Navigating G-SII designation: a practical guide for global banking institutions

The assessment is indicator-based, and it is relative. The Basel Committee methodology scores banks in a global sample across a set of indicators, aggregates those scores, and identifies the institutions above a cut-off as G-SIBs, sorting them into buckets that set how much extra capital they hold. The UK G-SII framework mirrors this structure. The PRA methodology weighs five broad categories equally, and an institution’s score in each is measured against the global sample rather than against a fixed threshold.

The five categories are size, interconnectedness, substitutability, complexity and cross-jurisdictional activity. Size is measured through total exposures on the Basel III leverage ratio basis, which captures the institution’s overall footprint. Interconnectedness looks at the financial relationships that could transmit distress through the system, including intra-financial-system assets and liabilities, on the reasoning that a bank wired tightly into other banks poses more systemic risk than a similarly sized but more isolated one. Substitutability measures how hard it would be to replace the institution’s critical functions, such as payment processing, custody and market-making: a bank providing a service no one else can readily provide is systemically important even at modest size. Complexity captures how difficult the institution would be to resolve, driven by large derivatives portfolios and structured products. Cross-jurisdictional activity measures the global footprint and the cross-border contagion risk it creates.

That relative basis changes how an institution has to think. A bank that holds its balance sheet flat while peers grow can drift up the rankings without doing anything. A bank that cuts complexity can find the benefit muted because the whole sample moved with it. So you need to know where you sit in the cohort and where the cohort is heading; both feed your eventual designation and buffer, which makes competitive positioning analysis part of the capital plan rather than a footnote to it.

The headline consequence is the higher loss absorbency requirement, an additional capital buffer that scales with systemic importance. The buffers run from 1.0 percent to 3.5 percent of risk-weighted assets, allocated by bucket, with the higher buckets held for the most systemically important institutions. Three features make the buffer more demanding: It must be met entirely with Common Equity Tier 1 capital, the highest-quality and most expensive form; there is no satisfying it with cheaper instruments. It sits on top of other regulatory buffers and requirements and cannot be double-counted against them, so it is genuinely additional. And it applies at the consolidated group level, capturing the whole institution rather than any single entity. The regime also includes an additional leverage ratio buffer, the ALRB, a parallel constraint expressed against the leverage exposure measure rather than risk-weighted assets, so a G-SII satisfies both a risk-weighted and a leverage-based version of its systemic surcharge.

Beyond the buffer, G-SIB status pulls in the wider FSB requirements that set these institutions apart. They must meet the total loss-absorbing capacity standard, holding a defined minimum of capital and bail-in-able debt so that, in a failure, losses can be absorbed and the institution recapitalised without taxpayer support. They face enhanced resolvability requirements, including group-wide resolution planning and regular resolvability assessments reviewed through the FSB’s process by the supervisors in their crisis management groups. And they attract heightened supervisory expectations across the board. The resolvability and supervisory obligations shape how the group can be structured and run.

The feature that most often catches institutions out is the lag. Roughly two years separate the data an assessment is based on from the point at which the resulting buffer applies. An institution’s indicators are drawn from a prior fiscal year-end, the assessment runs, the designation and bucket are published, and the buffer then phases in on a defined timeline well after the data that produced it.

That lag runs in both directions and has to be planned for. An institution whose systemic footprint grew two years ago is committed to a higher buffer now, whatever it has done since, so capital planning has to look through the lag rather than at the current position alone. Deliberate action to reduce systemic importance does not pay off at once either; the benefit shows up in a lower buffer two years later. An institution that fails to model this lag will misjudge both its capital trajectory and the timing of any strategic move to manage its designation. G-SII capital planning has to run multi-year, mapping the path from today’s actions through the assessment cycle to the eventual buffer.

The methodology depends entirely on the indicators an institution submits, and those indicators are drawn from across a vast, complex balance sheet. An institution that cannot produce them accurately and consistently risks being scored on figures that overstate its systemic footprint, and paying a higher buffer than its true position warrants. An error in an indicator submission can directly raise the capital surcharge, not just create a finding.

