Commodity Futures Trading Commission v. James A. Donelson

In the
United States Court of Appeals
For the Seventh Circuit
____________________
No. 23-1809
C OMMODITY FUTURES T RADING C OMMISSION,
Plaintiff-Appellee,
v.
J AMES A. DONELSON,
Defendant-Appellant.
____________________
Appeal from the United States District Court for the
Northern District of Illinois, Eastern Division.
No. 1:20-cv-03758 — Thomas M. Durkin, Judge.
____________________
A RGUED J ANUARY 18, 2024 — D ECIDED SEPTEMBER 5, 2024
____________________
Before R IPPLE , BRENNAN, and SCUDDER , Circuit Judges.
BRENNAN, Circuit Judge. Long Leaf Trading Group pro-
vided trade recommendations to customers in the commodi-
ties markets and connected them with a merchant who could
execute the trade. The company made money when it re-
ceived commissions on completed trades. When James Donel-
son became CEO, Long Leaf proved itself adept at taking
commissions but inept at making recommendations. The
company collected $1,235,413 in commissions from its

-- 1 of 24 --

2 No. 23-1809
customers who participated in one particular investment pro-
gram. Yet during that same time customers incurred losses to-
taling $2,376,738.
After an investigation, the Commodity Futures Trading
Commission filed a civil enforcement action against the com-
pany, Donelson, and others, charging them with options
fraud and five other violations of commodities laws. The dis-
trict court granted summary judgment to the CFTC against
Donelson on all but one count. Donelson appeals.
I
A
Long Leaf is an introducing broker, which means it is a
middleman between the client and the merchant responsible
for making trades. As a middleman, Long Leaf recommended
trades to its clients. When the client accepted a recommenda-
tion through an associated person (an individual acting as a
salesperson for the firm), Long Leaf would forward the cus-
tomer order to its merchant to be executed. Then Long Leaf
would get a cut.
As Long Leaf was collecting commissions, its trading strat-
egies were not paying off. While James Donelson was CEO of
Long Leaf, the company lost more than $2 million for custom-
ers who participated in one particularly bad strategy: “Time
Means Money” (“TMM”). The TMM program began in June
2015, and 80 percent of Long Leaf’s customers participated.
When Donelson became Long Leaf’s CEO in December 2017,
TMM was still active.
That strategy was impressive for its failure. Throughout
Donelson’s time at Long Leaf, “nearly all Long Leaf patrons
who participated in TMM lost money.” Donelson was not

-- 2 of 24 --

No. 23-1809 3
blind to the program’s performance or to his customers’
frustration with it. He received daily customer account state-
ments; he conducted a “five-month look-back” of the
program, where he learned of the consistent losses; and cus-
tomers called and emailed him with complaints. Donelson
claims he eventually abandoned TMM for a new strategy after
evaluating TMM’s performance. And at his deposition, he dis-
cussed what might be a third strategy—the “gut strangle.”
Whatever strategy it pursued, Long Leaf lost money.
During his tenure as CEO, Donelson (sometimes through
Long Leaf) engaged in acts and omissions that the CFTC
would eventually allege constituted fraud.
1. Long Leaf’s history of losses. Throughout his tenure, Do-
nelson never communicated that Long Leaf was consistently
losing money. His predecessor implemented a policy prohib-
iting Long Leaf’s associated persons from disclosing TMM’s
performance history. If customers asked, the associated per-
sons would not answer, citing the policy. Donelson continued
this policy but had the associated persons redirect the conver-
sations using template responses which did not answer the
customers’ questions. Neither current nor prospective cus-
tomers were ever informed of Long Leaf’s poor track record.
2. Target return rates. Donelson also directed Long Leaf’s
associated persons to tell prospective customers that their
“strategies” targeted a 6 to 12 percent annual return.
3. Trade history & new track record emails. In February 2019,
Donelson sent a prospective customer an email with a “trade
history.” This included a selection of seven of the trades Long
Leaf recommended from August 2018 to January 2019. In ag-
gregate, these trades had an average return of 1.68 percent.

