A position limit is a wall. A position accountability level is a doorbell. Cross a limit and you have broken a rule; cross an accountability level and you have, at first, only told the exchange something about yourself. The two words sit side by side in one sentence of federal commodities law, the choice between them belongs to the exchange, and almost nobody outside a compliance department could say which one governs the market they are trading.
Both United States exchanges listing event contracts chose the doorbell, and both put it in the same place. On 14 November 2026 we read 120 of the 3,229 distinct contract terms documents behind the 16,441 series on Kalshi's public series endpoint, and 70 of the 202 public product filings of Polymarket US. The figure was twenty five thousand dollars in almost all of them, for a national election and for one baseball game alike. The public record does not say where it came from.
One Rule Lets an Exchange Choose Between a Wall and a Doorbell
Under Core Principle 5 of the Commodity Exchange Act, 7 U.S.C. 7(d)(5), a board of trade "shall adopt for each contract of the board of trade, as is necessary and appropriate, position limitations or position accountability for speculators," to reduce the potential threat of market manipulation or congestion. The Commodity Futures Trading Commission repeats that sentence almost word for word in Commission Regulation 38.300, and Commission Regulation 38.301 adds one line of plumbing: a designated contract market "must meet the requirements of parts 150 and 151 of this chapter, as applicable." Two phrases do the work. For each contract means the statute contemplates a judgment made market by market, not one setting for a venue. As is necessary and appropriate is a standard rather than a number, and one someone has to apply in writing if anyone else is to check it.
The Commission does define the doorbell. In Commission Regulation 150.1, position accountability means a rule submitted under part 40 "in lieu of, or along with, a speculative position limit" which requires a person above the level to consent to provide information about the position and to halt increasing it further or reduce it "in an orderly manner, in each case as requested" by the exchange. That is a complete definition of what accountability does. It says nothing about where the level should sit.
The Paragraph That Sets a Level Stops at Physical Commodities
The level standard exists one paragraph away. Commission Regulation 150.5 governs exchange-set limits, and its paragraph (b) is titled "Requirements for exchange-set limits on commodity derivative contracts in a physical commodity that are not subject to the limits set forth in 150.2." Inside it, (b)(2)(i) is the sentence a prediction market reader wants: an exchange "shall adopt either speculative position limits or position accountability outside of the spot month at a level that is necessary and appropriate to reduce the potential threat of market manipulation or price distortion." Aggregation is scoped the same way: paragraph (d) requires conforming aggregation rules under 150.4 "For commodity derivative contracts in a physical commodity."
An election result and the outcome of a game are not physical commodities. Read against the words of the rule and the "as applicable" in 38.301, the sentence that would attach a level standard to an event contract and the sentence that would attach aggregation to it both stop before they arrive. What remains is Core Principle 5's single sentence.
Appendix F Prices a Limit Four Ways and Accountability Not at All
The asymmetry is impossible to miss in Appendix F to part 150, the Commission's acceptable practices for 150.5. For a position limit, paragraph (a)(1) offers four quantitative routes, any one of which suffices: historical speculative position sizes as a share of average month-end open interest; the contract's own spot-month limit; 5,000 contracts scaled by notional; or ten percent of average month-end open interest up to 50,000 contracts, with a marginal increase of 2.5 percent beyond that. For position accountability, paragraph (a)(2) is one clause: the exchange "adopts position accountability, as defined in 150.1." No level, no formula, no proportion of open interest. The appendix then has a paragraph (b) reading "[Reserved]", while 150.5(b)(2)(ii) tells exchanges they "may refer to the non-exclusive acceptable practices in paragraph (b) of appendix F." The practices are in paragraph (a).
One further sentence rewards slow reading. Paragraph (e) of 150.5 requires a part 40 submission for either choice, then adds: "To the extent that a designated contract market or swap execution facility adopts speculative position limit levels, such part 40 submission shall also include the methodology by which such levels are calculated." The duty to publish a methodology, and the duty in the next sentence to review levels regularly, both attach to limit levels. An exchange picking accountability owes neither.
