Analysis date: April 24, 2026. Reporting period: December 31, 2025 / Q4 2025. This article uses OCC bank-level tables unless it explicitly says holding-company data.[1]
Bottom Line: The Dealer Core Is Four Banks
If the question is which U.S. banks use derivatives at truly meaningful scale, the public-data answer starts with four names: Goldman Sachs Bank USA, Citibank N.A., JPMorgan Chase Bank N.A., and Bank of America N.A. As of Q4 2025, the OCC reported $208.1 trillion of total derivative notional across insured U.S. commercial banks and savings associations; those four banks held 85.1% of the industry notional, while the top 25 held 99.3%. A total of 1,197 banks and savings associations reported trading and derivatives activity.[1]
| Bank | Total derivatives | Derivatives / assets | Held for trading share | Screen read |
|---|---|---|---|---|
| Goldman Sachs Bank USA | $53.2T | 82.5x | 99.8% | Trading-linked dealer profile |
| Citibank N.A. | $50.8T | 27.7x | 99.7% | Trading-linked dealer profile |
| JPMorgan Chase Bank N.A. | $50.2T | 13.4x | 97.6% | Trading-linked dealer profile |
| Bank of America N.A. | $22.8T | 8.6x | 92.9% | Trading-linked dealer profile, with more non-trading use than the other three |
The held-for-trading percentages above are from the OCC top-four table. The OCC excludes credit derivatives from that trading/non-trading split because the Call Report does not currently differentiate credit derivatives by trading status.[1]
The Next Tier: Large, But Not the Same Question
The next group still matters, but the interpretation changes. A bank can have a large derivatives book because it is a rates dealer, a custody bank with heavy FX activity, a global bank subsidiary, or a regional bank using interest-rate swaps for asset-liability management. The screen should identify the follow-up question, not pretend every notional dollar means the same thing.
| Bank | Total derivatives | Derivatives / assets | Dominant product cue | Screen read |
|---|---|---|---|---|
| Wells Fargo Bank N.A. | $17.8T | 9.8x | 65.8% interest rate, 30.0% FX | Large-scale rates/FX profile; review trading revenue and ALM disclosures |
| State Street Bank & Trust Co. | $2.9T | 8.0x | 94.3% FX | Custody/FX-heavy profile; not automatically a dealer-style risk signal |
| U.S. Bank N.A. | $1.4T | 2.1x | 87.4% interest rate | Likely ALM and client-rate activity first; confirm in disclosures |
| HSBC N.A. | $1.4T | 8.3x | 84.7% FX | Global-bank FX-heavy profile; legal-entity context matters |
| Bank of New York Mellon | $1.3T | 3.5x | 77.9% FX | Custody/FX-heavy profile; compare with State Street and Northern Trust |
| Morgan Stanley Bank N.A. | $1.0T | 4.0x | 52.4% interest rate, 32.3% equity | Capital-markets linked; bank-level figure is not the full BHC story |
| PNC Bank N.A. | $738B | 1.3x | 90.4% interest rate | Material, but the first read is ALM/rates rather than dealer exposure |
| Truist Bank | $515B | 1.0x | 80.0% interest rate | Rates-heavy regional profile; review hedge accounting and NII sensitivity |
| Northern Trust Co. | $412B | 2.3x | 93.7% FX | Custody/FX-heavy profile |
| Capital One N.A. | $349B | 0.5x | 85.3% interest rate | Below dealer scale; review only if fair-value or capital metrics stand out |
The rest of the OCC top-25 bank list was TD Bank N.A., Citizens Bank N.A., Regions Bank, BMO Bank N.A., Fifth Third Bank N.A., KeyBank N.A., Huntington National Bank, Morgan Stanley Private Bank N.A., Manufacturers and Traders Trust Co., Western Alliance Bank, and Comerica Bank. These names are worth keeping in the screen, but they are not the dealer core by Q4 2025 notional scale.[1]
What Meaningful Scale Means Here
Meaningful scale is not the same thing as risk. Notional amount is a reference amount used to calculate payments; it is not the amount the bank expects to lose. But notional does tell you where derivatives are central enough to change the review process. A $70 billion notional book at a regional bank deserves a different read from a $53 trillion notional book at a dealer bank.
For this article, meaningful scale means one of three things: the bank appears in the OCC top-25 derivatives table, derivatives notional is large relative to assets, or the trading/non-trading split points to dealer-style activity. That is a screening definition, not a regulatory threshold.
Use These Rules as a Screen, Not a Verdict
| Measure | Source type | Practical trigger | Why it matters |
|---|---|---|---|
| Total notional / total assets | DDV heuristic using OCC Table 13 | Above 5x is meaningful; above 10x is dealer-scale unless context says otherwise | It normalizes size and separates ordinary regional use from balance-sheet-shaping scale |
| Held-for-trading share | Call Report Schedule RC-L items 13 and 14; OCC Table 17 for the top four | Above 75% is trading-linked by this screen | It is the cleanest public-data split between dealer activity and non-trading ALM use |
| Total credit exposure / risk-based capital | OCC Table 16, derived from Schedule RC-L and RC-R | Above 50% is a capital-relevance flag | It moves the question from notional scale to counterparty exposure relative to capital |
| Credit derivatives / total notional | DDV diagnostic using OCC Table 15 | Above 2% is a complexity flag | Credit derivatives are not split by trading status in the Call Report, so legal-entity review matters |
| Gross positive and gross negative fair values | Schedule RC-L item 15 | Read separately; do not net them away too early | Netting can hide large bilateral exposures that matter under counterparty stress |
The exact Call Report fields behind this screen are Schedule RC-L item 13 for total gross notional derivative contracts held for trading, item 14 for total gross notional contracts held for purposes other than trading, and item 15 for gross positive and gross negative fair values. For credit exposure, the OCC table also uses Schedule RC-R, column B, lines 20 and 21.[2]
A Worked Example: PNC Is Not Goldman
PNC Bank N.A. had $738 billion of derivative notional at Q4 2025, equal to about 1.3x assets. Its product mix was 90.4% interest-rate contracts, and its total derivative credit exposure was 6% of risk-based capital. That is not trivial, but the first analytical question is asset-liability management: what is the bank hedging, how does it affect net interest income sensitivity, and what happens if rates move?