-- 3 of 24 --

4 No. 23-1809
But across all its accounts during that period, Long Leaf cus-
tomers had lost $323,932. Then, in May 2019, Donelson sent
Long Leaf’s associated persons an email with a file called
“track record,” which showed several transactions, reflecting
a return of $460.23 on trades recommended between July 2018
and May 2019. The trades were “real” in that Long Leaf rec-
ommended them and customers asked for them to be exe-
cuted. But no single customer received that exact set of trades.
And again, the full picture was much bleaker: customers lost
$538,618 during that period.
4. Donelson’s experience. In February 2018, Donelson told
customers that he had “ten years of financial and business de-
velopment experience at two of the largest proprietary trad-
ing firms.” Donelson had never worked in trading before. His
only options trading experience was in a single trade in which
he lost $30,000.
B
In June 2020, the CFTC brought a civil enforcement action
against Donelson, Long Leaf, Donelson’s predecessor, and
two of Long Leaf’s associated persons.
The CFTC charged Long Leaf and Donelson with six vio-
lations of the commodities laws. The Commission alleged Do-
nelson was liable for Long Leaf’s conduct as to each violation.
Secondary liability in financial-regulatory enforcement ac-
tions attaches based on how much control a person has over
the principal violator. The CFTC alleged Donelson was liable
for Long Leaf’s conduct because he was one of its controlling
persons.
The charges were:
• Claim I: Options fraud;

-- 4 of 24 --

No. 23-1809 5
• Claim II: Fraud by a commodity trading advisor
(“CTA”);
• Claim III: Fraudulent advertising by a CTA;
• Claim IV: Failure to register as a CTA;
• Claim V: Failure to provide CTA disclosures;
• Claim VI: Failure to register as associated persons.
The CFTC moved for summary judgment against Donel-
son and Long Leaf in November 2021. The district court
agreed with the CFTC on every claim except the part of Claim
VI charging Donelson with secondary liability for allowing a
Long Leaf employee to work as an associated person without
proper registration.1
The district court’s resolution of two claims—options
fraud and failure to register as a CTA—are especially im-
portant for Donelson’s appeal. First, the court agreed with the
CFTC that the four incidents italicized above constituted
fraudulent acts or omissions supporting the options fraud
claim. Donelson had argued that he did not need to disclose
Long Leaf’s past failures, but the court disagreed: “A reason-
able investor would want to know” such a “history of losing
customers’ investments.” Second, the court determined that
1 Donelson addresses Claim VI in his brief but he does not suggest the
district court’s ruling was erroneous. He says, “[l]ate registration is settled
by a small fine” and “[l]ate registration filing could not be a cause of any
loss to a customer either.”
Even assuming this is an argument for error, it fails. The CFTC has
authority to bring a civil action for “any act … constituting a violation of
any provision” of the commodities laws. 7 U.S.C. § 13a-1(a). And it has
discretion to impose civil penalties for those violations. Id. § 13a-1(d)(1).

-- 5 of 24 --

6 No. 23-1809
Long Leaf should have registered as a CTA. Donelson had ar-
gued that Long Leaf was exempt under 17 C.F.R. § 4.14(a)(6).
Relying on an interpretive letter the CFTC published, the
court explained that Long Leaf did not qualify for the exemp-
tion because it was “guiding” its customers’ accounts.
The court ordered Donelson to pay restitution in the
amount of the customers’ losses during his tenure as Long
Leaf’s CEO. This included trading losses as well as commis-
sions—$2,376,738—and disgorgement of the commissions
Long Leaf collected during the same period—$1,235,413 (as
offset by anything Donelson paid toward restitution). Donel-
son is the only one who has appealed.
II
Donelson appeals the grant of summary judgment to the
CFTC on every claim it brought against him except Claim VI,
failure to register as an associated person. We will address the
five claims Donelson asks us to reverse by addressing four in-
terrelated issues. The chart below identifies those issues and
the claims to which they apply.

-- 6 of 24 --

No. 23-1809 7
Options
fraud
CTA
fraud
Fraudu-
lent adver-
tising by a
CTA
Failure to
register as
a CTA
Failure to
make CTA
disclo-
sures
Did Donel-
son’s and
Long Leaf’s
actions con-
stitute
fraud?
X
Is Long Leaf
a CTA?
X X X X
Was Long
Leaf required
to register as
a CTA?
X X
Is Donelson a
controlling
person of
Long Leaf?
X X X X X
First, we will discuss whether Donelson and Long Leaf
committed options fraud. We will focus on the misstatements
the CFTC alleged Donelson and Long Leaf made about Long
Leaf’s investment strategy, performance history, and Donel-
son’s experience. Second, we will examine whether Long Leaf
acted as a CTA by giving advice to its clients on the viability
of trades in commodity options. Third, we will answer the
question whether a regulation exempted Long Leaf from a
rule requiring CTAs to register. We will conclude by

