Comment Text:
To the Secretary of the Commission:
As a Certified Public Accountant, a 50 percent owner of a public accounting firm, and a
professional with an academic background in statistics and industrial-organizational
psychology, I am submitting this comment in response to the Commission's Advance Notice of
Proposed Rulemaking on event contract derivatives traded on prediction markets, published in
the Federal Register on March 16, 2026 (91 FR 12516). The Commission has issued this
ANPRM seeking public comment on statutory core principles and Commission regulations that
apply to prediction markets, the types of event contracts that may be prohibited as contrary to
the public interest, cost-benefit considerations, and other topics. (Federal Register, 91 FR
12516, March 16, 2026)
My comments address three of the Commission's enumerated questions and are supported by
peer-reviewed research, documented market incidents, and the CFTC's own regulatory record.
Response to Question 10: Price Risk, Price Discovery, and Hedging Utility
In the aggregate, prediction markets fail to provide reliable commercial hedging utility at scale.
To function as a legitimate hedge, a derivative must offer deep liquidity and meaningful
resistance to artificial price distortion. Event contracts (particularly those with low trading
volumes) structurally fail this test.
A "head market" siphon effect dominates the prediction market landscape: in top markets such
as a presidential election, liquidity is excellent and price discovery is efficient. However,
thousands of more niche, vertical markets face a severe lack of liquidity. In these long-tail
markets, a large order from a few individuals can easily manipulate the price, causing the
market's core value as a "truth engine" to vanish. (KuCoin Ventures, The Prediction Market
Playbook, November 2025)
This is not a marginal concern: it describes the overwhelming majority of contracts currently
listed. The Commission's own ANPRM notes that event contract activity has grown from an
average of approximately five contracts listed per year between 2006 and 2020, to
approximately 1,600 in 2025 alone. The multiplication of contract volume has dramatically
outpaced the liquidity available to support meaningful price discovery across that full catalog.
(Cleary Gottlieb, Prediction Markets for Those Who Don't Predict, April 2026)
Academic research corroborates the structural problem. Research from the Wharton School's
Initiative for Financial Policy and Regulation notes that uninformed noise bettors can derail the
benefits of liquidity and decrease market efficiency, and that insufficient liquidity means being
unable to fully diversify away risk. A contract that cannot support a modest commercial hedge
without material price distortion does not satisfy the requirements of CEA Section 3, which
identifies price discovery and the deterrence of market disruption as core statutory purposes.
(Wharton IFPR, A Primer on Prediction Markets, January 2025)
On the applicable standard: DCM Core Principle 3 requires that contracts must not be readily
susceptible to manipulation, and Core Principle 5 specifies that speculative position limits or
position accountability must be adopted where necessary and appropriate to reduce the
potential threat of market manipulation or congestion. The liquidity conditions prevailing in the
vast majority of listed event contracts create a structural failure against both standards. (CFTC,
Economic Requirements — Core Principles)
Response to Question 30: Manipulation Risk and Token-Holder Governance
Many modern prediction markets (particularly those utilizing blockchain infrastructure) rely on
governance models that confer disproportionate resolution authority to participants who hold
voting tokens. This concentration of power introduces material manipulation risk that is not
hypothetical; it has already materialized in documented form on one of the largest prediction
market platforms currently operating.
In March 2025, a contentious $7 million bet on Polymarket (speculating on whether Ukraine
would agree to a mineral deal with President Trump before April) saw its "yes" probability surge
from 9% to 100% over a 24-hour period, despite no official agreement being reached. Multiple
users alleged that a large holder of UMA tokens cast a significant number of votes to yield a
resolution in their favor. (CoinDesk, Polymarket, UMA Communities Lock Horns After $7M
Ukraine Bet Resolves, March 27, 2025) Polymarket acknowledged the governance attack, admitting the market resolved against expectations, but declined to issue refunds, stating it was not a "market failure." The platform's own team characterized the incident as "unprecedented." The mechanism that enabled it was not a bug, but the platform's designed governance structure. (The Block, Polymarket Says Governance Attack by UMA Whale Is 'Unprecedented', March 27, 2025) Analysis of the UMA voting distribution reveals that a small number of wallets control a majority of the active voting power. This concentration creates a centralization vector that undermines the value proposition of the platform. The governance model effectively functions as an oligarchy rather than a democracy. The problem extends beyond any single market, casting
doubt on the reliability of every market relying on this resolution mechanism. (Webopedia, Why
Is Polymarket's UMA Controversial?, January 2026) The CFTC previously stated in its legal case against Kalshi that unregulated prediction markets, including Polymarket, are prone to manipulation. (The Block, Polymarket Says Governance Attack by UMA Whale Is 'Unprecedented', March 27, 2025)
The attempted remediation has introduced its own problems. In August 2025, UMA updated its
oracle to restrict resolution proposals to a whitelist of experienced proposers: a transition from
open community governance to a centralized council criticized for limiting community
participation and reducing decentralization. A resolution mechanism that begins as
token-holder democracy and is reformed toward a whitelist of platform employees and affiliated
parties does not resolve the manipulation concern; it merely relocates it. (The Block, UMA
Oracle Update Limits Resolution Proposals to Whitelisted Parties, August 12, 2025)
DCM Core Principle 4 requires that a DCM have the capacity and responsibility to prevent
manipulation, price distortion, and disruptions of the cash settlement process through market
surveillance, compliance, and enforcement practices and procedures. Core Principle 12
requires a DCM to establish and enforce rules to protect markets and market participants from
abusive practices and to promote fair and equitable trading. Token-weighted governance
structures in their current form are structurally incompatible with these requirements. The
Commission should consider whether any event contract listed on a platform whose resolution
mechanism relies on concentrated token-holder voting can, as a matter of first principles,
satisfy Core Principle 3's requirement that contracts not be readily susceptible to manipulation.
