In the
United States Court of Appeals
For the Seventh Circuit
____________________
No. 16‐3017
UNITED STATES OF AMERICA,
Plaintiff‐Appellee,
v.
MICHAEL COSCIA,
Defendant‐Appellant.
____________________
Appeal from the United States District Court for the
Northern District of Illinois, Eastern Division.
No. 1:14‐cr‐00551‐1 — Harry D. Leinenweber, Judge.
____________________
ARGUED NOVEMBER 10, 2016 — DECIDED AUGUST 7, 2017
____________________
Before RIPPLE, MANION, and ROVNER, Circuit Judges.
RIPPLE, Circuit Judge. Today most commodities trading
takes place on digital markets where the participants utilize
computers to execute hyper‐fast trading strategies at speeds,
and in volumes, that far surpass those common in the past.
2 No. 16‐3017
This case involves allegations of spoofing1 and commodities
fraud in this new trading environment. The Government al‐
leged that Michael Coscia commissioned and utilized a com‐
puter program designed to place small and large orders sim‐
ultaneously on opposite sides of the commodities market in
order to create illusory supply and demand and, conse‐
quently, to induce artificial market movement. Mr. Coscia
was charged with violating the anti‐spoofing provision of the
Commodity Exchange Act, 7 U.S.C. §§ 6c(a)(5)(C) and
13(a)(2), and with commodities fraud, 18 U.S.C. § 1348(1). He
was convicted by a jury and later sentenced to thirty‐six
months’ imprisonment.2
Mr. Coscia now appeals.3 He submits that the anti‐spoof‐
ing statute is void for vagueness and, in any event, that the
evidence on that count did not support conviction. With re‐
spect to the commodities fraud violations, he submits that the
Government produced insufficient evidence and that the trial
court applied an incorrect materiality standard. Finally, he
contends that the district court erred in adjudicating his sen‐
tence by adding a fourteen‐point loss enhancement.
We cannot accept these submissions. The anti‐spoofing
provision provides clear notice and does not allow for arbi‐
trary enforcement. Consequently, it is not unconstitutionally
vague. Moreover, Mr. Coscia’s spoofing conviction is sup‐
1 The term “spoofing,” as will be explained in greater detail below, is de‐
fined as “bidding or offering with the intent to cancel the bid or offer be‐
fore execution.” 7 U.S.C. § 6c(a)(5)(C).
2 The district court had jurisdiction over this case under 18 U.S.C. § 3231.
3 We have jurisdiction over this appeal under 28 U.S.C. § 1291.
No. 16‐3017 3
ported by sufficient evidence. With respect to the commodi‐
ties fraud violation, there was more than sufficient evidence
to support the jury’s verdict, and the district court was on
solid ground with respect to its instruction to the jury on ma‐
teriality. Finally, the district court did not err in applying the
fourteen‐point loss enhancement.
I
BACKGROUND
A.
The charges against Mr. Coscia are based on his use of pre‐
programmed algorithms to execute commodities trades in
high‐frequency trading.4 This sort of trading “is a mechanism
for making large volumes of trades in securities and commod‐
ities based on trading decisions effected in fractions of a sec‐
ond.”5 Before proceeding with the particular facts of this case,
we pause to describe the trading environment in which these
actions took place.
4 Mr. Coscia’s opening brief conflates algorithmic trading and high‐
frequency trading. See Appellant’s Br. 5. High‐frequency trading, or HFT,
is perhaps better conceptualized as “a subset of algorithmic trading.” Tara
E. Levens, Comment, Too Fast, Too Frequent? High‐Frequency Trading and
Securities Class Actions, 82 U. Chi. L. Rev. 1511, 1527 (2015).
5 United States v. Aleynikov, 676 F.3d 71, 73 (2d Cir. 2012); see also United
States v. Aleynikov, 737 F. Supp. 2d 173, 175 (S.D.N.Y. 2010) (explaining that
HFT “involves the rapid execution of high volumes of trades in which
trading decisions are made by sophisticated computer programs that use
complex mathematical formulae known as algorithms”); United States v.
Pu, 814 F.3d 818, 821 (7th Cir. 2016) (defining HFT as “the rapid buying
and selling of publicly traded stocks”).