-- 7 of 24 --

8 No. 23-1809
addressing Donelson’s argument that he was not so in control
of Long Leaf that he could be held responsible for its viola-
tions of the commodities laws.
This court reviews an appeal from a summary judgment
order de novo, “construing the record in the light most favor-
able to [the non-movant] and drawing all reasonable infer-
ences in his favor.” James v. Hale, 959 F.3d 307, 314 (7th Cir.
2020).
A
The CFTC charged Donelson and Long Leaf with options
fraud under 7 U.S.C. § 6c(b) and 17 C.F.R. § 33.10.2 To resolve
this claim, we must decide whether Donelson’s and Long
Leaf’s actions constituted fraud.
The Commodity Exchange Act broadly prohibits fraud in
options transactions. A person or business association may
not “enter into or confirm the execution of[] any transaction
involving” an option if that transaction is “contrary to any
rule, regulation, or order of the [CFTC] … .” 7 U.S.C. § 6c(b).
Fraudulent transactions are contrary to CFTC regulations:
It shall be unlawful for any person directly or
indirectly:
(a) To cheat or defraud or attempt to cheat or
defraud any other person;
2 The CFTC also charged Donelson with violating two other antifraud
provisions: fraud by a CTA, 7 U.S.C. § 6o(1), and fraudulent advertising
by a CTA, 17 C.F.R. § 4.41(a). Donelson’s sole argument that the district
court erred by ruling for the CFTC on these claims is that Long Leaf was
not a CTA. We will discuss these two provisions in the next section.

-- 8 of 24 --

No. 23-1809 9
(b) To make or cause to be made to any other
person any false report or statement thereof
or cause to be entered for any person any
false record thereof;
(c) To deceive or attempt to deceive any
other person by any means whatsoever
in or in connection with an offer to enter into,
the entry into, the confirmation of the execution
of, or the maintenance of, any commodity op-
tion transaction.
17 C.F.R. § 33.10.
This court has not determined the elements of fraud under
Rule 33.10, but this is not a complicated task. “Congress in-
tends to incorporate the well-settled meaning of the common-
law terms it uses.” Neder v. U.S., 527 U.S. 1, 23 (1999); Universal
Health Servs., Inc. v. United States ex rel. Escobar, 579 U.S. 176,
187 (2016) (quotations omitted). In other statutes prohibiting
fraud, the Supreme Court has explained that “defraud” takes
the common-law definition of “fraud.” Neder, 527 U.S. at 22–
23 (mail fraud); Escobar, 579 U.S. at 187 (Fales Claims Act).
Common-law fraud has three elements: (1) the making of
a misrepresentation, misleading statement, or a deceptive
omission; (2) scienter; and (3) materiality. See Escobar, 579 U.S.
at 187 (misrepresentations—affirmative and by omission—
were elements of common-law fraud); United States ex rel.
Schutte v. SuperValu Inc., 598 U.S. 739, 750 (2023) (scienter,
same); Neder, 527 U.S. at 22–23 (materiality, same).
Other circuits have applied those elements to Rule
33.10(a). CFTC v. R.J. Fitzgerald & Co., 310 F.3d 1321, 1328 (11th
Cir. 2002); CFTC v. Kratville, 796 F.3d 873, 891, 885 n.8 (8th Cir.

-- 9 of 24 --

10 No. 23-1809
2015); see also SEC v. Lee, 720 F. Supp. 2d 305, 323, 324
(S.D.N.Y. 2010). At least one circuit has also included proof of
reliance in a CEA fraud action, see FDIC v. UMIC, Inc., 136 F.3d
1375, 1384 (10th Cir. 1998), but CFTC brought this action so
reliance is not at issue, Slusser v. CFTC, 210 F.3d 783, 785–86
(7th Cir. 2000) (analyzing Section 6b). The CFTC defends the
markets from misstatements. It need not be misled by a mis-
statement to carry out that duty.
The district court ruled that Long Leaf made four misrep-
resentations or omissions which constituted fraud under Rule
33.10(a). These arose from the four acts or omissions listed
above in section I.A. Donelson argues that none of these are
misrepresentations. We disagree.
“[C]ommon-law fraud has long encompassed certain mis-
representations by omission … .” Escobar, 579 U.S. at 187.
These “half-truths” are “representations that state the truth
only so far as it goes, while omitting critical qualifying infor-
mation … .” Id. at 188. Long Leaf’s four misrepresentations—
the trade history and new track record emails, the targeted
return rate messaging, Long Leaf’s history of losses, and Do-
nelson’s experience—fall into the half-truth bucket.
Trade history & new track record emails. Donelson sent at
least two trade histories to customers (either personally or
through Long Leaf’s associated persons), both of which fea-
tured a list of profitable trades. The most problematic one is
the “new track record” email, which included a set of trades
between July 2018 and May 2019 yielding a 0.63 percent re-
turn. Donelson contends these were demonstrative exhibits of
“what was believed to be the viability of [Long Leaf’s]
model,” meaning they did not misstate anything. He

