Capital requirements: going concern and gone concern

The architecture of 2012

The capital regime adopted in 2012 translated the commission’s three layers into law. For UBS and Credit Suisse, the total requirement of about 19 per cent of risk-weighted assets consisted of the basic requirement, the buffer and the progressive component. The first two layers were designed to keep the bank alive as a going concern. The progressive component, consisting of low-trigger convertible instruments, was intended to provide fresh equity when the bank was already in serious difficulty and thus to support the emergency plan. The Federal Council’s report of 2015 described the original framework as a combination of minimum requirements, buffers including convertible instruments with high and low triggers, and a progressive component. That component was designed to create incentives to reduce size while providing loss-absorbing capacity.¹

The leverage ratio, which relates capital to total unweighted exposure, was a novelty. The Basel Committee introduced it as an international backstop only in the Basel III framework, where it became a binding Pillar 1 requirement several years later. The Swiss legislator had rejected an explicit leverage ratio in the statute but allowed the ordinance to set one. The ordinance fixed it as a proportion of the risk-weighted requirement, producing the figure of 4.56 per cent for the two big banks mentioned above. The design reflected a compromise. By linking the leverage ratio to the risk-weighted requirement, the ordinance avoided a second, independent calibration and kept the leverage ratio in the background as long as risk weights remained plausible.

Simple to explain, the architecture of 2012 placed Switzerland ahead of most other jurisdictions in the period from 2012 to 2015. It also had weaknesses that became visible quickly. The first concerned risk weights. Because the requirements were expressed mainly as percentages of risk-weighted assets, their real bite depended on the internal models that the banks used to calculate those weights. When the big banks reduced their risk-weighted assets through model changes and a shift into business with low risk weights, their capital ratios rose, while the leverage ratio showed a much less comfortable picture. In December 2014 Jean-Pierre Danthine, vice-chairman of the SNB’s Governing Board, observed that the risk-weighted capital ratios of the Swiss big banks exceeded the global average while their leverage ratios did not. He added that the estimated loss potential for the big banks remained substantial relative to their capitalisation and that a lasting solution for the “too big to fail” issue was imperative.²

The second weakness concerned the gone-concern capacity. The low-trigger instruments of the progressive component were meant to recapitalise the bank at the point of non-viability. They were few, however, and their amount depended on the bank’s size and market share. International thinking had in the meantime moved to a broader concept. The Financial Stability Board was developing its standard on total loss-absorbing capacity, TLAC, which required global systemically important banks to hold a minimum amount of instruments that could be written down or converted in resolution. The Swiss regime of 2012 did not yet contain a comprehensive requirement of this kind.

The report of 2015 and the revision of 2016

Article 52 of the Banking Act, inserted in 2011, obliges the Federal Council to evaluate the TBTF provisions and to report to Parliament. Since 2015 reports have followed at roughly two-year intervals.³ The first report was prepared with the help of a group of experts chaired by the economist Aymo Brunetti, whose recommendations were already referred to by Danthine in December 2014.⁴ The Federal Council adopted its report on 18 February 2015. The overall assessment was favourable. Having judged the Swiss approach positively in an international comparison, the Federal Council concluded that no fundamental reorientation of the regulatory model was necessary. Yet the report also noted that the capital ratios were on track for the targets of 2019, whereas emergency planning and the improvement of resolvability in general had not yet been implemented.⁵

The report was frank about the persistence of the implicit guarantee. It referred to estimates by the International Monetary Fund and to data from rating agencies which indicated that implicit state guarantees for the Swiss big banks still existed, though they were lower than in 2009. It described such a guarantee as a hidden state subsidy to the systemically important banks and cited pre-crisis estimates by Andrew Haldane of the Bank of England. The Federal Council formulated nine recommendations. They included a review of the methods used to calculate risk-weighted assets, a recalibration of the capital requirements with a higher leverage ratio, stricter quality requirements for convertible capital and a modification of the relief granted at the level of individual entities. The Federal Council also recommended a public deadline for the implementation of the emergency plans, a binding TLAC requirement in line with the standard of the Financial Stability Board, a stronger legal framework for resolution, a better tax treatment of bail-in bonds and a biennial review of effectiveness.⁶