Security Futures Show What a Conditional Switch Looks Like
There is one corner of the rules where the swap from wall to doorbell is written down for a designated contract market, and it is instructive. Under Commission Regulation 41.25(b)(3)(i)(B), an exchange listing a security futures product on a single equity may adopt "a position accountability rule in lieu of a position limit" only where the underlying's six-month total trading volume exceeds 2.5 billion shares and there are more than 40 million shares of estimated deliverable supply. The threshold is named in the rule, at 25,000 one-hundred-share contracts or a lower level the exchange specifies, and so are the duties: provide information on request, consent to halt increasing the position when ordered. Three things are therefore fixed by the regulation itself: the switch is earned by measured depth, the level carries only downward discretion, and the consequences are spelled out.
Kalshi reproduces all three for its own perpetual security futures products. Rule 14.23(d) of the Kalshi DCM Rulebook version 1.31, a 50,134 word document last modified on 4 November 2026, permits accountability in lieu of a limit on the same volume and deliverable-supply test and fixes the level at 25,000 Standard Contract Equivalents "or such lower level as the Exchange may specify." Same exchange, same rulebook, same word. For a security futures product the switch is gated by two measurements of depth. For an event contract, nothing in these texts gates it.
Two Rulebooks Define the Term in Two Different Units
Here the venues part company, in their definitions sections. The Kalshi rulebook defines a Position Accountability Level as "the level of loss that can be incurred as a result of a position in a Contract that is held or controlled by a Member, above which Kalshi may impose restrictions or obligations on the Member." The Polymarket US rulebook, a 35,451 word document dated 30 September 2026, defines the same term as "the size of positions in a Contract above which a Participant holding the positions is required to submit reports to the Company."
Loss and size are not the same quantity, and on a binary contract the gap is wide: a trader long at ninety cents and a trader short at ten cents can hold the same number of contracts while facing very different downside. Kalshi's own contract terms then express the level as a flat dollar figure rather than a loss calculation, qualified "per strike, per Member" in 90 of our 120 documents and "per Member" in 26. None of this is hidden and none of it is wrongdoing. It is that a trader wanting to know how close they stand must first work out which quantity is measured. Directive 02 asks for exactly this to be stated plainly where the trade happens, which is why we read terms sheets rather than marketing pages, as we did for a different line of the same filing in our piece on legs nobody played.
Crossing the Level Does Different Things on the Two Venues
On Kalshi, Rule 5.18 sets out a ladder. Accountability levels apply to all contracts as specified in each contract's terms, except where a position limit applies instead. A participant above the level must give compliance staff "all information regarding its position" that staff deems necessary, must then refrain from increasing it or reduce it if instructed, and failing that the exchange "shall have the authority to liquidate the applicable position" back below the level. Non-compliance can reach disciplinary proceedings under Chapter 9.
On Polymarket US, Rule 5.9(b) is two clauses. The exchange may at any time require information on the nature and size of a position, the strategy behind it, and support for any claim that it is a hedge. It may then "order the reduction of any such position of a Participant who fails to provide the information as directed, or in its discretion, order that a position above the position accountability threshold not be increased." The trigger for an ordered reduction is non-cooperation rather than the position itself, no clock is stated, and the rule confers no power to liquidate. Rule 3.10(g)(i) separately attaches a records duty at the same level, for the life of the contract and five years after.
The Aggregation Clause Sits on the Rule Neither Exchange Uses
This part has consequences for a thin market, because concentration is the risk accountability is meant to surface and it is usually built across accounts rather than inside one. Kalshi's limit rule handles that: Rule 5.19(f) applies limits to all positions in accounts where any person directly or indirectly holds positions or controls trading, and to positions held by two or more persons "acting pursuant to an express or implied agreement or understanding" as if one person held them. Rule 5.18, the accountability rule, has no equivalent clause. Polymarket US is more explicit still: Rule 5.9(a)(iv) provides that "For purposes of this Rule 5.9(a), the Company will aggregate the Accounts for all Participants as provided by CFTC Regulation 150.4." Rule 5.9(a) is the limit paragraph; accountability in 5.9(b) falls outside that sentence.