-- 10 of 24 --

No. 23-1809 11
explained at his deposition that these are the trades consistent
with a new strategy he implemented called the “gut strangle.”
But this was a hypothetical portfolio. Although the trades
actually happened, Long Leaf did not advise a single cus-
tomer to make that set of trades. Donelson testified to this too:
“I don’t believe any one person had every one of those
trades.” At Donelson’s request, Long Leaf’s associated per-
sons sent this “new track record,” with no explanation, to cli-
ents and prospective customers. So the trade history docu-
ments implied—falsely—that Long Leaf achieved those kinds
of returns for its customers on each trade set.
Courts have long frowned upon statements that mislead
by omission. In Escobar, the Supreme Court explained the
common-law fraud approach to half-truth misrepresenta-
tions. 579 U.S. at 188–89. The Court discussed a 1931 New
York Court of Appeals contract law case about an actionable
half-truth. Id. In that case, a real estate seller disclosed two
streets projected to be constructed near the property for sale.
Junius Constr. Co. v. Cohen, 178 N.E. 672, 674 (N.Y. 1931)
(Cardozo, J.). But a third road was projected which, if con-
structed, would bisect the property. Id. The seller’s incom-
plete disclosure “was a tacit representation that the land to be
conveyed was subject to no others.” Id. Thus, it was a fraud.
Here, the trading history documents are like the seller’s
promise that two roads might appear near the property. True
on its face (Long Leaf recommended all the trades), the docu-
ments inaccurately represented that a single Long Leaf cus-
tomer could have held this exact portfolio. As in Junius
Construction, what Long Leaf omitted—this is hypothetical,
this is a projection, and no Long Leaf customer experienced

-- 11 of 24 --

12 No. 23-1809
this return—“materially affect[ed] the value of” Long Leaf’s
communication. Id.
Targeted return rates. Donelson directed the associated per-
sons to tell prospective and current customers that Long
Leaf’s “strategies target 6 to 12 percent returns on an annual
basis.” He argues this was not a misrepresentation because it
is a reasonable projection. Again, we disagree. In the related
Rule 10b-5 context, a projection is misleading if “it was not
made in good faith or was made without a reasonable basis.”
Stransky v. Cummins Engine Co., 51 F.3d 1329, 1333 (7th Cir.
1995). Against the backdrop of Long Leaf’s pattern of losses,
Donelson does not offer a reasonable basis for the projection,
so the district court did not err in finding that the target return
rate statement was misleading.
History of losses. Donelson argues that he was under no
duty to disclose Long Leaf’s history of losing its clients’
money, meaning the district court erred by concluding it was
a misrepresentation.
Donelson may have a point. Omissions or failures to dis-
close “will not support a common law fraud claim.” Nelson v.
Great Lakes Educ. Loan Servs., Inc., 928 F.3d 639, 649 (7th Cir.
2019) (discussing “common law tort of fraud” in Illinois law);
Midwest Com. Banking Co. v. Elkhart City Ctr., 4 F.3d 521, 524
(7th Cir. 1993) (same, but Indiana law); RESTATEMENT
(SECOND) OF T ORTS § 551 (1977) (discussing this rule); see Esco-
bar, 579 U.S. at 188 (stating general principle that failures to
speak are not sufficient for common-law fraud). On the other
hand, these statements “may be actionable as constructive
fraud or fraudulent concealment if the defendant was under
a particular duty to speak … .” Nelson, 928 F.3d at 649;
R ESTATEMENT § 551(2)(a).