4 No. 16‐3017
The basic process at the core of high‐frequency trading is
fairly straightforward: trading firms use computer software
to execute, at very high speed, large volumes of trades. A
number of legitimate trading strategies can make this practice
very profitable. The simplest approaches take advantage of
the minor discrepancies in the price of a security or commod‐
ity that often emerge across national exchanges. These price
discrepancies allow traders to arbitrage between exchanges
by buying low on one and selling high on another. Because
any such price fluctuations are often very small, significant
profit can be made only on a high volume of transactions.
Moreover, the discrepancies often last a very short period of
time (i.e., fractions of a second); speed in execution is there‐
fore an essential attribute for firms engaged in this business.6
6 The Southern District of New York has noted that “[s]ome commentators
and, at points, the SEC, have stated that HFT firms have a positive effect
on the market by creating significant amounts of liquidity, thereby per‐
mitting the national stock market to operate more efficiently and benefit‐
ting ordinary investors.” In re Barclays Liquidity Cross & High Frequency
Trading Litig., 126 F. Supp. 3d 342, 350 (S.D.N.Y. 2015).
Nonetheless, HFT is not unambiguously good. Rather, some have
sharply criticized the HFT firms’ trading practices. Chief
among their criticisms … is that the HFT firms use the
speed at which they are capable of trading to identify the
trading strategies being pursued by ordinary investors
and react in a manner that forces ordinary investors to
trade at a less advantageous price, with the HFT firm tak‐
ing as profit a portion of the “delta”—that is, the differ‐
ence between the price at which the ordinary investor
would have traded and the price at which it actually
traded as a result of the HFT firm’s actions.
No. 16‐3017 5
Although high‐frequency trading has legal applications, it
also has increased market susceptibility to certain forms of
criminal conduct. Most notably, it has opened the door to
spoofing, which Congress criminalized in 2010 as part of the
Dodd‐Frank Wall Street Reform and Consumer Protection
Act, Pub. L. No. 111‐203, 124 Stat. 1376 (2010). The relevant
provision proscribes “any trading, practice, or conduct that …
is, is of the character of, or is commonly known to the trade
as, ‘spoofing’ (bidding or offering with the intent to cancel the
bid or offer before execution).” 7 U.S.C. § 6c(a)(5).7 For present
purposes, a bid is an order to buy and an offer is an order to
sell.
In practice, spoofing, like legitimate high‐frequency trad‐
ing, utilizes extremely fast trading strategies. It differs from
legitimate trading, however, in that it can be employed to ar‐
tificially move the market price of a stock or commodity up and
down, instead of taking advantage of natural market events
(as in the price arbitrage strategy discussed above). This arti‐
ficial movement is accomplished in a number of ways, alt‐
hough it is most simply realized by placing large and small
orders on opposite sides of the market. The small order is
placed at a desired price, which is either above or below the
current market price, depending on whether the trader wants
to buy or sell. If the trader wants to buy, the price on the small
batch will be lower than the market price; if the trader wants
Id.
7 The provision has almost no legislative history. The only meaningful ref‐
erence reads as follows: “The CFTC requested, and received, enforcement
authority with respect to insider trading, restitution authority, and disrup‐
tive trading practices.” 156 Cong. Rec. S5992 (daily ed. July 15, 2010) (state‐
ment of Sen. Lincoln) (emphasis added).
6 No. 16‐3017
to sell, the price on the small batch will be higher. Large or‐
ders are then placed on the opposite side of the market at
prices designed to shift the market toward the price at which
the small order was listed.
For example, consider an unscrupulous trader who wants
to buy corn futures at $3.00 per bushel in a market where the
current price is $3.05 per bushel. Under the basic laws of sup‐
ply and demand, this trader can drive the price downward by
placing sell orders for large numbers of corn futures on the
market at incrementally decreasing prices (e.g., $3.04, then
$3.03, etc.), until the market appears to be saturated with in‐
dividuals wishing to sell, the price decreases, and, ultimately,
the desired purchase price is reached. In short, the trader
shifts the market downward through the illusion of down‐
ward market movement resulting from a surplus of supply.
Importantly, the large, market‐shifting orders that he places
to create this illusion are ones that he never intends to execute;
if they were executed, our unscrupulous trader would risk ex‐
tremely large amounts of money by selling at suboptimal
prices. Instead, within milliseconds of achieving the desired
downward market effect, he cancels the large orders.
