Skip to content
FinObservatory

Systemic risk

How much each big US bank co-moves with system-wide tail risk

Four academic market-based measures, computed from bank equity returns only: LRMES (the returns-based core of SRISK), MES, Delta-CoVaR, and the system-level absorption ratio. This is not a dollar SRISK capital-shortfall level (that needs consolidated-holding-company market cap and book debt, which this repo does not hold), and it is not investment advice.

36.6%
System-mean LRMES
26 of 28 banks, Jul 6, 2026
28
Banks covered
listed BHCs
0.869
Absorption ratio (63d)
Jul 6, 2026
2020–2026
Data span
Jan 3, 2020 to Jul 6, 2026

Data as of Jul 6, 2026 (system-mean LRMES from bank equity returns)

What this is not. Listed banks only (28 large US bank holding companies; no private banks, credit unions, the thousands of small FDIC banks, or non-bank financials). These are market-derived statistics, not supervisory data. The dollar SRISK level, and therefore the classic market-cap-driven March-2020 SRISK spike, is a documented open gap; what is shown is the returns-based core. See the methodology for every formula, window, citation, and limitation.

Systemic risk

System-mean LRMES barely moved through the COVID crash

Monthly mean across covered US bank holding companies from 2020 through the latest observation. LRMES is each bank’s expected equity loss given a fixed −40% systemic shock.
Hover for values; dashed line = full-sample mean
This returns-based exposure measure is near-flat by design and does not track the current level of market stress or the dollar capital shortfall measured by full SRISK.

Source: FinObservatory systemic engine (absorbed from argus), current revision Month-end mean across covered banks; GARCH(1,1)+DCC(1,1) filter, Brownlees-Engle 2017. Methodology

Top LRMES contributors

The ten banks with the highest latest LRMES, the returns-based systemic-vulnerability core of SRISK. Share is each bank’s LRMES as a fraction of the aggregate across all 28 covered banks. Regionals dominate the top of the list, consistent with the 2023 episode.

#BankLRMESLRMES shareMES
1ZION Zions Bancorporation, N.A.0.49604.9%-0.0464
2WAL Western Alliance Bancorporation0.49524.9%-0.0597
3CFG Citizens Financial Group, Inc.0.44644.4%-0.0314
4RF Regions Financial Corporation0.43314.3%-0.0318
5HBAN Huntington Bancshares Incorporated0.43044.2%-0.0282
6ALLY Ally Financial Inc.0.42834.2%-0.0354
7FITB Fifth Third Bancorp0.41244.0%-0.0298
8FHN First Horizon Corporation0.40934.0%-0.0380
9EWBC East West Bancorp, Inc.0.40904.0%-0.0344
10KEY KeyCorp0.40684.0%-0.0308

Source: FinObservatory systemic engine (absorbed from argus), current revision Each bank at its own most-recent observation; LRMES per Brownlees-Engle 2017, MES per Acharya et al. 2017. Methodology

Delta-CoVaR leaders

Banks whose own distress most worsens the system’s tail Value-at-Risk (most negative first). Delta-CoVaR weights toward the largest, most-connected names, a different lens from LRMES.

#BankDelta-CoVaRAs of

Source: FinObservatory systemic engine (absorbed from argus), current revision Rolling quantile regression, 252-day window, q = 0.05; Adrian-Brunnermeier 2016. Methodology

Systemic risk

Bank co-movement spiked during the COVID crash

Daily 63-day rolling ratio for a fresh panel of listed US bank holding companies since the COVID crash. It measures the share of bank-return variance captured by the top eigenvalues.
Hover for values
High co-movement indicates a tightly coupled system, not an observed loss, and the listed-bank panel excludes private banks, credit unions, and non-bank financial firms.

Source: FinObservatory systemic engine (absorbed from argus), current revision Rolling PCA, top 20% of eigenvalues, 27-bank fresh panel; Kritzman et al. 2011. Shaded: COVID crash and the 2023 SVB / regionals episode. Methodology

All covered banks

Every covered bank at its own most-recent observation. Click a column header to sort; click a row to expand its four measures and definitions. A single dashboard rather than per-bank pages: the universe is small (28 banks) and the listed-BHC tickers do not map cleanly onto the FDIC subsidiary-bank CERTs used by the bank scorecards.