-- 12 of 24 --

No. 23-1809 13
There is no genuine dispute of material fact that Donelson
made affirmative misleading statements implying that Long
Leaf’s trading was profitable. The trading history documents,
for example, communicated two half-truths. First, as dis-
cussed, they falsely implied that a customer received that rec-
ommendation. But second, they implied that this period was
a profitable one for Long Leaf. In fact, Long Leaf customers
actually lost $538,618 during that same period.
The targeted return rate projection also communicated
two half-truths. First, as discussed, it falsely implied that Do-
nelson and Long Leaf had a reasonable basis to make the pro-
jection. But second, it implied that the projection was made in
good faith. Long Leaf’s consistently poor track record makes
it hard to believe that anything other than bad faith brought
about that projection.3
3 Donelson challenges the district court’s materiality and scienter de-
terminations only on its decision that he should have disclosed Long
Leaf’s losses to his customers. We have agreed with the district court that
his failure to disclose is grounds for liability through the half-truths this
failure to disclose created based on his other two affirmative misstate-
ments.
Even in this context, Long Leaf’s omitted performance history is ma-
terial and Donelson omitted it with scienter. It was material because it
“‘would have been viewed by the reasonable investor as having signifi-
cantly altered the ‘total mix’ of information made available’”—the trade
history and new track record emails and the targeted return rate messag-
ing materials were only persuasive as indicators of positive upward
movement for Long Leaf’s investment recommendations. Smykla v. Moli-
naroli, 85 F.4th 1228, 1236 (7th Cir. 2023) (quoting TSC Indus., Inc. v. North-
way, Inc., 426 U.S. 438, 449 (1976)) (discussing materiality in the context of
the related Rule 10b-5 context). And Donelson acted with scienter, which
is the “intent to deceive or manipulate,” when he took deliberate steps to
obscure its performance history by directing Long Leaf’s associated

-- 13 of 24 --

14 No. 23-1809
Donelson’s experience. Last, Donelson told customers that
he had “10 years” of experience at major proprietary trading
firms, when he had never before worked in trading and his
only prior experience was limited to a single trade.
This is a half-truth too. It is “critical qualifying infor-
mation” that his experience at the proprietary trading firms
was not in a trading role. Escobar, 579 U.S. at 187. He said in
the letter that he would be working on the trading strategy
and, in testimony, acknowledged that customers wouldn’t
have trusted him if they knew he had no profitable trading
experience.
The district court correctly decided that there is no genu-
ine dispute of material fact that Donelson committed options
fraud.
B
CFTC also charged Long Leaf with four violations of com-
modities regulations that apply only to CTAs: fraud by a
CTA, 7 U.S.C. § 6o(1); fraudulent advertising by a CTA, 17
C.F.R. § 4.41(a); failure to register as a CTA, 7 U.S.C. § 6m(1);
and failure to make the disclosures required of some CTAs,
17 C.F.R. §§ 4.31(a), (b) & 4.36(d)(1). To resolve these four
claims, we must answer the question whether Long Leaf was
a CTA.
A CTA is any person who “for compensation or profit, en-
gages in the business of advising others, either directly or
through publications, writings, or electronic media, as to the
persons not to disclose that history and to redirect questions about it. SEC
v. Jakubowski, 150 F.3d 675, 681 (7th Cir. 1998) (discussing scienter in the
context of the related Rule 10b-5 context); Schutte, 598 U.S. at 750.

-- 14 of 24 --

No. 23-1809 15
value of or the advisability of trading in … any commodity
option authorized under section 6c of this title … .” 7 U.S.C.
§ 1a(12)(A)(i)(III). This includes a publisher of financial rec-
ommendations, even where the publisher’s recommendations
are impersonal. Commodity Trend Serv., Inc. v. CFTC, 149 F.3d
679, 687 (7th Cir. 1998) (“The broad sweep of 7 U.S.C.
§§ 1a(5)(A)4 & 6m(1), by their plain terms, covers imper-
sonal—as well as personalized—advice.”).
The district court correctly found that Long Leaf is a CTA
and was acting as one throughout Donelson’s tenure as CEO.
Donelson’s own words confirm this. As he explains, Long
Leaf “recommended transactions to consumers” as part of its
brokerage activities. Once customers confirmed the recom-
mended trade, Long Leaf forwarded it to a futures commis-
sion merchant and earned a commission. In other words,
Long Leaf advised clients for profit as to the advisability of
trading in commodity options, which fits it within the Act’s
definition. See 7 U.S.C. § 1a(12)(A)(i).
Donelson objects to that finding and cites two out-of-cir-
cuit, district court cases as support for reversing. Per Donel-
son, the first case, Taucher v. Born, 53 F. Supp. 2d 464 (D.D.C.
1999), found that the plaintiffs were not CTAs. Quoting
Taucher, Donelson says they did not “engage in individual
consultations with their customers regarding their standard
advice and recommendations and under no circumstances do
they make trades for their customers,” having instead to go
through “some other licensed broker.”
4 This is the earlier version of the statute defining a CTA. See Commod-
ity Trend Serv., 149 F.3d at 681. The language is the same, as relevant for
this case.