On 21 October 2015 the Federal Council announced the key parameters of the revision. For the two big banks, the going-concern requirement was set at a leverage ratio of 5 per cent and a risk-weighted ratio of 14.3 per cent. Of this, 3.5 per cent of the leverage exposure and 10 per cent of risk-weighted assets had to be met with common equity tier 1, while the rest could be covered with high-trigger additional tier 1 instruments converting at a common equity ratio of 7 per cent. A gone-concern requirement of equal size was added, to be met with bail-in instruments. Taken together, the big banks were to hold total loss-absorbing capacity of 10 per cent of the leverage exposure, or 28.6 per cent of risk-weighted assets. The requirements were to be phased in linearly until the end of 2019, the date by which the emergency plans also had to be executable. FINMA’s director Mark Branson said that the measures would strengthen the resilience of the big banks, the stability of the Swiss financial system and thus the entire economy.⁷

The banks resisted parts of the package. In February 2016 the Swiss Bankers Association criticised the proposed thresholds and progression rules, arguing that the differences at stake could alter requirements by about CHF 5 billion and create annual costs of up to CHF 500 million. The Federal Council made concessions. It replaced a “double progression” with a simple progression and allowed convertible bonds governed by foreign law if their conversion remained assured in a crisis. It rejected, however, the banks’ request to calculate the requirements on the basis of the size of the Swiss subsidiary only. The ordinance amendments, adopted by the Federal Council on 11 May 2016, entered into force on 1 July 2016. For all systemically important banks, the base going-concern requirements were set at 4.5 per cent of the leverage exposure and 12.9 per cent of risk-weighted assets, with the progressive surcharges raising the level for the two big banks to 5 and 14.3 per cent.⁸

The revision of 2016 changed the nature of the regime. The original model had relied on convertible capital to bridge the gap between going concern and gone concern. The new model distinguished clearly between two sets of instruments with different purposes. Going-concern capital served to keep the bank alive, while gone-concern capital, in the form of bail-in bonds issued by the group holding company, was to be converted into equity by FINMA in resolution in order to recapitalise the bank. The logic followed the international TLAC standard, but the Swiss calibration exceeded the FSB minimum. The FSB peer review of 2024 recalled that Credit Suisse had been required to meet a TLAC requirement of 28.6 per cent of risk-weighted assets and a leverage ratio of 10 per cent at group level at the end of 2020, against an international minimum of 18 per cent and 6.75 per cent.⁹

Rebates, relief and the problem of the parent bank

The gone-concern requirement was accompanied by a system of rebates. If a big bank improved its global resolvability beyond the statutory minimum, FINMA could reduce its gone-concern requirement. The maximum rebate was capped. FINMA’s first public resolution report of 2020 explained that the rebate potential was limited to 2 percentage points of the leverage exposure and 5.7 percentage points of risk-weighted assets. From 2022, the gone-concern requirement could not fall below 3.75 per cent of the leverage exposure and 10 per cent of risk-weighted assets. For 2019, Credit Suisse received 40 per cent and UBS 42.5 per cent of the maximum possible rebate.¹⁰ The system was meant to reward the banks for structural changes that made resolution more credible, such as the creation of separate Swiss subsidiaries, the establishment of service companies and the reduction of intra-group dependencies.

The rebate system reached its limits within a few years. By the early 2020s the big banks had obtained most of the rebates available, which exhausted the incentive effect. The Federal Council’s report of 2021 announced its intention to abolish the system and noted that the legal anchoring of the incentive mechanism still required action.¹¹ The revision of the Banking Ordinance adopted in November 2022 replaced the rebates with a mechanism under which FINMA may increase the requirements for additional loss-absorbing funds or liquidity if obstacles to resolvability persist. The explanatory report gave as the reason that the incentive effect of the discount system had been exhausted because the banks concerned had reached the maximum reductions.¹² The FSB described the change as a move from rebates to an “add-on” system, in which capital surcharges apply if deficiencies identified in resolvability assessments are not remedied.¹³