Goldman Sachs Bank USA had $53.2 trillion of derivative notional, equal to about 82.5x assets. The OCC classified 99.8% of its non-credit derivative notional as held for trading and mark-to-market, and total derivative credit exposure equaled 148% of risk-based capital. That is a different analytical problem. The follow-up is about dealer activity, counterparty exposure, collateral, valuation, revenue mix, and stress behavior.[1]
The same word, derivatives, covers both situations. The screen is useful because it keeps those two cases from being flattened into the same story.
The Bank-versus-BHC Trap
Bank-level Call Reports are the right starting point for insured bank subsidiaries, but they are not always the full economic map of a financial group. The OCC reported $206.7 trillion of derivative notional for the top 25 banks at Q4 2025, versus $265.5 trillion for the top 25 holding companies. Morgan Stanley shows why that distinction matters: Morgan Stanley Bank N.A. had about $1.0 trillion of derivatives, while Morgan Stanley at the holding-company level had $32.5 trillion.[1]
That does not make the bank-level screen wrong. It means the scope must be explicit. For bank credit analysis, start with the bank. For a parent-company equity, debt, or systemic-risk view, reconcile the bank data with the FR Y-9C holding-company data.
How To Read the Named Banks
Dealer core: Goldman Sachs Bank USA, Citibank N.A., JPMorgan Chase Bank N.A., and Bank of America N.A. These are the banks where derivatives are unmistakably central to trading and market-making analysis.
Large but context-dependent: Wells Fargo Bank N.A., State Street Bank & Trust Co., HSBC N.A., Bank of New York Mellon, Morgan Stanley Bank N.A., and U.S. Bank N.A. The scale is meaningful, but the product mix and legal-entity structure determine whether the next read is trading, custody FX, client accommodation, or ALM.
Mostly ALM/client-service watchlist: PNC, Truist, TD, Citizens, Regions, BMO, Fifth Third, KeyBank, Huntington, M&T, Western Alliance, and Comerica. The screen should still be run, especially for fair-value exposure and hedge accounting, but these names should not be treated like the dealer core just because they report derivatives.
What To Review Next
- Start with the top-25 table. It captures nearly all bank-level derivative notional in Q4 2025.
- Separate dealer scale from operating-bank scale. Use notional/assets and held-for-trading share before reading narrative disclosures.
- Read product mix before drawing risk conclusions. Interest-rate and FX-heavy books often point to ALM, custody, or client-service activity; equity, credit, and high trading share point to more capital-markets sensitivity.
- Compare gross fair values and credit exposure to capital. Notional gives footprint; fair value and credit exposure give a better stress-review starting point.
- Reconcile bank and holding-company scope. This is essential for firms where trading activity sits outside the insured bank.
For a faster shortlist, the derivatives-usage view in the Banking app is built around this workflow: find the banks that stand out, compare them with peers, and then decide which disclosures deserve a deeper read.
FAQ
Does high derivatives notional automatically mean high risk?
No. Notional is a scale measure, not a loss estimate. A swap book can have huge notional and much smaller fair-value exposure. That is why the screen also looks at gross positive fair value, gross negative fair value, net current credit exposure, collateral, and capital context.
Does an interest-rate-heavy derivatives book mean the bank is mostly hedging?
Not by itself. Interest-rate dominance is consistent with ALM and balance-sheet hedging, but large banks can also run trading businesses in rates. The held-for-trading split and trading revenue context are what make the distinction more credible.
Why can Call Report data understate a group’s full derivatives exposure?
Call Reports describe the insured bank or savings association. A bank holding company can also have material derivatives in non-bank subsidiaries. That is why a parent-company analysis should compare bank-level Schedule RC-L data with holding-company FR Y-9C data.
What is TCE, and should it be used here?
TCE means tangible common equity, usually common equity less goodwill and other intangible assets. Net derivatives fair value / TCE can be useful in a bank-specific stress model, but this article uses the OCC’s comparable public metric, total derivative credit exposure / risk-based capital, as the primary capital lens.
Why are credit derivatives called out separately?
The Call Report does not currently split credit derivatives by trading versus non-trading purpose in the same way it does for other derivatives. A bank with a meaningful credit-derivative share therefore deserves an extra legal-entity and business-purpose review before you classify the exposure.
Sources
- [1] OCC, Quarterly Report on Bank Trading and Derivatives Activities: Fourth Quarter 2025, published March 31, 2026. Source for Q4 2025 bank-level and holding-company derivatives tables, notional amounts, trading/non-trading split, fair values, and credit exposure metrics. URL: https://www.occ.gov/publications-and-resources/publications/quarterly-report-on-bank-trading-and-derivatives-activities/files/pub-derivatives-quarterly-qtr4-2025.pdf
- [2] FDIC / FFIEC 031 and 041 Schedule RC-L instructions. Source for Call Report derivative item definitions, including held-for-trading notional, non-trading notional, and gross fair values. URL: https://www.fdic.gov/bank-financial-reports/ffiec-031-041-schedule-rc-l-derivatives-and-balance-sheet-items-december