-- 15 of 24 --

16 No. 23-1809
But Taucher holds the opposite: “Each of the plaintiffs in
this case falls squarely within the definition of a CTA. They
engage in the business of advising others, through publica-
tions, writings and electronic media, as to the value of or the
advisability of trading in the futures market, and they do so
for compensation or profit.” Id. at 475. Taucher was an as-ap-
plied First Amendment challenge to the registration require-
ment by publishers of financial advice. Donelson quotes a
part of Taucher explaining that as applied to publishers, the
registration requirement is not a regulation of a profession (al-
lowed) but a regulation of speech (suspect). Id. at 476, 478.
The second case he cites, CFTC v. AVCO Fin. Corp., 979 F.
Supp. 232, 234 (S.D.N.Y. 1997), also does not help him. There,
the court held a company that “promoted itself to actual and
potential customers as providing highly profitable commod-
ity futures trading advice through [a trading] program” was
a CTA. 979 F. Supp. at 236.
Even if Taucher or AVCO drew a line between personal
and impersonal advice, the text of the statute does not. The
CTA definition is “broad”—“broad” enough to include both
types of advice. Commodity Trend Serv., 149 F.3d at 687. Thus,
we agree with the district court that there is no genuine dis-
pute of material fact that Long Leaf was acting as a CTA.
C
Two of these claims—failure to register as a CTA, 7 U.S.C.
§ 6m(1), and failure to make the disclosures required of some
CTAs, 17 C.F.R. §§ 4.31(a), (b) & 4.36(d)(1)—warrant further
discussion.
CTAs cannot “unless registered under this chapter …
make use of the mails or any means or instrumentality of

-- 16 of 24 --

No. 23-1809 17
interstate commerce in connection with [their] business as
such commodity trading advisor … .” 7 U.S.C. § 6m(1). If a
CTA is required to be registered, it must make certain disclo-
sures. See, e.g., 17 C.F.R. § 4.31(a).
Donelson argued before the district court, as he does here,
that Long Leaf was not required to be registered because a
CFTC regulation exempted the company. The CFTC has said
that a CTA registered as an introducing broker need not also
register as a CTA if its “trading advice is solely in connection
with its business as an introducing broker.” Id. § 4.14(a)(6).
Therefore, to resolve Donelson’s appeal of these two claims,
we must determine whether Long Leaf was exempt.
The district court asked whether Long Leaf “‘guid[ed]’
trading in a majority of its customers’ accounts”—if yes, the
exemption would not apply. Because the court found that
Long Leaf was indeed guiding trades in its customers’ ac-
counts, Long Leaf was required to register.
This “guiding trades” language comes not from the text of
Regulation 4.14(a)(6) but from a CFTC comment letter. See
Clarification of Issues Concerning Guided Accounts of Introducing
Brokers and Futures Commission Merchants, CFTC Div. Trading
& Markets Comment Letter, CFTCLTR No. 95-82, 1995 WL
707862 (Sept. 19, 1995). In the letter, the CFTC explained that
a correspondent had asked whether an introducing broker
that “guides” customer accounts is required to register as a
CTA. Id. at *1. By “guided” account, the CFTC and the corre-
spondent meant 1) the customer has orally authorized the in-
troducing broker to initiate a trade, and 2) the introducing
broker “inform[s]” the account “to purchase or sell, the
price …[,] and the number of contracts … .” Id.