Robust data governance for the assessment means clear ownership of each indicator, a documented and repeatable process for generating it, reconciliation against the institution’s other regulatory and financial reporting, and validation before submission. The indicators span size, interconnectedness, substitutability, complexity and cross-jurisdictional activity, so they pull data from treasury, from the derivatives and securities businesses, from payments and custody operations, and from every jurisdiction the group operates in. Producing accurate indicators is a cross-functional capability, and doing it well keeps the designation tied to actual systemic importance rather than data limits. There is a transparency dimension too: the framework requires public disclosure of the indicator data, so the figures are seen not only by supervisors but by analysts and peers, which raises the cost of getting them wrong.

G-SII status is a capital cost and a franchise consideration; a board should weigh both. On the cost side, beyond the CET1 buffer and the total loss-absorbing capacity requirement, designation tends to raise the institution’s overall cost of capital, because it has to issue more equity and more bail-in-able debt, the latter priced to reflect its loss-absorbing role. It draws more intensive supervision, more frequent and demanding stress testing, and a standing obligation to remain resolvable, which limits how the group can be structured and how freely it can reorganise.

On the other side, designation is also an acknowledgement of the institution’s central role in the global financial system, and that role carries franchise value. The largest, most systemically important banks are often the ones clients and counterparties most want to deal with, precisely because of their scale, reach and the implicit robustness the designation reflects. So the task is not to minimise the designation at all costs, because the activities that create systemic importance are frequently the same activities that create the franchise. The task is to read the trade-off clearly: which elements of systemic footprint earn their keep through the business they support, and which are complexity that lifts the surcharge without commensurate return. A board should distinguish value-creating footprint from surplus complexity before cutting.

The five equally weighted categories, what each one measures, and the lever available to an institution that wants to influence its score.

Because the higher loss absorbency requirement is allocated in discrete buckets rather than as a smooth function, the practical stakes of the assessment concentrate at the boundaries between them. An institution sitting comfortably in the middle of a bucket has little to gain from marginal changes to its score. An institution near a boundary faces a step change: a small movement in its relative score can tip it into a higher bucket and add a meaningful slice of CET1, or drop it into a lower one and release capital. The 2025 list shows this plainly. Bank of America and ICBC moved from bucket 2 to bucket 3, and Deutsche Bank moved from bucket 2 to bucket 1, each shift carrying a real change in required capital.

Boards should track where the firm sits relative to the nearest bucket boundary, and which way the cohort is moving it. An institution near the top edge of its bucket should know which indicators are pushing it towards the next one, and whether those indicators reflect business it values or complexity it could rationalise. An institution near the bottom edge should know what is holding it down and whether that is stable. The relative basis makes this harder, because the boundary itself moves as the global sample changes, so the analysis has to track both the institution’s own trajectory and the cohort’s. Monitor the boundary through the year rather than waiting for the annual list.

Take a large banking group sitting near the boundary between two buckets, facing the prospect of moving up and absorbing a higher CET1 surcharge. A proactive approach starts earlier and manages the drivers the methodology measures.

The group builds a robust data governance process first, because the assessment is only as accurate as the indicators submitted, and an institution that submits poorly governed data can find itself scored higher than its true position warrants. It analyses each of the five categories to understand which are driving its score and how it sits relative to the cohort, recognising that the relative basis means peer behaviour matters as much as its own. Where complexity is a driver, it examines whether legal-entity simplification or a reduction in structured-product inventory would move the needle. Where interconnectedness is the driver, it looks at its wholesale funding and intra-financial exposures. It models the two-year lag so that any action is timed against when the benefit would land in a lower buffer. And it engages its supervisors early on material strategic changes, so that a restructuring intended to manage systemic importance is understood by the regulator rather than discovered by it.

None of this is gaming the system. The systemic risk being measured is real, and reducing it makes the institution safer. An institution that understands the methodology can align its capital strategy, structure and supervisory relationships deliberately. Over a multi-year horizon, the difference between the two approaches is measured in basis points of CET1 on a very large balance sheet, which is a material sum.

This insight is provided for general informational purposes only and doesn’t constitute legal, investment, or regulatory advice.