. The trade-off is the operational complexity of drawing the distinction consistently across reporting cycles.

A fourth option, often overlooked, is explicit articulation of the carve-out itself, disclosing not only that sovereign exposures are excluded from targets, but the institutional logic and the boundary conditions under which the exclusion holds. This does not change the substantive position; it changes its visibility and interpretability, which addresses the reporting-credibility tension without requiring methodological reconstruction.

Closing perspective

The pattern described here is not a methodological failure on the part of the institutions that have adopted it. It reflects a coherent set of considerations about the difference between sovereign and corporate exposures, about the holding purposes that drive sovereign portfolio construction, and about the operational instruments available to institutional investors. The position has its own intellectual structure and deserves to be treated as a choice rather than as an oversight.

The choice has consequences. It creates internal asymmetries between measurement and target-setting, it affects the interpretability of headline alignment scores, and it raises questions about consistency across asset classes. These consequences do not invalidate the choice, but they do reward being made explicit. The institutional credibility of a transition plan depends not on whether sovereigns are included or excluded, but on whether the rationale for the chosen position is articulated clearly enough that external readers can understand what the plan commits to and what it does not.

The methodological options available to institutions that want to revisit the position exist and are workable. None of them is the obvious answer. Choosing among them - or maintaining the existing carve-out - is, in the end, an institutional decision rather than a methodological one. What the methodological frameworks can do is make the decision visible and its consequences legible. The rest belongs to the institutions themselves.