Fair value as a fence: the 2018 Volcker proposal let the balance sheet decide what counted as trading

In May 2018 regulators proposed to define the trading account by fair value accounting. The borrowed definition promised certainty but would have handed the boundary to standard setters and to banks’ own accounting elections, which is why it failed.

Looking Back revisits past weeks one at a time, up to the most recent, and asks what it meant for the banks, health systems and public agencies we serve. The series was written in 2026, with the benefit of hindsight, and each article says which week it looks back on.

Two-minute summaryThe dek and the first sentence of each section, quoted from the article

In May 2018 regulators proposed to define the trading account by fair value accounting. The borrowed definition promised certainty but would have handed the boundary to standard setters and to banks’ own accounting elections, which is why it failed.

  1. The week On Wednesday, May 30, 2018, the Federal Reserve Board asked for comment on a proposal to simplify and tailor compliance requirements under the Volcker rule.
  2. Scope: what a fair value test would have captured The first part concerns breadth.
  3. Control: who would have moved the boundary The second part concerns authority.
  4. Proof: what a bank would have needed to show The third part concerns evidence of scope.
  5. The exit: what the agencies kept and what they borrowed instead The fourth part concerns how the idea ended.
  6. Objections, and where the argument is weakest The argument that the accounting prong failed because it borrowed from the wrong owner has to answer four objections.
  7. Checks for any rule that borrows a definition The four parts translate into checks a chief compliance officer or head of regulatory reporting can run when a rule cross-references another regime.
  8. With hindsight The proposal was published in the Federal Register on July 17, 2018, with comments due by September 17, 2018.

The week

On Wednesday, May 30, 2018, the Federal Reserve Board asked for comment on a proposal to simplify and tailor compliance requirements under the Volcker rule. The rule implements section 13 of the Bank Holding Company Act. By statute, it generally prohibits banking entities from proprietary trading and from owning or controlling hedge funds or private equity funds. The Board’s release said all five agencies that administer the rule had developed the proposal jointly. They are the Board, the Commodity Futures Trading Commission, the Federal Deposit Insurance Corporation, the Office of the Comptroller of the Currency and the Securities and Exchange Commission.

The release listed six changes. The proposal would tailor compliance requirements to the size of a firm’s trading assets and liabilities. It would revise the definition of “trading account,” in part by relying on commonly used accounting definitions. It would treat trading within appropriately developed internal risk limits as permissible market making or underwriting. It would streamline the hedging exemption, limit the rule’s reach into foreign banks’ activity abroad and simplify the trading data firms report to the agencies.

The staff memorandum to the Board, dated May 25 and published with the meeting materials, set out the trading account change in detail. The 2013 rule defined the trading account through three prongs: a short-term intent prong, a market risk capital prong and a dealer prong. The short-term intent prong carried a rebuttable presumption that an instrument held for fewer than 60 days was in the trading account. Staff proposed to keep the market risk capital and dealer prongs, to drop the 60-day presumption and to replace the short-term intent prong with an accounting prong.

Under the accounting prong, the trading account would include any account used to buy or sell financial instruments that are recorded at fair value on a recurring basis under applicable accounting standards. The memorandum said this would generally cover derivatives, trading securities and available-for-sale securities. To soften the effect, desks caught solely by the accounting prong would be presumed compliant. The presumption would fall away if the absolute value of a desk’s daily net realized and unrealized gains or losses exceeded $25 million at any point over a 90-calendar-day period. Each agency would keep authority to decide case by case that a trade was or was not in the trading account.

The memorandum was explicit about what the presumption would and would not do. Desks under the threshold would remain subject to the statutory prohibition on proprietary trading and the applicable regulatory requirements. They would not have to demonstrate their compliance with the proprietary trading restrictions on an ongoing basis. A desk whose figure exceeded $25 million would have to demonstrate compliance with the statutory prohibitions and the regulation. The accounting prong and the presumption were designed as a pair.