Bank
0.49604.9%-0.0464
0.49524.9%-0.0597
0.44644.4%-0.0314
0.43314.3%-0.0318
0.43044.2%-0.0282
0.42834.2%-0.0354
0.41244.0%-0.0298
0.40934.0%-0.0380
0.40904.0%-0.0344
0.40684.0%-0.0308
0.39753.9%-0.0270
0.39323.9%-0.0256
0.39003.8%-0.0272
0.38783.8%-0.0281
0.37553.7%-0.0263
0.37243.7%-0.0454
0.37163.6%-0.0374
0.35083.4%-0.0426
0.33383.3%-0.0181
0.31013.0%-0.0265
0.29762.9%-0.0234
0.29122.9%-0.0310
0.29052.9%-0.0126
0.27552.7%-0.0148
0.27052.7%-0.0137
0.25952.5%-0.0218
0.25292.5%-0.0196
0.19721.9%-0.0229

Source: Equity returns via Yahoo Finance (argus collector), internal use LRMES/MES/CoVaR/Delta-CoVaR from bank equity returns; internal-use Yahoo-derived input, display aggregates only. Methodology

G-SIB designations

The official supervisory counterpart to the market-based measures above: the Financial Stability Board’s annual list of global systemically important banks and their capital-surcharge buckets, a higher bucket meaning a higher systemic tier. The 2025 list names 29 banks; JP Morgan Chase sits alone in the top occupied bucket (4). These are consolidated holding-company designations, so the names are not linked to the FDIC-subsidiary bank scorecards, which key on insured-bank CERTs, a different legal entity.

FSB 2025 list, by bucket

BucketBanks
41JP Morgan ChaseUSA
34Bank of AmericaUSA, CitigroupUSA, HSBCGBR, Industrial and Commercial Bank of ChinaCHN
29Agricultural Bank of ChinaCHN, BNP ParibasFRA, Bank of ChinaCHN, BarclaysGBR, China Construction BankCHN, Goldman SachsUSA, Groupe Crédit AgricoleFRA, Mitsubishi UFJ FGJPN, UBSCHE
115Bank of CommunicationsCHN, Bank of New York MellonUSA, Deutsche BankDEU, Groupe BPCEFRA, INGNLD, Mizuho FGJPN, Morgan StanleyUSA, Royal Bank of CanadaCAN, SantanderESP, Société GénéraleFRA, Standard CharteredGBR, State StreetUSA, Sumitomo Mitsui FGJPN, Toronto DominionCAN, Wells FargoUSA

Bucket history, 20112025

Each cell is the bank’s FSB bucket that year; a brighter fill is a higher bucket (larger capital surcharge). A blank cell means the bank was not on that year’s list. The 2011 column carries no bucket (the FSB introduced buckets in 2012), so it marks designation only. Current members first, then banks that have since left the list.

Bank111213141516171819202122232425
JP Morgan ChaseUSA·44444444344444
Bank of AmericaUSA·22223322223323
CitigroupUSA·43334333333333
HSBCGBR·44443333333333
Industrial and Commercial Bank of ChinaCHN1112222222223
Agricultural Bank of ChinaCHN111111111222
Bank of ChinaCHN·11111222222222
BarclaysGBR·33332222222222
BNP ParibasFRA·33333222232222
China Construction BankCHN11211221222
Goldman SachsUSA·22222222122222
Groupe Crédit AgricoleFRA·12111111111122
Mitsubishi UFJ FGJPN·22222222222222
UBSCHE·22111111111222
Bank of CommunicationsCHN111
Bank of New York MellonUSA·21111111111111
Deutsche BankDEU·43333332222221
Groupe BPCEFRA·1111111111111
INGNLD·11111111111111
Mizuho FGJPN·11111111111111
Morgan StanleyUSA·22221111111111
Royal Bank of CanadaCAN111111111
SantanderESP·11111111111111
Société GénéraleFRA·11111111111111
Standard CharteredGBR11111111111111
State StreetUSA·11111111111111
Sumitomo Mitsui FGJPN·11111111111111
Toronto DominionCAN1111111
Wells FargoUSA·11112222111111
Credit SuisseCHE(left 2022)·22222111111
UniCreditITA(left 2022)·11111111111
NordeaSWE(left 2017)·111111
Royal Bank of ScotlandGBR(left 2017)·222111
BBVAESP(left 2014)111
CommerzbankDEU(left 2011)·
DexiaBEL(left 2011)·
Lloyds Banking GroupGBR(left 2011)·
Bucket:1234designated, no bucket (2011)

Designation history, 20112025

A bucket is a capital-surcharge tier: higher means more systemic. The 29 null buckets are an explicit state, not zero and not an unlisted bank: they are the 2011 designation-only rows, before the FSB began publishing bucket assignments in 2012.

442
designation rows
20112025
year range
37
distinct banks
29
null buckets
all are 2011 designation-only rows

All 29 null buckets are 2011 designation-only rows: the FSB named G-SIBs that year but did not begin publishing bucket assignments until 2012.