Once our unscrupulous trader has acquired the commod‐
ity or stock at the desired price, he can then sell it at a higher
price than that at which he purchased it by operating the same
scheme in reverse. Specifically, he will place a small sell order
at the desired price and then place large buy orders at increas‐
ingly high prices until the market appears flooded with de‐
mand, the price rises, and the desired value is hit. Returning
to the previous example, if our unscrupulous trader wants to
sell his corn futures (recently purchased at $3.00 per bushel)
for $3.10 per bushel, he will place large buy orders beginning
No. 16‐3017 7
at the market rate ($3.00), quickly increasing that dollar value
(e.g., $3.01, then $3.02, then $3.03, etc.), creating an appear‐
ance of exceedingly high demand for corn futures, which
raises the price, until the desired price is hit. Again, the large
orders will be on the market for incredibly short periods of
time (fractions of a second), although they will often occupy
a large portion of the market in order to efficiently shift the
price.
B.
On October 1, 2014, a grand jury indicted Mr. Coscia for
spoofing and commodities fraud based on his 2011 trading
activity. Prior to trial, he moved to dismiss the indictment, ar‐
guing that the anti‐spoofing provision was unconstitutionally
vague. He further argued that he did not commit commodi‐
ties fraud as a matter of law. The district court rejected both
arguments.
Trial began on October 26, 2015, and lasted seven days.
The testimony presented at trial explained that the relevant
conduct began in August of 2011, lasted about ten weeks, and
followed a very particular pattern. When he wanted to pur‐
chase, Mr. Coscia would begin by placing a small order re‐
questing to trade at a price below the current market price. He
then would place large‐volume orders, known as “quote or‐
ders,”8 on the other side of the market. A small order could be
as small as five futures contracts, whereas a large order would
represent as many as fifty or more futures contracts. At times,
8 Government’s Br. 3.
8 No. 16‐3017
his large orders risked up to $50 million.9 The large orders
were generally placed in increments that quickly approached
the price of the small orders.
Mr. Coscia’s specific activity in trading copper futures
helps to clarify this dynamic. During one round of trading,
Mr. Coscia placed a small sell order at a price of 32755,10
which was, at that time, higher than the current market
price.11 Large orders were then placed on the opposite side of
the market (the buy side) at steadily growing prices, which
started at 32740, then increased to 32745, and increased again
to 32750.12 These buy orders created the illusion of market
movement, swelling the perceived value of any given futures
contract (by fostering the illusion of demand) and allowing
Mr. Coscia to sell his current contracts at the desired price of
32755—a price equilibrium that he created.
9 R.88 at 94 (Tr. 699); R.90 at 66–67 (Tr. 1042–43).
10 As explained at trial:
The tick size for copper futures is one‐half of one‐
thousandth of a cent. So for purposes of the way these
prices are here, the tick size is an increment of five. …
….
… [N]umerical increments of five … represent one
tick, so a five amount increase in the number is one tick
in the copper futures.
R.89 at 63 (Tr. 820). In other words, increments of five represent (at least
for copper futures) one‐half of one‐thousandth of a cent.
11 Id. at 63–65 (Tr. 820–22); R.177‐24.
12 R.177‐24; R.89 at 64–65 (Tr. 821–22).
No. 16‐3017 9
Having sold the five contracts for 32755, Mr. Coscia now
needed to buy the contracts at a lower price in order to make
a profit. Accordingly, he first placed an order to buy five cop‐
per futures contracts for 32750, which was below the price
that he had just created.13 Second, he placed large‐volume or‐
ders on the opposite side of the market (the sell side), which
totaled 184 contracts. These contracts were priced at 32770,
and then 32765, which created downward momentum on the
price of copper futures by fostering the appearance of abun‐
dant supply at incrementally decreasing prices. The desired
devaluation of the contracts was almost immediately
achieved, allowing Mr. Coscia to buy his small orders at the
artificially deflated price of 32750. The large orders were then
immediately cancelled.14 The whole process outlined above
took place in approximately two‐thirds of a second, and was
repeated tens of thousands of times, resulting in over 450,000
large orders, and earning Mr. Coscia $1.4 million. All told, the
trial evidence suggested that this process allowed Mr. Coscia
to buy low and sell high in a market artificially distorted by
his actions.
