2015年-FSB全球金融稳定委员会_Summary_of_Findings_from_the_TLAC_Impact_Assessment_Studies_30页_3mb

Summary of Findings from the TLAC Impact Assessment Studies

Executive Summary

The Financial Stability Board (FSB) conducted a comprehensive impact assessment in 2015 to inform the final calibration of the minimum total loss-absorbing capacity (TLAC) requirement for global systemically important banks (G-SIBs). The TLAC requirement was proposed to be in the range of 16–20% of risk-weighted assets (RWA) or 6–10% of the exposure measure (EM). The study included a Quantitative Impact Study (QIS), market survey, economic impact assessment, and analysis of historical losses and recapitalisation needs.

The findings indicate that the impact of TLAC on economy-wide funding costs is relatively contained. The median estimated increase in G-SIB lending rates for the 16/6 calibration is 5.4bps, and for the 20/6 calibration is 9.2bps. These translate into an increase in lending rates for the average borrower of 2.2bps and 3.7bps, respectively. The macroeconomic costs of TLAC are also limited, with a median long-run GDP loss of less than 2bps for the 16/6 calibration and less than 10bps under more conservative assumptions.

The macroeconomic benefits of TLAC are estimated to exceed these costs. The central estimate suggests that TLAC generates benefits of 48bps of annual GDP, and even under the most conservative assumptions, the benefits are 15–20bps. These benefits arise from the reduced likelihood and cost of financial crises, as well as the disciplinary effect of bail-inable debt on G-SIBs' risk-taking behavior.

Key Findings

QIS Shortfall Analysis

• External TLAC:

  • The average eligible external TLAC ratio for G-SIBs as of end-2014 is 13.1% RWA and 7.2% EM (Case 1).
  • Including unsubordinated liabilities that are otherwise eligible, the ratios increase to 16.5% RWA and 8.7% EM (Case 3).
  • Shortfalls for the 16/6 calibration range from €307bn to €790bn, and for the 18/6.75 calibration from €457bn to €1,130bn.
  • Excluding emerging market G-SIBs, shortfalls are significantly lower, ranging from €42bn to €520bn for 16/6, and €107bn to €776bn for 18/6.75.

• Internal TLAC:

  • The average internal TLAC risk-based ratio is 17.5% RWA (Case 1), increasing to 27.2% RWA when including unsubordinated liabilities (Case 3).
  • Internal TLAC shortfalls are generally smaller than external ones, with the largest shortfalls being €19bn for the 90% of 2*3% LR requirement.
  • No data is available for material sub-groups, as the QIS was conducted before the final definition of material sub-groups was established.

• Maturity Analysis:

  • Nine G-SIBs could meet the 16/6 requirement if they replaced maturing liabilities with eligible TLAC instruments.
  • Most of the 17 G-SIBs with no shortfall under Case 4 could meet the 16/6 requirement by replacing maturing liabilities with eligible TLAC.
  • Five G-SIBs, including emerging market ones, still have shortfalls even after replacing all maturing liabilities under Case 4.

• Holdings of TLAC within the system:

  • G-SIBs and non-G-SIBs hold relatively small amounts of TLAC instruments.
  • In the case of the widest definition (Case 4), G-SIB cross-holdings range from €0.2bn to €31bn, and non-G-SIB holdings range from €0 to €40bn.

Market Capacity Findings

• Market size:

  • The total outstanding G-SIB debt securities as of end-2014 are €4,527bn.
  • The Case 1 shortfalls represent 1.0% to 1.7% of the €80 trillion global debt securities market.
  • When including unsubordinated liabilities, the percentages drop to 0.7% to 1.3%.

• Impact on spreads:

  • Market participants estimated that new TLAC issuance would cause bond spreads to rise by 30bps.
  • This estimate is subject to uncertainty.

Economic Impact Assessment

• Costs of meeting shortfalls:

  • At prevailing funding spreads, the median annual funding cost increase for the 16/6 calibration is €195 million, and for the 20/6 calibration is €511 million.
  • The aggregate costs do not change substantially under robustness calculations, but the impact is concentrated on G-SIBs with large shortfalls and those operating in thin market segments.

• Impact on economy-wide funding costs:

  • The median increase in G-SIB lending rates is 5.4bps for the 16/6 calibration and 9.2bps for the 20/6 calibration.
  • These rates translate to an increase in lending rates for the average borrower of 2.2bps and 3.7bps, respectively.

• Macroeconomic costs:

  • The median long-run GDP loss is estimated at less than 2bps for the 16/6 calibration and 2.8bps for the 18/6.75 calibration.
  • Under the most conservative assumptions, the GDP loss could be as high as 10bps for 16/6 and 20bps for 20/6.

• Spillovers:

  • For G-SIB home countries, the estimated median peak annual output loss is 3–4bps, with 1bp attributed to spillovers.
  • For host and other countries, the estimated median peak output loss is 2–3bps.