The week’s context mattered. The Economic Growth, Regulatory Relief, and Consumer Protection Act, enacted on May 24, had exempted community banks with limited trading from the statutory prohibitions. The memorandum said those changes would be handled in a separate rulemaking. Governor Lael Brainard supported the proposal in a statement that day. She highlighted the tailored tiers and the move to internal risk limits, which, she noted, would remain subject to supervisory review. The release said comments would be accepted for 60 days after publication in the Federal Register.

Much of the May 2018 Volcker proposal was about tailoring: three tiers of firms, lighter programs for smaller traders and a presumption of compliance for the smallest. Those changes survived. One of its central ideas did not. The agencies proposed to let accounting treatment decide which positions fell inside the trading account, the boundary that determines where the proprietary trading ban applies at all. A position recorded at fair value on a recurring basis would be in. A position carried at amortized cost would be out, unless another prong caught it.

Our argument is that the accounting prong was an attempt to borrow a definition from another rulebook, and that a borrowed definition brings its owner, its purpose and its choices with it. The agencies wanted certainty, and the ledger offered certainty. But accounting classifications answer questions about how to measure and report value, not about why a position was taken. The comment process exposed that gap, and the final rule in 2019 abandoned the idea.

The analysis has four parts. The first is scope: what the fair value test would have swept into the trading account, and why the agencies paired it with a profit and loss presumption. The second is control: who would have decided, through standard setting and accounting elections, where the boundary fell. The third is proof: what evidence a bank would have needed to show compliance under the new test, and which function would have owned it. The fourth is the exit: why the agencies dropped the prong and what they borrowed instead. Two closing sections test the argument against objections and turn it into checks a compliance officer can run on any rule that borrows a definition.

The parts do not overlap. Scope is about what the definition captured. Control is about who could move it. Proof is about how a bank would have shown where it stood. The exit is about how the rule actually ended up. Read together, they explain why a proposal that looked like simplification would have shifted work from trading desks to finance and accounting teams.

The audience is wider than the largest dealers. Regional banks with sizable securities portfolios, foreign banks with U.S. operations, broker-dealers inside bank holding companies and their technology teams all had a stake in where the boundary fell. Nothing below is advice on any firm’s position. Firms subject to the rule should rely on the rule text and their counsel.

Scope: what a fair value test would have captured

The first part concerns breadth. The statute defines the trading account by purpose: accounts used principally for selling in the near term, or otherwise with the intent to resell to profit from short-term price movements. The 2013 rule translated that purpose into a short-term intent prong, backed by a 60-day rebuttable presumption. The 2018 proposal kept the purpose in the statute but replaced the regulatory test with a measurement attribute.

What fair value on a recurring basis means

Under U.S. accounting standards, some instruments are remeasured at fair value at every reporting date. The proposal’s preamble defined fair value under GAAP as the price that would be received to sell an asset or paid to transfer a liability in an orderly transaction between market participants. The memorandum said the prong would generally cover derivatives, trading securities and available-for-sale securities. Each of those categories exists for reasons that have nothing to do with the Volcker statute.

The proposal itself noted that derivatives are required to be recorded at fair value. That requirement applies whether a swap is a dealer’s inventory or a hedge of a mortgage portfolio. Available-for-sale is an accounting category, and its label says nothing about whether the bank intends to sell in the near term. Securities a treasury function buys for liquidity, and holds for years, can sit in that category.

Why the agencies paired it with a presumption

The staff memorandum recognized the breadth. It proposed a presumption of compliance for desks caught only by the accounting prong, as long as their absolute daily profit and loss stayed under $25 million over any 90-calendar-day period. The memorandum explained the choice of absolute value: a desk could not avoid the threshold by offsetting a large gain on one day with a large loss on another.

The presumption reveals the design logic. The accounting prong would make the trading account bigger and clearer. The profit and loss test would then remove much of the new territory from day-to-day compliance work. Two rules would do the work one rule did before. The first would draw the line by classification. The second would excuse much of what fell inside it by outcome.

That pairing had a cost. A desk that crossed the threshold would have to demonstrate compliance with the full proprietary trading prohibition, including activities it had never treated as trading. A threshold set in dollars does not scale with a desk’s size or volatility. The same $25 million figure would apply to a large dealer’s rates desk and to a regional bank’s investment portfolio. A volatile quarter in rates could push a long-term portfolio over the line through unrealized moves alone.