The Government also introduced evidence regarding
Mr. Coscia’s intent to cancel the large orders prior to their ex‐
ecution. The primary items of evidence in support of this al‐
legation were the two programs that Mr. Coscia had commis‐
sioned to facilitate his trading scheme: Flash Trader and
Quote Trader. The designer of the programs, Jeremiah Park,
testified that Mr. Coscia asked that the programs act “[l]ike a
13 R.177‐24; R.89 at 66–67 (Tr. 823–24).
14 R.177‐24; R.89 at 65–67 (Tr. 822–24).
10 No. 16‐3017
decoy,” which would be “[u]sed to pump [the] market.”15
Park interpreted this direction as a desire to “get a reaction
from the other algorithms.”16 In particular, he noted that the
large‐volume orders were designed specifically to avoid be‐
ing filled and accordingly would be canceled in three partic‐
ular circumstances: (1) based on the passage of time (usually
measured in milliseconds); (2) the partial filling of the large
orders; or (3) complete filling of the small orders.17
A great deal of testimony was presented at trial to support
the contention that Mr. Coscia’s programs functioned within
their intended parameters. For example, John Redman, a di‐
rector of compliance for Intercontinental Exchange, Inc.,18 tes‐
tified that Mr. Coscia
would place a small buy or sell order in the mar‐
ket, and then immediately after that, he would
place a series of much larger opposite orders in
the market, progressively improving price levels
toward the previous order that he placed. That
small initial order would trade, and then the large
order would be canceled and be replaced by a
15 R.86 at 231 (Tr. 498), at 235 (Tr. 502).
16 Id. at 235 (Tr. 502).
17 R.87 at 71–72 (Tr. 577–78).
18 Mr. Coscia used his algorithms on both the Chicago Mercantile Ex‐
change and the Intercontinental Exchange, although he was charged only
for his conduct on the Chicago Mercantile Exchange. Nonetheless, the in‐
dictment also does mention the Intercontinental Exchange trading and a
substantial amount of information related to that trading was offered at
trial.
No. 16‐3017 11
small order, and the large orders in the opposite
direction will have previously taken place.[19]
Redman further testified that Mr. Coscia placed 24,814 large
orders between August and October 2011, although he only
traded on 0.5% of those orders.20 During this same period he
placed 6,782 small orders on the Intercontinental Exchange
and approximately 52% of those orders were filled.21 Mr.
Redman additionally explained that this activity made the
small orders “100 times” more likely to be filled than the
large‐volume orders.22 Mr. Redman made clear that this was
highly unusual:
What we normally see is people placing or‐
ders of roughly the same size most of the time
and, therefore, there aren’t two order sizes in
use with a different cancellation rate between
them. There’s just one order size in use and the
cancellation rate is, there’s just one.[23]
19 R.82 at 254 (Tr. 254).
20 R.86 at 22 (Tr. 289).
21 Id. at 23–24 (Tr. 290–91).
22 Id. at 24 (Tr. 291).
23 Id. at 25 (Tr. 292); see also id. at 85 (Tr. 352) (“Mr. Coscia was the only
person we looked at in this time frame who would put in small orders
with one cancellation rate and big orders with a completely different can‐
cellation rate. That was unusual.”).
12 No. 16‐3017
Finally, Mr. Redman also noted that Mr. Coscia’s order‐to‐fill
ratio (i.e., the average size of the order he showed to the mar‐
ket divided by the average size of the orders filled)24 was ap‐
proximately 1,600%, whereas other traders generally pre‐
sented ratios of between 91% and 264%.25
Other traders testified to the effect of Mr. Coscia’s trading
on their businesses. For example, Anand Twells of Citadel,
LLC, explained that his firm lost $480 in 400 milliseconds as a
result of trading with Mr. Coscia.26 Similarly, Hovannes Der‐
menchyan of Teza Technologies testified that he “lost $10,000
over the course of an hour” of trading with Mr. Coscia.27 Fi‐
nally, Alexander Gerko of XTX Markets described how his
firm “probably lost low hundreds of thousands of dollars” as
a result of Mr. Coscia’s actions.28
The Government also introduced Mr. Coscia’s prior testi‐
mony from a deposition taken by the Commodity Futures
Trading Commission. In that deposition, Mr. Coscia ex‐
plained the logic behind his trading as follows:
The logic is I wanted to make a program
with two sides. I noticed there was more trading
done when one side was larger than the other,
24 See id. at 28 (Tr. 295); see also infra at 26–27.
25 See R.86 at 30–33 (Tr. 297–300).
26 R.88 at 30 (Tr. 635).
27 Id. at 51 (Tr. 656).
28 Id. at 105 (Tr. 710).
No. 16‐3017 13
and I made a program to make a market as tight
as possible with different lopsided markets.