A second form of relief proved more consequential. The TBTF requirements applied both at the level of the consolidated group and at the level of individual entities within it, most importantly the parent bank. Credit Suisse AG and UBS AG were at the same time operating banks and the owners of the group’s foreign subsidiaries. When the requirements were first applied at single-entity level, the legislator obliged FINMA to grant relief so that the parent bank would not be treated more strictly than necessary. The Federal Council adjusted the treatment of participations in subsidiaries on 1 January 2019 and clarified on 1 January 2020 that the parent entities of the two G-SIBs remained subject to the special requirements.¹⁴ In its report on the lessons of the Credit Suisse crisis, FINMA later wrote that the obligation to grant relief at single-entity level, introduced by the legislator in 2013, had led to a weakening of the parent bank. In FINMA’s view, the treatment of participations in subsidiaries introduced in 2017 had had a pro-cyclical effect during the crisis.¹⁵

The regulatory filter granted to Credit Suisse AG in the same context became one of the most debated findings of the parliamentary investigation. The Parliamentary Investigation Committee reported in December 2024 that a change in rules in 2017 had allowed FINMA to grant Credit Suisse a regulatory filter worth CHF 15.3 billion, which improved the parent bank’s capital ratios and concealed a weakness that would otherwise have become visible.¹⁶ The filter illustrates a general lesson about capital regimes: requirements set at group level offer limited protection if the entity that actually fails is a subsidiary or a parent bank whose own capital depends on accounting choices. The Federal Council’s proposals of 2024 to 2026 on the capital backing of foreign participations, discussed in chapter G.VI, respond directly to this weakness.

Gone-concern requirements for the domestic banks

The revision of 2016 had introduced gone-concern requirements for the two big banks only. The report that the Federal Council adopted on 28 June 2017 concluded that the regulatory model did not need fundamental adjustment, but proposed extending gone-concern requirements to the domestically oriented systemically important banks.¹⁷ The corresponding ordinance amendment took effect in January 2019. It required Raiffeisen, PostFinance and the Zürcher Kantonalbank to hold gone-concern capital for their possible restructuring and resolution, calibrated at 40 per cent of their going-concern requirements. The lower level compared with the big banks was justified by the lesser international complexity of these institutions. The same revision replaced the deduction of participations in subsidiaries from capital by a risk weighting and brought group companies providing essential services under FINMA oversight.¹⁸

The application of these requirements to banks with public ownership raised a delicate question. Could an explicit guarantee by a canton, or the ownership of the Confederation, count as loss-absorbing capacity? The supervisor’s answer, as it emerges from the resolution reports, was cautious: public support could not replace the obligation of the banks to build up genuine gone-concern capacity over time. FINMA found in 2022 that none of the three domestic banks had reserved enough capital to be recapitalised and continue operations in a crisis. It expected full compliance with the Capital Adequacy Ordinance by 2026.¹⁹

