EN
Back to the archive

The encyclopedia · Finance & Accounting · Financial decision · 2015–2019

Fed priced systemic risk: banks with bigger footprints hold more capital

The 2015 GSIB surcharge maps each big bank's systemic score to a capital charge of 1.0–4.5%, making failure costlier for the firm that causes it.

Federal Reserve Board · JPMorgan Chase · Bank of America · Citigroup

the move

In July 2015 the Federal Reserve adopted a final rule requiring the eight largest, most systemically important US bank holding companies to hold extra capital. The surcharge is calibrated to each firm's overall systemic risk, estimated at 1.0% to 4.5% of risk-weighted assets.

Firms compute the surcharge under two methods and hold the higher: one based on the Basel framework (size, interconnectedness, cross-jurisdictional activity, substitutability, complexity), the other replacing substitutability with reliance on short-term wholesale funding.

Chair Janet Yellen said the rule confronts firms with a choice: hold substantially more capital or shrink their systemic footprint. The surcharge began phasing in on January 1, 2016 and became fully effective on January 1, 2019.

why it works

  • A formula score makes the charge predictable and tied to measurable inputs.
  • Interconnectedness and wholesale funding capture how failure propagates.
  • Taking the higher of two methods guards against gaming a single metric.
  • The price is paid in capital, which both buffers failure and raises its private cost.
the payoffCapital charge matches systemic footprintclever

what transfers

Price an externality by scoring the firm's contribution to it with a transparent formula: the firm then chooses between holding more buffer and shrinking its risky footprint.

what came after

The surcharge became part of US bank capital rules and was mirrored internationally under the Basel framework. Banks responded by trimming some systemic indicators, and the methodology continues to be recalibrated with each annual GSIB identification.

references

spotted an error? The archive wants to know.

same kind of clever