-- 17 of 24 --

18 No. 23-1809
The CFTC answered in the negative. Id. Citing “concerns
as to the possible misuse of the exemption,” the agency estab-
lished a line: If the introducing broker is “guiding” trading in
more than half of its accounts, it must register; if not, it might
be required to register. Id. at *2. The comment letter neither
attempts nor achieves a close reading of the text of Regulation
4.14(a)(6). To understand what the regulation means, we un-
dertake that close reading.
We start with a definition of “trading advice”—a recom-
mendation as to the value of a financial transaction for profit.
This is confirmed by the Commodity Exchange Act, which
necessarily defines trading advice while defining “commod-
ity trading advisor”:
The term ‘commodity trading advisor’ means
any person who for compensation or profit, en-
gages in the business of advising others, either
directly or through publications, writings, or
electronic media, as to the value of or the advis-
ability of trading in [a commodity] [or] for com-
pensation or profit, and as part of a regular busi-
ness, issues or promulgates analyses or reports
concerning any of [these] activities … .
7 U.S.C. § 1a(12)(A) (internal subsection organization omit-
ted).
Next, the regulation restricts the type of trading advice an
introducing broker/CTA can give without complying with the
registration requirement. First, it requires that the trading ad-
vice be “in connection with the business as an introducing
broker.” 17 C.F.R. § 4.14(a)(6). Second, it requires that the

-- 18 of 24 --

No. 23-1809 19
trading advice be “solely in connection with” that business. Id.
(emphasis added).
The term “business as an introducing broker” also refers
to the CEA, which defines “introducing broker.” Per the CEA,
an introducing broker is
[A]ny person who is engaged in soliciting or in
accepting orders for
the purchase or sale of any commodity for
future delivery, security futures product, or
swap;
any agreement, contract, or transaction de-
scribed in [other parts of the CEA];
any commodity option authorized under
section 6c of this title; or
any leverage transaction authorized under
section 23 of this title; and
does not accept any money, securities, or prop-
erty (or extend credit in lieu thereof) to margin,
guarantee, or secure any trades or contracts that
result or may result therefrom … .
7 U.S.C. § 1a(31)(A) (internal subsection organization omit-
ted). The function of an “introducing broker”—its business—
is therefore its role as a financial-transaction middleman. It
solicits or accepts buy/sell orders for commodities, id.
§ 1a(31)(A)(i)(I), and does not margin, guarantee, or secure re-
sulting trades or contracts, id. § (31)(A)(i)(II).
The regulation specifies that any trading advice must be
“solely” in connection with this business. “Solely” is an exclu-
sive word, requiring a perfect match between the clause that

-- 19 of 24 --

20 No. 23-1809
precedes it and the word or phrase that it modifies. See
A NTONIN S CALIA & BRIAN A. G ARNER , R EADING LAW: T HE
INTERPRETATION OF L EGAL T EXTS 174–75 (2012). Contrast
“David was solely concerned with defeating Goliath” with
“David was concerned with defeating Goliath.” “Solely” in
the first sentence communicates that David has a single con-
cern: defeating Goliath. Its absence in the second sentence re-
lays a different proposition: David might have many con-
cerns, one of which is defeating Goliath.
Here, “solely” is a qualitative limiter. The “sole” type of
trading advice an IB/CTA may give is advice in connection
with its business as an introducing broker. To state the con-
trapositive, if an IB/CTA gives any advice not in connection
with its business as an introducing broker, it is required to
register. See Udiskey v. Commodity Res. Corp., CFTC No. 98-
R081, at *73–75 (April 2, 1999) (reaching the same conclusion
when interpreting Regulation 4.14(a)(6)). Thus, the exception
does not apply if an introducing broker/CTA gives any advice
not in connection with its role as a solicitor or accepter of
buy/sell orders in commodities.
The CFTC comment letter raises a policy objection to this
broader reading of the exemption. We understand that the
CFTC might have concerns with the scope of their regulation.
The agency has been concerned about the scope of Regulation
4.14(a)(6) since it first drafted the rule. See Udiskey, CFTC No.
98-R081 at *75 n.171. In a Federal Register entry discussing the
rule, the CFTC stated it would monitor the exemptions and
make changes if abuse occurred. Id. The agency has often
cited this practical concern to cabin the exemption—for exam-
ple, “as a basis to hold that, if an IB provides advice too often
or to too many customers, it does not qualify for the