Notes

1. Federal Council, Bericht des Bundesrates «Too big to fail», 18 February 2015, https://www.parlament.ch/centers/eparl/curia/2014/20143002/Bericht%20BR%20D.pdf, p. 11. ↩
2. Jean-Pierre Danthine, “Swiss banks from a financial stability perspective”, introductory remarks, media news conference, Berne, 1 December 2014, BIS central bankers’ speeches, https://bis.org/review/r141215c.htm. ↩
3. Banking Act, Art. 52, as amended by the Federal Act of 30 September 2011 (strengthening the stability of the financial sector), in force since 1 March 2012, https://www.droit-bilingue.ch/de-fr/9/95/952.0-52-107.html. The provision requires the Federal Council to review the provisions no later than three years after the entry into force of the fifth and sixth sections of that amendment and thereafter every two years. The Federal Council’s report under Art. 52 BankG, https://www.newsd.admin.ch/newsd/message/attachments/66982.pdf, p. 3, confirms the two-year rhythm. ↩
4. Danthine, “Swiss banks from a financial stability perspective” (as above), which refers to the report of the Brunetti expert group on the evaluation of the TBTF framework. ↩
5. Federal Council, Bericht «Too big to fail», 18 February 2015 (as above), pp. 6–7. ↩
6. Federal Council, Bericht «Too big to fail», 18 February 2015 (as above), p. 8 for the implicit guarantee and pp. 14–15 for a summary of the nine recommendations. ↩
7. FINMA, press release “Swiss ‘too big to fail’ regime significantly strengthened”, 21 October 2015, https://www.finma.ch/en/news/2015/10/mm-tbtf-20151021/; see also FINMA, fact sheet “Die Schweiz erlässt international führende Kapitalvorschriften”, https://www.finma.ch/de/~/media/finma/dokumente/dokumentencenter/myfinma/faktenblaetter/faktenblatt-to-big-to-fail-regime-verstaerkt.pdf. ↩
8. Année politique Suisse, “‘Too-big-to-fail’-Bericht und daraufhin beschlossene Massnahmen (2015)”, https://politchronik.swiss/de/prozesse/56308-too-big-to-fail-bericht-und-daraufhin-beschlossene-massnahmen-2015. ↩
9. Financial Stability Board, Peer Review of Switzerland: Review Report, 29 February 2024, https://www.fsb.org/uploads/P290224.pdf, p. 9. ↩
10. FINMA, Resolution Report 2020, February 2020, https://www.finma.ch/en/~/media/finma/dokumente/dokumentencenter/myfinma/finma-publikationen/resolution-bericht/20200225-resolution-bericht-2020.pdf?la=en, pp. 10 and 27. ↩
11. Federal Council, Bericht des Bundesrates zu den systemrelevanten Banken (evaluation under Art. 52 BankG), https://www.newsd.admin.ch/newsd/message/attachments/66982.pdf, pp. 14–15. ↩
12. Federal Department of Finance, explanatory report on the amendment of the Banking Ordinance (insolvency, deposit protection, segregation and resolvability), 23 November 2022, https://www.newsd.admin.ch/newsd/message/attachments/73991.pdf, p. 9. ↩
13. FSB, Peer Review of Switzerland (as above), p. 10. ↩
14. Federal Council, Bericht zu den systemrelevanten Banken (as above), pp. 2–3. See also Federal Council, press release “Federal Council adopts amendment of Capital Adequacy Ordinance”, 27 November 2019, https://www.admin.ch/gov/en/start/documentation/media-releases.msg-id-77254.html, which states that the amendment was to ensure that the parent entities of systemically important banks are sufficiently well capitalised in the event of a crisis. ↩
15. FINMA, Lessons Learned from the CS Crisis, 19 December 2023, https://www.finma.ch/en/~/media/finma/dokumente/dokumentencenter/myfinma/finma-publikationen/cs-bericht/20231219-finma-bericht-cs.pdf, p. 8. ↩
16. Blick, “Die zehn wichtigsten Erkenntnisse der PUK zum Ende der Credit Suisse”, 20 December 2024, https://www.blick.ch/politik/puk-untersuchungsbericht-enthuellt-das-sind-die-10-wichtigsten-erkenntnisse-zum-cs-niedergang-id20430902.html. The full report of the Parliamentary Investigation Committee is published on the website of the Federal Assembly. ↩
17. Federal Council, press release “Der Bundesrat verabschiedet den Bericht über systemrelevante Banken”, 28 June 2017, https://www.admin.ch/gov/de/start/dokumentation/medienmitteilungen/bundesrat.msg-id-67336.html. ↩
18. NZZ, “Inlandbanken: Bundesrat verschärft die Kapitalvorschriften”, https://www.nzz.ch/wirtschaft/inlandbanken-bundesrat-verschaerft-die-kapitalvorschrifen-ld.1438321. ↩
19. FINMA, Resolution Report 2022, “Focus on domestic systemically important banks” (as above). ↩

For general information only. This article does not constitute legal or tax advice, does not replace advice on an individual case and creates no lawyer-client relationship. It reflects the legal position as of September 2025.