….
I watched the screen, and through watching
the screen for years or weeks, I noticed that
when there was a larger order and smaller or‐
der, a lopsided market, there was more of a ten‐
dency for trading to occur.[29]
When pressed on why he designed the program to cancel
when the large orders risked being filled, without placing
similar parameters on the small orders, Mr. Coscia simply
stated “[t]hat’s just how it was programmed. I don’t give it
much thought beyond that.”30 At trial, Mr. Coscia further tes‐
tified that, “Obviously, there’s less risk there. I thought it was
common sense. But I should have given more of an explana‐
tion.”31 Ultimately, as explained by his counsel in summation,
Mr. Coscia’s defense was that he “placed real orders that were
exactly that, orders that were tradeable.”32
The jury convicted Mr. Coscia on all counts. Mr. Coscia
then filed a motion for acquittal. The district court denied the
motion in a memorandum opinion and order issued on April
6, 2016. The district court determined that the evidence was
sufficient to prove that Mr. Coscia committed commodities
29 R.87 at 52 (Tr. 558).
30 Id. at 61 (Tr. 567).
31 R.89 at 168 (Tr. 925).
32 R.92 at 59 (Tr. 1472).
14 No. 16‐3017
fraud and that his deception was material. Moreover, with re‐
spect to the spoofing charge, the court held that the statute
was not void for vagueness. Finally, the court denied a chal‐
lenge to the definition of materiality provided in the commod‐
ities fraud jury instructions.
Thereafter, the district court, applying a fourteen‐point en‐
hancement for the estimated loss attributable to the illegal ac‐
tions, sentenced Mr. Coscia to thirty‐six months’ imprison‐
ment to be followed by two years’ supervised release.
II
DISCUSSION
A.
We begin with Mr. Coscia’s contention that the anti‐spoof‐
ing provision is unconstitutionally vague. For the conven‐
ience of the reader, we set forth the statutory provision in its
entirety:
(5) Disruptive practices
It shall be unlawful for any person to engage in
any trading, practice, or conduct on or subject to the
rules of a registered entity that—
…
(C) is, is of the character of, or is com‐
monly known to the trade as, “spoofing”
(bidding or offering with the intent to cancel
the bid or offer before execution).
No. 16‐3017 15
7 U.S.C. § 6c(a)(5). The Fifth Amendment’s guarantee that
“[n]o person shall … be deprived of life, liberty, or property,
without due process of law” forbids vague criminal laws. U.S.
Const. amend. V.; Johnson v. United States, 135 S. Ct. 2551, 2556
(2015). This constitutional proscription gives rise to the gen‐
eral rule that “prohibits the government from imposing sanc‐
tions under a criminal law so vague that it fails to give ordi‐
nary people fair notice of the conduct it punishes, or so stand‐
ardless that it invites arbitrary enforcement.” Welch v. United
States, 136 S. Ct. 1257, 1262 (2016) (internal quotation marks
omitted). We review a challenge to a statute’s constitutional‐
ity, including vagueness challenges, de novo. See United States
v. Leach, 639 F.3d 769, 772 (7th Cir. 2011).
1.
Mr. Coscia first submits that the statute gives inadequate
notice of the proscribed conduct. He submits that Congress
did not intend the parenthetical included in the statute to de‐
fine spoofing.33 Mr. Coscia contends that, by “placing ‘spoof‐
ing’ in quotation marks and referring to a ‘commonly known’