Which firms the prong would have reached hardest

The effect of the prong would not have fallen evenly. The proposal’s preamble said that for entities not subject to the market risk capital prong or the dealer prong, the accounting prong would be the sole avenue into the proprietary trading requirements. Large dealers were already inside the trading account through the capital and dealer prongs. For them, the accounting prong would mostly add positions held outside their dealer businesses.

For a bank that did not calculate capital under the market risk capital rule and was not a registered dealer, the picture was different. Its trading account under the 2013 rule depended on the short-term intent prong and its 60-day presumption. Under the proposal, it would depend on which of its instruments were recorded at fair value. A regional bank with a large available-for-sale portfolio and interest rate swaps hedging its loans would find both inside the new boundary, whatever their purpose.

Commenters made this point to the agencies. The final rule’s summary of the comments records the argument that the prong would have a disproportionately negative impact on midsize and regional banks. It also records concerns about securitization liquidity and about investment advisers’ use of seed capital for new products. Those are businesses far from what the statute calls short-term trading.

Foreign banks faced a separate version of the problem. The prong referred to applicable accounting standards. A foreign bank could report under its home framework at group level while its U.S. entities report under GAAP. Commenters warned that foreign banks might have to apply both frameworks to non-U.S. positions, one for ordinary reporting and one for Volcker compliance.

Broker-dealers inside bank holding companies were in a different place again. As registered dealers they were already inside the trading account through the dealer prong for their dealing activity. The accounting prong would have mattered for positions they held outside that activity, such as investments carried at fair value for other reasons. In each case, the same instrument class would sit inside or outside the boundary depending on the firm’s accounting, its licenses and its capital rules.

What the agencies asked

The proposal itself flagged the risk. Among the numbered questions on the accounting prong, the agencies asked whether all fair value instruments should be included, whether including available-for-sale securities was appropriate and whether including all derivatives was appropriate. They also asked whether practical expedients to fair value, such as net asset value measurements for certain investments, should count. The questions show the agencies knew the borrowed definition did not map neatly onto the statutory purpose.

For a bank’s technology team, scope translated into an inventory question. Which systems hold positions recorded at fair value, at which legal entity, under which accounting framework? A Volcker inventory built around trading desks would not answer that. Positions in the treasury book, the investment portfolio and structured hedges can sit in different systems, with different owners and different data models. Mapping them onto desks would have been a project in its own right.

Control: who would have moved the boundary

The second part concerns authority. Under the 2013 rule, the agencies wrote the short-term intent prong and interpreted it through guidance, examination and enforcement. Under the accounting prong, the content of the boundary would have depended on decisions the Volcker agencies did not make.

Standard setters

U.S. GAAP and International Financial Reporting Standards are written by accounting standard setters, not by the five Volcker agencies. Their standards serve financial reporting, not the Bank Holding Company Act. When a standard setter changes how an instrument is measured, an accounting prong would carry that change into the Volcker trading account automatically.

The proposal recognized the cross-border version of this problem. One of its questions asked whether there were differences in how IFRS and GAAP apply that the agencies should consider, and how they would affect the prong’s scope. A foreign bank with U.S. operations reports under its home framework. Its U.S. subsidiaries may report under GAAP. The same instrument could be inside the trading account in one ledger and outside it in another.

The bank’s own elections

Accounting standards also give preparers choices. A bank can elect the fair value option for certain instruments. It can classify securities as available for sale or held to maturity, within the standards’ conditions. Under the accounting prong, those elections would have doubled as Volcker scoping decisions.

The agencies saw this too. The proposal asked whether the prong was appropriate given that entities may have discretion over whether certain instruments are recorded at fair value. It asked directly whether the prong could incentivize banks to change their accounting treatment to evade the prohibition, and whether that could affect their accounting practices.

That second question points to a cost that runs the other way. If electing fair value pulled an instrument into Volcker compliance, a bank might decline the election. Commenters later told the agencies exactly this: the prong could discourage fair value elections, reduce transparency in financial reporting and frustrate risk management built on the fair value option. In our reading, a definition that changes accounting behavior has stopped being neutral.