(Norton Rose Fulbright, CFTC Advances Regulatory Framework for Prediction Markets, March
2026)
Response to Questions 10 & 11: Price Dissemination, Forecasting Accuracy, and the Public Interest
A central justification offered by proponents of expanded prediction market regulation is that
these markets function as superior forecasting tools: aggregating dispersed information and
producing probability estimates that exceed what conventional polling or statistical modeling
can achieve. The empirical record does not support this claim, particularly for political and
public affairs markets, which represent the most prominent category of contracts currently at
issue.
Proponents of prediction markets will correctly note that Polymarket priced Donald Trump as a
meaningful favorite throughout the final weeks of the 2024 presidential election, while polling
averages showed a near-dead-heat. Trump's decisive victory is cited as evidence of prediction
markets' superior forecasting ability. That result deserves acknowledgment. (TokenMetrics, Are
Prediction Markets More Accurate Than Polls?, February 2026)
However, a single high-profile correct call (on the highest-volume, most-liquid event contract in
the market's history) does not establish a reliable forecasting track record, and it should not be
allowed to obscure the broader empirical picture that existed before prediction markets became
mainstream instruments commanding the volume necessary to attract sophisticated
participants.
The relevant comparison is the 2022 U.S. midterm cycle, the last major election before
prediction markets achieved their current scale and retail prominence, and therefore a cleaner
test of the underlying mechanism rather than of post-hoc liquidity effects.
Even in the 2022 U.S. midterm elections, with no exceptional confounding factors and fully
modern prediction markets in operation, prediction markets fared worse than expert
forecasters. In a systematic comparison graded on a log-odds scoring method, the election
forecasts from FiveThirtyEight's model were more accurate than prediction markets from
Manifold Markets, Polymarket, and PredictIt. The two best predictors of the 2022 midterm
results were the two sites least reliant on betting and market mechanisms and most reliant on
specialist expertise. Prediction markets deviated from the publicly available FiveThirtyEight
predictions, which were proven to be well-calibrated, and added negative value. (Asterisk
Magazine, Prediction Markets Have an Elections Problem, Issue 5)
An independent scoring analysis of midterm forecasts confirmed the same ranking, with
real-money prediction markets Polymarket and PredictIt scoring below both FiveThirtyEight and
the play-money Manifold platform. (First Sigma, Scoring Midterm Election Forecasts,
November 2022)
This finding is reinforced by peer-reviewed research in the International Journal of Forecasting.
A systematic historical assessment of U.S. presidential election betting markets found that
market prices are far better predictors in periods without polls than when polls are available,
and that once scientific polling is available, last-minute market prices add nothing to election
prediction beyond what polls already provide. (Erikson & Wlezien, Markets vs. Polls as Election
Predictors, International Journal of Forecasting, Vol. 38, Issue 3, 2022)
The mechanism underlying this forecasting failure is well-characterized. Analysis of Kalshi
demonstrates that the market is split into two distinct populations: a taker class that
systematically overpays for low-probability, affirmative outcomes, and a maker class that
extracts this premium through passive liquidity provision. When the topic allows for tribalism
and hope (sports, entertainment, political contests), the market transforms into a mechanism
for transferring wealth from the optimistic to the calculated, rather than a reliable aggregator of
probabilistic information. The 2024 result is consistent with this framework: a massive
post-election liquidity surge attracted sophisticated market makers who corrected prior
inefficiencies. That is not the same as the market being structurally reliable. It became more
accurate precisely when volume made it profitable for professionals to trade it efficiently. The
same liquidity conditions do not exist across the catalog of 1,600 contracts now listed annually.
(Becker, The Microstructure of Wealth Transfer in Prediction Markets, January 2026)
A market that measures social sentiment and speculative momentum rather than objective
probability does not fulfill the purposes of the Commodity Exchange Act. The Commission has
asked how the objectives of CEA Section 3 should be weighed in any public interest
determination: deterring manipulation, protecting market participants, ensuring financial
integrity, and promoting responsible innovation. The answer, on the present record, is that the
first three objectives are in tension with expanding the footprint of retail-facing event contracts
without meaningful structural controls. (Cleary Gottlieb, Prediction Markets for Those Who
Don't Predict, April 2026)
Conclusion:
The aggregate data and structural mechanics of prediction markets demonstrate that they are,
in their current form, highly susceptible to capital gaming, token-holder manipulation, and
sentiment-driven price distortion that decouples contract pricing from objective probability.
These are not theoretical concerns. They are documented in published academic research, in
the CFTC's own enforcement record, and in incidents that have already harmed market
participants on platforms currently seeking DCM registration.
I urge the Commission to:
1. Establish minimum liquidity thresholds as a precondition for event contract listing, to
ensure that contracts offered as commercial instruments can actually sustain commercial
use without being susceptible to small-capital price manipulation;
2. Require that any platform utilizing token-weighted or concentrated governance
mechanisms for market resolution demonstrate, with specificity, how those mechanisms
satisfy DCM Core Principles 3, 4, and 12 before self-certification is accepted; and
3. Weigh the empirical forecasting record of prediction markets (which shows
systematic underperformance relative to rigorous statistical models in political and public
affairs contracts prior to the attainment of high-volume liquidity) when evaluating whether
specific event contracts serve the price discovery and public interest purposes enumerated in CEA Section 3.
The Commission has an opportunity in this rulemaking to establish a durable,
evidence-grounded framework rather than one premised on the assumption that market
participation is itself a sufficient proxy for public benefit.
Respectfully submitted
Certified Public Accountant