definition in the trade, Congress clearly signaled its (mis‐
taken) belief that the definition of ‘spoofing’ had been estab‐
lished in the industry as a term of art.”34 In support of this
argument, he further submits that this statutory structure
33 Appellant’s Br. 40.
34 Id.
16 No. 16‐3017
mirrors the “wash sale” provision of the Commodity Ex‐
change Act35 and that this “parallel approach in statutory
structure strongly suggests that Congress intended for the
‘spoofing’ definition, like the ‘wash sale’ definition, to be es‐
tablished by sources outside the statutory text.”36 We cannot
accept this argument; it overlooks that the anti‐spoofing pro‐
35 7 U.S.C. § 6c(a) provides, in relevant part:
(1) Prohibition
It shall be unlawful for any person to offer to enter into,
enter into, or confirm the execution of a transaction de‐
scribed in paragraph (2) involving the purchase or sale of
any commodity for future delivery (or any option on such
a transaction or option on a commodity) or swap if the
transaction is used or may be used to—
(A) hedge any transaction in interstate commerce
in the commodity or the product or byproduct of the
commodity;
(B) determine the price basis of any such transac‐
tion in interstate commerce in the commodity; or
(C) deliver any such commodity sold, shipped, or
received in interstate commerce for the execution of
the transaction.
(2) Transaction
A transaction referred to in paragraph (1) is a transac‐
tion that—
(A)(i) is, of the character of, or is commonly
known to the trade as, a “wash sale” or “accommoda‐
tion trade”; ….
36 Appellant’s Br. 41.
No. 16‐3017 17
vision, unlike the wash sale provision, contains a parenthe‐
tical definition, rendering any reference to an industry defini‐
tion irrelevant.37
Relying on Chickasaw Nation v. United States, 534 U.S. 84
(2001), Mr. Coscia next submits that the “use of parentheses
emphasizes the fact that that which is within is meant simply
to be illustrative,” id. at 89. The provision at issue in Chickasaw
Nation, a portion of the Indian Gaming Regulatory Act, Pub.
L. No. 100‐497, 102 Stat. 2467 (1988), referred to “[t]he provi‐
sions of Title 26 (including sections 1441, 3402(q), 6041, and
6050I, and chapter 35 of such title).” 25 U.S.C. § 2719(d)(1)
(emphasis added). The anti‐spoofing statute, on the other
hand, reads:
It shall be unlawful for any person to engage
in any trading, practice, or conduct on or subject
to the rules of a registered entity that—
…
(C) is, is of the character of, or is com‐
monly known to the trade as, “spoofing”
(bidding or offering with the intent to cancel
the bid or offer before execution).
37 Compare 7 U.S.C. § 6c(a)(5) (explaining that “any trading, practice, or
conduct … that … is, is of the character of, or is commonly known to the
trade as, ‘spoofing’ (bidding or offering with the intent to cancel the bid or offer
before execution)” is illegal) (emphasis added), with id. § 6c(a)(2)(A)(i) (out‐
lining the wash sale provision, which prohibits any transaction that “is, of
the character of, or is commonly known to the trade as, a ‘wash sale’ or
‘accommodation trade’”).
18 No. 16‐3017
7 U.S.C. § 6c(a)(5). Comparing the statutes, it is clear that, in
the Indian Gaming Regulatory Act, the use of the word “in‐
cluding” rendered the parenthetical illustrative. The anti‐
spoofing provision, however, has no such language and is
thus meaningfully different. The Supreme Court has read par‐
enthetical language like the language before us today as defi‐
nitional instead of illustrative. See, e.g., Lopez v. Gonzales, 549
U.S. 47, 52–53 (2006).38 In any event, this argument does little
to aid Mr. Coscia because, here, the charged conduct clearly
falls within the ambit of the statute regardless whether the
parenthetical is an example or a definition.
In the same vein, Mr. Coscia contends that the lack of a
Commodity Futures Trading Commission regulation defin‐
ing the contours of spoofing adds to his lack of notice. None‐
theless, the Supreme Court has explained that “the touch‐
stone [of a fair warning inquiry] is whether the statute, either
standing alone or as construed, made it reasonably clear at the
relevant time that the defendant’s conduct was criminal.”
United States v. Lanier, 520 U.S. 259, 267 (1997). Consequently,
because the statute clearly defines “spoofing” in the parenthe‐
tical, Mr. Coscia had adequate notice of the prohibited con‐
duct.
Mr. Coscia also makes a broader notice argument. He con‐
tends, in effect, that the absence of any guidance external to
the statutory language—no legislative history, no recognized
38 Cf. Novacor Chems., Inc. v. United States, 171 F.3d 1376, 1381 (Fed. Cir.
1999) (stating that “general principles of construction support the view
that a parenthetical is the definition of the term which it follows”).