A contrast: limits the firm sets but the supervisor reviews

The same proposal moved another boundary onto firm-produced artifacts, and the comparison is instructive. Under the statute, market making and underwriting are permitted when designed not to exceed the reasonably expected near-term demands of clients, customers or counterparties. Staff wrote that the 2013 rule’s requirements for demonstrating compliance with that standard may unnecessarily constrain use of the exemptions. The proposal would instead presume compliance when a desk set internal risk limits and implemented, maintained and enforced them.

Here too, the bank’s own numbers would define compliance. The difference lies in who could review them. Internal limits under the proposal remained subject to supervisory review, and Governor Brainard’s statement said so expressly. The limits were written for the Volcker purpose. An examiner could challenge a limit as too loose for a desk’s business. An accounting classification, by contrast, is reviewed by auditors for compliance with accounting standards, not for its fit with the Volcker statute.

In our reading, that difference explains why one firm-produced boundary survived and the other did not. A boundary the firm sets for the regulatory purpose, under supervisory review, can be tested against that purpose. A boundary inherited from another discipline can only be tested against that discipline’s rules.

Five agencies, one boundary

The proposal added a second source of movement. The memorandum said each agency would reserve authority to determine, case by case, that a purchase or sale is or is not for the trading account, and to require a desk to demonstrate compliance. The authority was a safety valve against the breadth and the gaps of the accounting test.

A safety valve held by five agencies has its own cost. A bank holding company can own a national bank supervised by the OCC, a broker-dealer registered with the SEC and a swap dealer registered with the CFTC. Each entity answers to its own agency for Volcker purposes. Under the proposal, the same instrument class could be treated differently by different agencies within one group, each exercising its reserved authority.

The 2013 rule already had five agencies applying one regulation. The accounting prong would have raised the stakes, because the boundary itself would be broader and more dependent on judgment calls at the margin. For a group compliance function, that meant tracking determinations by agency and by legal entity, and reconciling them when they differed.

What control meant inside the bank

The governance consequence is concrete. Under the accounting prong, a change in accounting policy would also be a change in Volcker scope. The controller’s office, the accounting policy group and the external auditor would become participants in proprietary trading compliance. A memo approving a classification would need a second review for regulatory effect.

Where a bank runs change controls on accounting policy, the proposal would have required those controls to include a compliance reviewer. The Volcker program would also have needed to monitor standard setters’ agendas. That is a different job from monitoring trading desks.

Proof: what a bank would have needed to show

The third part concerns evidence of scope. Under the 2013 rule, a bank argued scope from its trading records: why a position was taken and how long it was held. The accounting prong would have moved that argument into accounting records, and its presumption would have added a daily figure that many non-trading books did not produce.

Daily desk-level profit and loss

The presumption depended on a daily figure: the absolute value of net realized and unrealized gains and losses on each trading desk’s portfolio. To use it, a bank would need that figure for every desk caught by the accounting prong, every day, retained so that any 90-day window could be reconstructed.

Where a desk is already managed on daily results, the figure exists. Where an investment portfolio or a hedging book is measured only for periodic financial reporting, it may not, and the general ledger may not map the book to anything called a desk. The proposal would have pushed daily valuation, attribution and retention into places that had not needed it.

Lineage from the ledger to the desk

The accounting prong also made data lineage a compliance control. The scoping question would be answered by a measurement attribute stored in the accounting sub-ledger. The compliance question would be answered at the desk. Linking the two requires a reliable mapping from accounting records to the organizational units the rule calls trading desks.

Examiners would test that mapping. They could ask how a position’s measurement basis was recorded, who could change it and how changes flowed to Volcker reporting. A gap in lineage would be a gap in the evidence that the firm knew where its trading account began.

Where the technology work would have landed

The proposal would have required joins between systems that had not needed to talk to each other for Volcker purposes. The measurement basis of a position is recorded in accounting systems. The desk hierarchy is recorded in risk and front-office systems. The accounting prong needed both, for every position, every day the presumption was in use.