Year-to-year list changes

Entries and exits are computed mechanically from adjacent annual lists. A bank is an exit only when it is present in year N and absent in year N+1, so the table stops at the 2025 list and does not infer a departure after the latest publication.

TransitionBanks on listEntriesExitsBucket changes
2011 to 201229 to 28BBVAESP, Standard CharteredGBRCommerzbankDEU, DexiaBEL, Lloyds Banking GroupGBR
First published buckets: 26 continuing 2011 designations, bucket 1 (12), bucket 2 (8), bucket 3 (2), bucket 4 (4).
2012 to 201328 to 29Industrial and Commercial Bank of ChinaCHNnone
Moved up: Groupe Crédit Agricole 1 to 2FRA
Moved down: Bank of New York Mellon 2 to 1USA, Citigroup 4 to 3USA, Deutsche Bank 4 to 3DEU
2013 to 201429 to 30Agricultural Bank of ChinaCHNnone
Moved down: Groupe Crédit Agricole 2 to 1FRA, UBS 2 to 1CHE
2014 to 201530 to 30China Construction BankCHNBBVAESP
Moved down: Royal Bank of Scotland 2 to 1GBR
2015 to 201630 to 30nonenone
Moved up: Bank of America 2 to 3USA, Citigroup 3 to 4USA, Industrial and Commercial Bank of China 1 to 2CHN, Wells Fargo 1 to 2USA
Moved down: Barclays 3 to 2GBR, HSBC 4 to 3GBR, Morgan Stanley 2 to 1USA
2016 to 201730 to 30Royal Bank of CanadaCANGroupe BPCEFRA
Moved up: Bank of China 1 to 2CHN, China Construction Bank 1 to 2CHN
Moved down: BNP Paribas 3 to 2FRA, Citigroup 4 to 3USA, Credit Suisse 2 to 1CHE
2017 to 201830 to 29Groupe BPCEFRANordeaSWE, Royal Bank of ScotlandGBR
Moved down: Bank of America 3 to 2USA, China Construction Bank 2 to 1CHN
2018 to 201929 to 30Toronto DominionCANnone
Moved down: Deutsche Bank 3 to 2DEU
2019 to 202030 to 30nonenone
Moved up: China Construction Bank 1 to 2CHN
Moved down: Goldman Sachs 2 to 1USA, JP Morgan Chase 4 to 3USA, Wells Fargo 2 to 1USA
2020 to 202130 to 30nonenone
Moved up: BNP Paribas 2 to 3FRA, Goldman Sachs 1 to 2USA, JP Morgan Chase 3 to 4USA
2021 to 202230 to 30nonenone
Moved up: Bank of America 2 to 3USA
Moved down: BNP Paribas 3 to 2FRA, China Construction Bank 2 to 1CHN
2022 to 202330 to 29Bank of CommunicationsCHNCredit SuisseCHE, UniCreditITA
Moved up: Agricultural Bank of China 1 to 2CHN, China Construction Bank 1 to 2CHN, UBS 1 to 2CHE
2023 to 202429 to 29nonenone
Moved up: Groupe Crédit Agricole 1 to 2FRA
Moved down: Bank of America 3 to 2USA
2024 to 202529 to 29nonenone
Moved up: Bank of America 2 to 3USA, Industrial and Commercial Bank of China 2 to 3CHN
Moved down: Deutsche Bank 2 to 1DEU

Count by jurisdiction over time

This is the one aggregate worth keeping: how many listed G-SIBs sit in each home jurisdiction each year. Zero is a real zero here. It is not the same thing as the 2011 no-bucket state above.

Jurisdiction111213141516171819202122232425
USAlatest 8, peak 8888888888888888
CHNlatest 5, peak 5112344444444555
FRAlatest 4, peak 4444444344444444
GBRlatest 3, peak 4444444433333333
JPNlatest 3, peak 3333333333333333
CANlatest 2, peak 2000000112222222
CHElatest 1, peak 2222222222222111
DEUlatest 1, peak 2211111111111111
ESPlatest 1, peak 2122211111111111
NLDlatest 1, peak 1111111111111111
BELlatest 0, peak 1100000000000000
ITAlatest 0, peak 1111111111111000
SWElatest 0, peak 1111111100000000

Source: FSB, 2025 List of Global Systemically Important Banks (G-SIBs), and 2011-2025 archive Transcribed verbatim from the FSB's annual PDF lists; bucket 5 has never been populated, and the 29 null buckets are the 2011 designation-only list before FSB began publishing bucket assignments in 2012. Jurisdiction is the bank's headquarters/home-regulator domicile, assigned here (the FSB lists do not print it). Methodology

Long-run academic benchmarks

Three independent academic series shown as context, not systemic-risk measures: none feed the LRMES, MES, CoVaR, or absorption-ratio computations above.