So in the regime both venues actually operate, neither rulebook states that two coordinated accounts are added together before the doorbell rings, and this is the question the Commission itself put to the industry. Question 2.e of the advance notice titled Prediction Markets, published at 91 FR 12516 on 16 March 2026, asks "what factors should the Commission consider regarding how position limits across similar event contracts should be aggregated," and whether existing accountability provisions are "helpful as precedent" for prediction markets. We counted the words of the Commission's proposed rule, published at 91 FR 35837 on 12 June 2026: across 76,474 words the phrase "position accountability" appears zero times, and so does "Core Principle 5."
The Same Figure Covers a Presidency and One Baseball Game
Back to the measurement. In the 70 Polymarket US product filings we read on 14 November 2026, all 70 state "No position limits apply" and all 70 set the accountability level at twenty five thousand dollars of notional exposure. In the Kalshi sample of 120 contract terms documents, 119 state an accountability level and exactly one states a position limit; 112 of those levels are twenty five thousand dollars, five are two hundred fifty thousand, and one is given in contract units instead, at 7,000,000 Contracts on the unemployment market. Those few exceptions are the most useful data points, because they prove the level is adjustable: nothing external pins it at twenty five thousand, which is simply the setting that applies when nobody revisits it. Nor do the filings explain it. The contract terms of 22 September 2026 for Polymarket's combination outcome contracts run 2,153 words; the accountability level appears once, as a figure, "No position limits apply" appears once, and Core Principle 5, 38.300 and 150.5 appear not at all. The compliance analysis sits in the confidential Attachment B, the structural problem we examined in our piece on confidential treatment.
Three Thousand Five Hundred Dollars Is the More Legible Number
There is a per-trader cap in the Commission's event-market record that works, and it is a fraction of the size. In CFTC Staff Letter No. 25-20 of 14 July 2025, the Division of Market Oversight amended the no-action position for the PredictIt market. It removed the old limit of 5,000 traders per contract, reasoning that the cap kept new traders out of popular contracts and distorted the data, and reset the investment limit for a single participant in any particular contract to "the limit for individual contributions to campaigns per election, per candidate, under the Federal Election Campaign Act, as that limit is adjusted going forward." A footnote names the figure: 3,500 dollars at the time of the letter, adjusted for inflation by the Federal Election Commission every two years. The Division said it was satisfied the caps would keep the market "small-scale" while permitting liquidity.
One cap is seven times smaller and governs an academic market the regulator explicitly wants kept small, but anyone can look up where the number came from and when it next moves. The other is uniform across sixteen thousand series and two hundred filings, and its derivation is in no document a trader can open. Size is not what makes a number a standard. Traceability is.
Four Things a Venue Could Publish Without Waiting for a Rule
None of this needs the Commission to move first, which is the premise of this initiative. Four things would turn an accountability level from a default into a standard.
- The unit. Say whether the level measures notional exposure, contract count or maximum loss, in the same words in the rulebook and in the contract terms. As of 14 November 2026 those two documents do not match on either venue.
- The derivation and the review cadence. Core Principle 5 asks for a judgment made for each contract, and 150.5(e) already requires a published methodology from an exchange choosing limits. One choosing accountability can volunteer the same.
- Whether accounts are aggregated. One sentence extending the existing aggregation clause from the limit rule to the accountability rule closes the gap the Commission asked about in March 2026, before anyone has to.
- What crossing it does, with a clock. Kalshi states an escalation ending in liquidation, Polymarket a reduction power triggered by non-cooperation. Either is defensible, but the consequence and the timetable should be readable before the position exists, not after the first request arrives.
That is Directive 02 applied to a line in a terms sheet and Directive 05 applied to the jurisdiction that regulates next rather than the one that has not yet. Less abstractly, it is the difference between a venue where a very large position in a thin book is noticed by design and one where it is noticed if somebody happens to look. Venue risk of this kind is why we wrote that you are not only trading the event, and firms wanting to hold this floor before it is set for them can sign at our commitment page.