Three pieces of work follow from that, as we read the proposal. The first is a stable desk identifier carried on every relevant position, including positions in treasury and investment books that had never been assigned to a desk. The second is a daily feed of the measurement basis, so a change in classification reaches the compliance population on the day it happens. The third is a valuation process able to produce daily realized and unrealized results for books measured only at period end.

Each piece has an owner outside compliance. Finance owns the measurement basis. Risk or product control owns desk-level results. Technology owns the feeds and their retention. The proposal would have made all three into Volcker control owners, with the compliance function depending on their work for its evidence.

What examiners would have asked for

An examination of the accounting prong would start with population. The examiner would ask for the list of accounts recorded at fair value on a recurring basis, the desks they mapped to and the reconciliation between that list and the general ledger. A bank that could not produce the reconciliation would struggle to show that its trading account was complete.

The next request would cover changes. Which instruments moved into or out of fair value measurement during the period, and why? Who approved each change, and did anyone consider its Volcker effect? Those questions reach the accounting policy committee, not the trading floor.

The third request would cover the presumption. For each desk relying on it, the examiner would want the daily profit and loss series, the method for computing the absolute value and evidence that no 90-day window exceeded $25 million. Where a window did exceed it, the examiner would want the demonstration of compliance that followed. Each of these is a record that must exist before the examination begins.

Desk boundaries under the presumption

The presumption applied desk by desk, which made the desk itself part of the evidence. Several commenters opposed the presumption, and some argued it could let banks evade the prohibition by splitting trades over multiple trading desks. One suggested it could invite proprietary trading within the $25 million threshold. Two said the statute gave no basis for permitting a de minimis amount of proprietary trading.

Those objections point to a proof problem specific to the prong. A bank relying on the presumption would have needed to show that its desk structure reflected how the business was actually run, not a design to keep each desk under the threshold. In our reading, that would have meant a documented rationale for each desk boundary, approved before the presumption was relied on, and a record of every change to desk membership. A desk reorganized just before a volatile quarter would have invited the question the commenters raised.

The memorandum’s choice of absolute value closed one route around the threshold: offsetting a large gain with a large loss. It did not close the other, which ran through the definition of the desk. An examiner would have had to test both, and the bank would have had to keep the records that let the examiner do so.

The exit: what the agencies kept and what they borrowed instead

The fourth part concerns how the idea ended. The agencies did not adopt the accounting prong. The reasons they gave, and the replacement they chose, show what a workable borrowed definition looks like.

Why the prong failed

The final rule was published in November 2019. In it, the agencies said the comments from banks and industry groups were generally supportive of the proposal, with the exception of the accounting prong. Commenters had argued that the prong would scope in over $400 billion in available-for-sale debt securities, long-term investments, asset-liability management, derivatives entered into for any purpose, seed capital and merchant banking investments. Some argued that it could discourage fair value elections and treat the same activity differently at two banks.

The agencies agreed. They said the prong would have inappropriately scoped in many instruments and activities the statute was not intended to capture, and entire categories of instruments regardless of purpose. They also declined to adopt the desk-level profit and loss presumption, explaining that it had been designed to address the accounting prong’s effects.

Not every commenter opposed the idea. Some supported adopting the prong, and one urged the agencies to keep the short-term intent prong and add the accounting prong as an extra test. Another warned the opposite risk: that a bank could avoid the rule by recording instruments at amortized cost instead. Both sides were describing the same feature. A definition that depends on a choice can be widened or narrowed by that choice.

What replaced it

The final rule kept a modified short-term intent prong. It reversed the 60-day presumption: instruments held for 60 days or longer would be presumed outside the short-term intent prong. A bank subject to the market risk capital prong would no longer also be subject to the short-term intent prong. A bank not subject to it could elect to use the market risk capital prong instead.

That choice is also a borrowed definition. The market risk capital rule decides which positions count as covered trading positions for capital purposes. The difference is ownership. The three banking agencies among the five Volcker agencies write and supervise the capital rule themselves. The preamble to the 2013 rule had already observed that the capital rule’s definition of trading position largely paralleled the statutory definition of trading account. Borrowing from a rule the same agencies control, written for a related prudential purpose, carried less risk of drift than borrowing from accounting standards.