No. 16‐3017 19
industry definition, no Commodity Futures Trading Commis‐
sion rule—leaves a person of ordinary intelligence to specu‐
late about the definition Congress intended when it placed
“spoofing” in quotation marks.39 In support of this argument,
Mr. Coscia relies on Upton v. S.E.C., 75 F.3d 92 (2d Cir. 1996).
In that case, the defendant had technically complied with the
requirements of a rule, but the SEC took the position that his
actions nevertheless violated the spirit and purpose of the
rule. Prior to the issuance of an interpretive memorandum ex‐
plaining that position, “[t]he Commission was aware that bro‐
kerage firms were evading the substance of Rule 15c3–3(e).”
Id. at 98. Nonetheless, “[a]part from issuing one consent order
carrying ‘little, if any, precedential weight,’ the Commission
took no steps to advise the public that it believed the practice
was questionable until August 23, 1989, after Upton had al‐
ready stopped the practice.” Id. (internal citation omitted).
Accordingly, “[b]ecause there was substantial uncertainty in
the Commission’s interpretation of Rule 15c3–3(e),” the court
held that “Upton was not on reasonable notice that [his] con‐
duct might violate the Rule.” Id.
The present situation is wholly different from the one in
Upton. Here, Congress enacted the anti‐spoofing provision
specifically to stop spoofing—a term it defined in the statute.
Accordingly, any agency inaction—the issue presented by
Upton—is irrelevant; Congress provided the necessary defini‐
tion and, in doing so, put the trading community on notice.
Lanier, 520 U.S. at 267 (explaining that “the touchstone is
whether the statute, either standing alone or as construed,
39 Appellant’s Br. 43.
20 No. 16‐3017
made it reasonably clear at the relevant time that the defend‐
ant’s conduct was criminal”).
For the same reason, the arguments about a lack of indus‐
try definition or legislative history are irrelevant. The statute
“standing alone” clearly proscribes the conduct; the term
“spoofing” is defined in the statute. Id.40
40 In support of these arguments, Mr. Coscia contends that the district
court’s own interpretation of the anti‐spoofing provision shifted through‐
out the proceedings and thus underscores the provision’s inherent vague‐
ness. The first passage to which he invites our attention is the district
court’s order denying the posttrial motion:
The purpose is clear: to prevent abusive trading practices
that artificially distort the market. That, in turn, only oc‐
curs when there is intent to defraud by placing illusory
offers (or put another way, by placing offers with the in‐
tent to cancel them before execution).
R.124 at 8. The second passage is from the defendant’s sentencing hearing
where the district court noted that defendant “manipulated the market,
that [his trading] caused the market for a specific lot to go up one tick and,
therefore, he was able to sell high.” R.162 at 9.
In context, neither of these passages is troubling. The first quote is
taken from a larger discussion that explains how Congress limited the stat‐
ute to manipulative cancellations:
Coscia had fair notice. It would be unreasonable to
believe that Congress had intended to criminalize all or‐
ders that are eventually cancelled at any point, for any
reason, under 7 U.S.C. § 6c(a)(5)(C). The definition of
spoofing must be read in conjunction with the companion
statutory provision that actually criminalizes the conduct:
[7] U.S.C. § 13(a)(2) prohibits the manipulation or at‐
tempted manipulation of commodity prices generally,
and prohibits knowing violation of the anti‐spoofing rule.
No. 16‐3017 21
2.
Mr. Coscia next contends that, even if the statute gives ad‐
equate notice, the parenthetical definition encourages arbi‐
trary enforcement. He specifically notes that high‐frequency
traders cancel 98% of orders before execution and that there
are simply no “tangible parameters to distinguish
[Mr.] Coscia’s purported intent from that of the other trad‐
ers.”41
This argument does not help Mr. Coscia. The Supreme
Court has made clear that “[a] plaintiff who engages in some
conduct that is clearly proscribed cannot complain of the
vagueness of the law as applied to the conduct of others.”
The purpose is clear: to prevent abusive trading practices
that artificially distort the market. That, in turn, only oc‐
curs when there is intent to defraud by placing illusory
offers (or put another way, by placing offers with the in‐
tent to cancel them before execution).