-- 20 of 24 --

No. 23-1809 21
exemption … .” Id. Regulation 4.14(a)(6) could be read to de-
prive the investing public of important disclosures in an area
prone to fraud.
Fortunately, if the CFTC wishes to narrow the exemption,
it can do so without help from us. It is a rule the CFTC wrote,
submitted for notice and comment, and enacted. The Admin-
istrative Procedure Act, 5 U.S.C. § 553, gives the CFTC all the
authority it needs to repeal and replace Regulation 4.14(a)(6).
The record contains much evidence of the trading advice
Long Leaf provided to its brokerage customers through TMM
and its other trading programs. But we have not located any
evidence of Long Leaf providing trading advice to non-bro-
kerage customers. Nor have we located any evidence of Long
Leaf providing advice on matters not within the scope of its
role in soliciting or accepting buy/sell orders for commodities.
In fact, Donelson testified at a deposition that Long Leaf’s in-
troducing brokerage activity was its sole source of business:
There’s two revenue streams, obviously the
trading revenue and then they also had an in-
troducing broker revenue from a European CFD
dealer—contract for difference dealer—from
one of their customers overseas. And that was
in their—in the parlance would be it’s almost
pure profit. They don’t really do anything. They
just receive an introducing broker fee.
We cannot say that, in the light most favorable to Donel-
son, the evidence shows that Long Leaf provided trading ad-
vice outside its business as an introducing broker. But a
deeper review consistent with the text of Regulation 4.14(a)(6)
might reveal new evidence that does so demonstrate. The

-- 21 of 24 --

22 No. 23-1809
district court should have the first chance to decide whether
the exception applies. The CFTC also charged Long Leaf with
failing to make disclosures required of some CTAs. The dis-
trict court should reconsider that claim—Claim V—and
reevaluate whether all CTAs must make those disclosures or
only CTAs required to be registered.
D
The CFTC also charged Donelson with secondary liability
for Long Leaf’s violation of every claim discussed. The final
question for us is whether Donelson was in fact a controlling
person of Long Leaf, such that he could be subject to liability
on its behalf.
Any person who directly or indirectly “controls” a pri-
mary violator of the commodities laws “may be held liable for
such violation in any action brought by [the CFTC] to the
same extent as [the] controlled person.” 7 U.S.C. § 13c(b). The
CFTC has the burden of proof, and must prove the controlling
person “knowingly induced, directly or indirectly, the … acts
constituting the violation” or “did not act in good faith.” Id.
Bad faith includes a reckless failure to “maintain a reasonably
adequate system of internal supervision and control … .”
Monieson v. CFTC, 996 F.2d 852, 860 (7th Cir. 1993). Negli-
gence is insufficient. Id.
Controlling person liability has two prongs. First, the per-
son must have control. Id. at 859. Second, the CFTC must
prove the person acted without good faith or knowingly in-
duced the acts. Id. at 860.
The district court found that Donelson was a controlling
person of Long Leaf. He does not dispute that he had control,
and in fact admitted it before the district court. His sole

-- 22 of 24 --

No. 23-1809 23
argument about his role in the charged conduct is that “[a]t
best, negligence is shown here.”
But the evidence is to the contrary. Donelson knowingly
induced the misstatements supporting the three fraud claims.
Long Leaf did not disclose its losses because Donelson main-
tained an existing policy and established a new one forbid-
ding associated persons to do so. Donelson personally sent
the track record document to the associated persons and
directed them to distribute to their current and prospective
customers. He also provided the associated persons with a
Powerpoint presentation showing the 6 to 12 percent return
projection and directed them to market in soliciting custom-
ers. In each instance the information suggested that Long
Leaf’s recommendations were more profitable than they
were. Long Leaf’s associated persons spread both, with Do-
nelson’s blessing.
Donelson also knowingly induced the CTA claims. He did
not register Long Leaf as a CTA or make the required disclo-
sures. So, his negligence argument does not follow; he inten-
tionally did not register Long Leaf because he believed it was
not a CTA.
We agree with the district court that there is no genuine
dispute of material fact that Donelson was a controlling per-
son of Long Leaf.
III
While at Long Leaf’s helm, Donelson made many deci-
sions that steered the company and himself into regulatory
trouble. Most importantly, he oversaw a dishonest marketing
approach that covered up substantial problems with the com-
pany’s investment-recommendation strategies. The district

-- 23 of 24 --

24 No. 23-1809
court correctly concluded that there was no genuine dispute
of material facts as to the CFTC’s three fraud claims, Long
Leaf’s identity as a CTA, and Donelson’s role as a controlling
person of the company. But it is not clear that Long Leaf was
required to register as a CTA, even though it was one.
For those reasons, we A FFIRM the district court’s entry of
summary judgment in part. As to Claims IV and V, the two
claims dependent on whether Long Leaf was required to reg-
ister as a CTA, we REVERSE the district court’s entry of sum-
mary judgment and R EMAND for proceedings consistent with
this decision.

-- 24 of 24 --