What the reversed presumption changed in operations

The reversal of the 60-day presumption is easy to read as a small edit. Operationally it was larger. Under the 2013 rule, a bank had to rebut a presumption that short holding periods meant trading. Under the final rule, instruments held for 60 days or longer were presumed outside the short-term intent prong, and the presumption could still be rebutted. The evidence burden for long-held positions moved off the bank’s desks.

The holding period is a fact the bank’s systems already record. Trade date, settlement date and disposal date are recorded in position-keeping systems. Using them as the marker kept the objective quality the agencies wanted from the accounting prong, without importing accounting categories. A compliance team can compute the marker from trade records alone, without a classification decision by the finance function.

How the final rule answered the original problem

The agencies explained the choice in terms of the difficulty that had started the exercise. The final rule said the impetus behind the accounting prong had been the uncertain application of the short-term intent prong to certain trades. Instead of replacing that prong, the agencies modified it and added exclusions from the definition of proprietary trading. They called those clarifying measures a more tailored approach to the difficulties that had arisen. The diagnosis survived. The accounting cure did not.

Objections, and where the argument is weakest

The argument that the accounting prong failed because it borrowed from the wrong owner has to answer four objections.

The objection that intent cannot be tested

The strongest case for the prong was that the short-term intent test asks about a state of mind. The proposal’s preamble said the change was meant to address concerns that the statutory definition may be read to call for an inquiry into the subjective intent behind a trade. Accounting classifications are objective and audited. That is a real advantage.

The answer is that objectivity in the wrong variable is not precision. A classification that is objective but unrelated to purpose will be over-inclusive in some places and under-inclusive in others, and firms can move instruments across it. The final rule’s reversed presumption kept an objective holding-period marker while tying it back to purpose.

The 2013 rule had already used an objective proxy: the 60-day presumption. Its weakness, in the agencies’ telling, was uncertainty about how the short-term intent prong applied to certain trades. The fix the agencies chose in 2019 kept a proxy of the same kind, holding period, and changed which way the presumption ran. That is a narrower change than replacing the definition, and it kept the evidence inside trade records.

The objection that the presumption fixed the breadth

A second objection holds that the $25 million presumption would have removed the over-breadth for practical purposes. Desks with modest results would never need to demonstrate compliance.

The comment record suggests otherwise. One commenter said not all businesses calculate daily profit and loss, and that a business that sells nothing for 90 days could still exceed $25 million in unrealized gains and losses. Commenters also said the desk-level presumption did not adequately address the breadth of the accounting prong. A patch that requires new daily data in order to excuse positions that should not have been covered is an expensive patch.

The objection that both tests could have run together

Some commenters proposed a middle course. One urged the agencies to keep the short-term intent prong and add the accounting prong as an extra test without any presumption of compliance. Another suggested making the accounting prong a presumption within the short-term trading prong, coupled with language from the accounting standards. Either version would have used the ledger as evidence of intent rather than as the definition itself.

That approach has merit, and it is closer to what the final rule did with holding periods. But a fair value presumption would still have relied on categories built for another purpose. A bank would still have had to rebut it for every swap hedge and every available-for-sale bond, and the burden would have landed on the same regional banks. The agencies chose a marker taken from the trade itself.

The objection that this was a narrow technical fight

A fourth objection is that the accounting prong was one technical element in a large package, and that its fate says little beyond the Volcker rule. We disagree. The Volcker rule itself cross-references the market risk capital rule, and the proposal would have added accounting standards. Every such cross-reference raises the same questions about scope, control and proof. The 2018 proposal is a well-documented case of an agency asking them in public, receiving answers and changing course.

Checks for any rule that borrows a definition

The four parts translate into checks a chief compliance officer or head of regulatory reporting can run when a rule cross-references another regime. They apply to Volcker, and equally to capital, liquidity, reporting and conduct rules that rely on accounting or other external terms.