• Macroeconomic benefits:

  • The introduction of TLAC is estimated to reduce the GDP cost of crises by 5.4 percentage points (ppt).
  • This is due to lower fiscal costs (no bail-outs) and lower sovereign yields in a crisis.
  • The disciplinary effect of TLAC on G-SIBs' risk-taking is estimated to reduce the likelihood of failure by at least one-third.

• Cost and benefit analysis:

  • The analysis provides a fair but conservative estimate of the net benefits.
  • It may underestimate the net benefits due to its conservative bias towards higher costs and lower benefits.

Additional Findings from the Market Survey

• G-SIB TLAC issuance strategy:

  • Most G-SIBs do not plan to replace existing ineligible TLAC instruments before maturity.
  • Subordination is the most frequently cited challenge for meeting TLAC requirements.

• Management buffer:

  • Market participants expect G-SIBs to hold 1.8% RWA above the minimum TLAC requirement as a management buffer.

• Credit rating implications:

  • Rating agencies expect rating downgrades for G-SIB debt as resolution regimes become operational.
  • However, they expect rating uplifts for senior unsecured debt once TLAC requirements are in place.

• Emerging market issuance:

  • Few comments were made about issuing in emerging market economies.
  • Some MPE subsidiaries expressed concerns about the limited local bond markets in emerging economies.

Historical Losses and Recapitalisation Needs

• Total losses:
  • The interquartile range of total losses for the banks studied is 3%–7% RWA, with a maximum of 12%–13% RWA.
  • When including recapitalisation needs, the interquartile range increases to 6%–15% RWA, with a maximum of 25% RWA.

Terminology

• AT1: Additional Tier 1 instruments
• Calibration: Four calibrations used in the economic impact analysis:
  • Calibration 1: 16% RWA or 6% EM ("16/6")
  • Calibration 2: 20% RWA or 6% EM ("20/6")
  • Calibration 3: 16% RWA or 10% EM ("16/10")
  • Calibration 4: 20% RWA or 10% EM ("20/10")
• EM: Exposure Measure, referring to the Basel III leverage ratio exposure measure
• Emerging market G-SIBs: G-SIBs headquartered in emerging market economies
• Market: Refers to the capacity and size of the markets for TLAC-eligible instruments
• G-SIBs: Global systemically important banks
• RWA: Risk-weighted assets
• T2: Tier 2 instruments
• Term Sheet: The term sheet from the November 2014 TLAC consultation document

1. Introduction

The FSB, in consultation with the Basel Committee on Banking Supervision (BCBS), published a consultative document on TLAC in November 2014. This document included high-level principles and a detailed Term Sheet. To finalize the TLAC standard, the FSB conducted a comprehensive impact assessment in 2015, involving the QIS, economic impact assessment, market survey, and analysis of historical losses and recapitalisation needs. The findings of these studies were used to inform the calibration of the minimum TLAC requirement and ensure that it achieves the objective of resolving G-SIBs without exposing public funds to loss.

2. QIS Shortfall Analysis

• The QIS assessed the impact of TLAC requirements on G-SIBs, including shortfalls for each resolution entity and material subsidiaries.
• Data from 29 G-SIBs was used, with the sample size reduced to 29 due to insufficient data for some banks.
• Four Cases were tested to reflect different levels of TLAC eligibility and substitution:
  • Case 1: Instruments meeting the Term Sheet criteria, excluding certain exemptions.
  • Case 2: Additional criteria required for Tier 2 instruments.
  • Case 3: Excludes subordination requirements.
  • Case 4: Broadest case, including all unsecured liabilities except those from derivatives, deposits, and non-contractual obligations.

3. Market Capacity Findings

• The TLAC shortfalls are a small fraction of the global debt securities market.
• The Case 1 shortfalls represent 1.0%–1.7% of the €80 trillion market.
• The Case 3 shortfalls are 0.7%–1.3% of the same market.
• The market survey highlighted the uncertainty in the impact on spreads and the market absorption capacity as a critical factor in the feasibility of TLAC implementation.

4. Estimates of Costs of Meeting Shortfalls

• The costs to G-SIBs of meeting the TLAC shortfall depend on the calibration and market conditions.
• The median increase in G-SIB lending rates is 5.4bps for the 16/6 calibration and 9.2bps for the 20/6 calibration.
• These rates translate into 2.2bps and 3.7bps increases for the average borrower.
• The costs are concentrated on G-SIBs with large shortfalls and those in thin markets.

5. Macroeconomic Impact

• The macroeconomic costs of TLAC are relatively low, with a median long-run GDP loss of less than 2bps for the 16/6 calibration.
• Under more conservative assumptions, the GDP loss could be as high as 10bps for 16/6 and 20bps for 20/6.
• The macroeconomic benefits of TLAC are estimated to be 48bps of annual GDP.
• Even under the most conservative assumptions, the benefits exceed the costs by 15–20bps.

6. Historical Losses and Recapitalisation Needs

• The interquartile range of total losses for the banks studied is 3%–7% RWA.
• When including recapitalisation needs, the range increases to 6%–15% RWA, with a maximum of 25% RWA.
• These findings support the need for a robust TLAC framework to ensure the resolution of G-SIBs without reliance on public funds.