Long-run academic benchmarks

US market volatility sits at the 42nd percentile of its record

Rolling 12-month annualized percentage volatility of the Fama-French US market excess return (Mkt-RF), from 1927 through 2026 for the whole US equity market. The absorption ratio above covers 28 banks since 2020, while this series extends back to 1926, giving a century-long scale for “how volatile is right now.” Latest reading 13.0% (May 31, 2026), in the full 1927–2026 distribution. The Depression peaked nearest 71% (1933); the single month of October 1987 (Black Monday) alone returned -23.19%.
Hover for values
This whole-market benchmark provides historical context only and does not enter the systemic-risk measures above.

Source: Ken French Data Library (Fama-French factors) Free with citation, academic research and teaching use; std. dev. of monthly Mkt-RF x sqrt(12). Methodology

Banking-industry cost of capital and beta

Damodaran’s US industry cross-section (2026-01-05): money-center and regional banks against the whole-market baseline. A textbook CAPM/WACC yardstick alongside the market-implied LRMES/MES/CoVaR measures above.

Industry (US)FirmsLevered betaUnlevered betaCost of equityWACC
Bank (Money Center)150.760.347.3%5.0%
Banks (Regional)5680.400.295.7%5.0%
Total Market59940.910.728.0%7.0%

Source: Damodaran Online, NYU Stern (industry cost of capital and betas) Free, acknowledgement welcomed but not required; single cross-sectional snapshot, not a time series. Methodology

Long-run academic benchmarks

Global real rates ended the record far below their historical high

Schmelzing’s 7-year-averaged global real rate, plotted 1317–2018 (the 7-year averaging consumes the first years of the paper’s 1311–2018 window; not extended to today). The deep-history backdrop for the low- and negative-rate environment that shaped the risk-taking these measures track: 1.53% in 2018, against an all-time high of 18.1% (1379) and a low of -10.4% (1945).
Hover for values; dashed line = full-sample mean
Bank of England copyright; non-commercial re-use permitted. Series ends 2018, not extended to today.

Source: Schmelzing (2020), Bank of England SWP 845 Methodology

Nine centuries of British macro, beginning in 1086

The Bank’s research compilation reaches from 1086 to 2016 across seven sections. Each chart below states its own, shorter span; the table shows how far each section actually reaches.

Nine centuries of British macro

Bank Rate closes its three-century record at the all-time low

Annual Bank Rate, percent at period end, United Kingdom, 16942016. It peaked at 17.00% in 1979 and closed at 0.25% in 2016, its record low.
Hover for values
The Bank publishes this as a research compilation: the data do not represent official Bank of England data or National Statistics. The headline series splice historical components together under assumptions. These data end in 2016 and are not extended to today.

Source: Bank of England, A millennium of macroeconomic data for the UK, version 3.1 Worksheet A31. Interest rates. Bank of England copyright; non-commercial re-use permitted. Methodology

Nine centuries of British macro

Britain’s pre-industrial inflation was far more volatile than its postwar record

Annual consumer price inflation, percent, United Kingdom, 12102016. Across the full record it reached 49.66% in 1225 and -31.19% in 1558. From 12101799, the mean was 0.98%, the sample standard deviation was 10.26 percentage points, and 160 of 590 years moved beyond ±10%. From 19462016, the corresponding readings were 4.97%, 4.30 percentage points, and 8 of 71 years.
Hover for values
The Bank publishes this as a research compilation: the data do not represent official Bank of England data or National Statistics. The headline series splice historical components together under assumptions. These data end in 2016 and are not extended to today.

Source: Bank of England, A millennium of macroeconomic data for the UK, version 3.1 Worksheet A47. Wages and prices. Bank of England copyright; non-commercial re-use permitted. Methodology

The compilation’s reach varies by section

First and last observation years describe section-level reach, not continuous coverage for every series. Counts are distinct published series names.

SectionSeriesFirst yearLast year
Labour, capital and productivity910862016
Money and Credit810862016
National Accounts1610862016
Wages and Prices1012092016
Fiscal516892016
Financial markets1216942016
Trade317722016

Source: Bank of England, A millennium of macroeconomic data for the UK, version 3.1 Section minima and maxima across published observations. Bank of England copyright; non-commercial re-use permitted. Methodology

See the full methodology for the measures, windows, universe, the repaired engine input contract, method labels and numerical fallbacks, original output provenance, crisis-anchor validation, and why the dollar SRISK level is a documented open gap.