  • Scope: list every external definition the rule relies on, and the population each one captures today. Compare that population with the rule’s stated purpose and note the gaps in both directions.
  • Control: for each external definition, name the body that can change it and the internal function that makes elections under it. Confirm that a change by either triggers a compliance review.
  • Elections: identify accounting or classification choices that move positions into or out of a regulatory perimeter. Require a documented regulatory-effect review before any such election is made or reversed.
  • Lineage: trace how the attribute that drives scope flows from its system of record to the compliance population. Test the mapping each quarter, rather than waiting for an examiner to ask.
  • Thresholds: for any presumption or exemption tied to a figure, confirm the bank produces that figure at the frequency the rule assumes, with retention long enough to rebuild any look-back window.
  • Desk design: where a presumption or threshold applies desk by desk, record why each boundary is drawn where it is, and review any change to it before relying on the presumption.

None of the checks requires a new system. They require the compliance function to treat borrowed terms as dependencies, with owners and change controls, the way technology teams treat external libraries. The May 2018 proposal made that dependency visible because it was so large. Smaller ones exist in every rulebook a bank follows.

The checks land differently by type of firm. A regional bank outside the market risk capital rule should start with its investment portfolio and loan hedges, where a borrowed definition can reach positions held for other purposes. A broker-dealer inside a holding company should start with positions held outside its dealing activity, where its dealer registration does not settle the question. A foreign bank should start with the frameworks it reports under, and map which one each regulatory definition refers to. A wealth manager that sponsors funds should start with seed investments and fund interests, which commenters flagged in 2018 as likely to be caught.

Credit unions and community banks outside the Volcker rule can still use the checks. Their own regimes borrow definitions too. A rule that relies on call report categories, accounting classifications or another agency’s thresholds carries the same risk of drift, and the same need for an owner who watches it.

With hindsight

The proposal was published in the Federal Register on July 17, 2018, with comments due by September 17, 2018. Its numbered questions ran to Question 342. Those on the accounting prong included the agencies’ own question about evasion through accounting choices.

The agencies published the final rule on November 14, 2019. They reported more than 75 unique comments from banks, industry groups, public interest groups and others, and about 3,700 short form letters opposing the proposal. They said industry comments were generally supportive except on the accounting prong, and they did not adopt it. They also dropped the desk-level profit and loss presumption that had been designed to offset it.

In place of the accounting prong, the final rule kept a modified short-term intent prong. It reversed the 60-day presumption, so that instruments held for 60 days or longer were presumed outside that prong. Banks subject to the market risk capital prong were no longer also subject to the short-term intent prong, and other banks could elect the capital prong instead.

The final rule also set out what the prong would have caught. It said the prong would have inappropriately scoped in entire categories of financial instruments regardless of the bank’s purpose, such as all derivatives and equity securities with a readily determinable fair value. It added that the prong would have captured certain seeding activity otherwise permitted under the rule. The proposal had asked whether including all derivatives was appropriate; the final rule answered that it was not.

A second rule on covered funds followed. It was published on July 31, 2020, and took effect on October 1, 2020. It modified and clarified the fund provisions that the 2019 rule had largely left for a later proposal.

The eCFR text of section 248.3 of the Federal Reserve’s Regulation VV, current to November 12, 2026, still defines the trading account through three prongs. They are a short-term intent prong, a market risk capital prong and a dealer prong, none of them based on accounting treatment. The section’s version history in the eCFR shows no substantive change since the amendments that took effect on January 1, 2020.

The record also shows how much of the 2018 debate turned on data. Commenters told the agencies that not all businesses calculate daily profit and loss. Because the agencies dropped the presumption along with the prong, the rule did not require any bank to build daily desk-level results for books outside its trading businesses.

What held up from May 2018 was the tailoring: tiers by trading activity, lighter programs for smaller traders and internal risk limits for market making and underwriting. What did not hold up was the idea that the ledger could stand in for purpose. The agencies kept a borrowed definition in the final rule, but it was their own capital rule, not an accounting standard written by others for investors. In our reading, that is the durable result: when a rule borrows a definition, the owner of the definition matters as much as its clarity.

Spotted an error? Tell us and we will check it and correct the article where needed.

  • Volcker rule
  • trading account
  • fair value accounting
  • tailoring
  • Federal Reserve