R.124 at 7–8. In short, the district court’s point here is one that we already
have made: the statute put Mr. Coscia on notice that, when he submitted
offers with the purpose of cancelling them, his actions constituted spoof‐
ing for purposes of 7 U.S.C. § 6c(a)(5)(C), which is part of a larger statutory
scheme to prevent manipulation of the market. As to the second quote,
although a conviction for spoofing does not require any showing of mar‐
ket manipulation, it is clear that the purpose of spoofing is to artificially
skew markets and accordingly make a profit. As a result, describing the
purpose of the anti‐spoofing provision as preventing practices that “arti‐
ficially distort the market” is factually accurate. All told, neither state‐
ment—issued years after the defendant’s actual conduct—suggests the
statute failed to put the defendant on notice as to the illegality of his ac‐
tions.
41 Appellant’s Br. 44–45.
22 No. 16‐3017
Holder v. Humanitarian Law Project, 561 U.S. 1, 18–19 (2010) (al‐
teration in original); see also United States v. Morris, 821 F.3d
877, 879 (7th Cir. 2016) (“Vagueness challenges to statutes that
do not involve First Amendment interests are examined in
light of the facts of the case at hand.”). Rather, the defendant
must prove that his prosecution arose from arbitrary enforce‐
ment. As explained by the Second Circuit, this inquiry “in‐
volve[s] determining whether the conduct at issue falls so
squarely in the core of what is prohibited by the law that there
is no substantial concern about arbitrary enforcement because
no reasonable enforcing officer could doubt the law’s applica‐
tion in the circumstances.” Farrell v. Burke, 449 F.3d 470, 494
(2d Cir. 2006).
Mr. Coscia cannot claim that an impermissibly vague stat‐
ute has resulted in arbitrary enforcement because his conduct
falls well within the provision’s prohibited conduct: he com‐
missioned a program designed to pump or deflate the market
through the use of large orders that were specifically designed
to be cancelled if they ever risked actually being filled. His
program would cancel the large orders (1) after the passage
of time, (2) if the small orders were filled, or (3) if a single large
order was filled. Read together, these parameters clearly indi‐
cate an intent to cancel, which was further supported by his
actual trading record. Accordingly, because Mr. Coscia’s be‐
havior clearly falls within the confines of the conduct prohib‐
ited by the statute, he cannot challenge any allegedly arbitrary
enforcement that could hypothetically be suffered by a theo‐
retical legitimate trader.42
42 Mr. Coscia further contends that we should construe the anti‐spoofing
provision to only apply to orders placed and cancelled during pre‐market
No. 16‐3017 23
Moreover, even if Mr. Coscia could challenge the statute,
we do not believe that it permits arbitrary enforcement. When
we examine the possibility of a statute’s being enforced arbi‐
trarily, we focus on whether the statute “impermissibly dele‐
gates to law enforcement the authority to arrest and prosecute
on ‘an ad hoc and subjective basis.’” Bell v. Keating, 697 F.3d
445, 462 (7th Cir. 2012). In undertaking this inquiry, we have
noted that, “[w]hen the government must prove intent and
knowledge, these requirements … do much to destroy any
force in the argument that application of the [statute] would
be so unfair that it must be held invalid[.]” United States v.
Calimlim, 538 F.3d 706, 711 (7th Cir. 2008) (second, third, and
fourth alterations in original) (internal citations omitted). We
also have underscored “that a statute is not vague simply be‐
cause it requires law enforcement to exercise some degree of
judgment.” Bell, 697 F.3d at 462.
The text of the anti‐spoofing provision requires that an in‐
dividual place orders with “the intent to cancel the bid or offer
before execution.” 7 U.S.C. § 6c(a)(5)(C). This phrase imposes
clear restrictions on whom a prosecutor can charge with
spoofing; prosecutors can charge only a person whom they
believe a jury will find possessed the requisite specific intent
to cancel orders at the time they were placed. Criminal pros‐
ecution is thus limited to the pool of traders who exhibit the
requisite criminal intent. This provision certainly does not
“vest[] virtually complete discretion in the hands of the po‐
lice.” Gresham v. Peterson, 225 F.3d 899, 907 (7th Cir. 2000) (in‐
ternal quotation marks omitted).
hours. We simply find no support for this argument in the statute’s plain
language, which is broad and unrestrained by any temporal limitations.
24 No. 16‐3017
Importantly, the anti‐spoofing statute’s intent require‐
ment renders spoofing meaningfully different from legal
trades such as “stop‐loss orders” (“an order to sell a security
once it reaches a certain price”)43 or “fill‐or‐kill orders” (“an
order that must be executed in full immediately, or the entire
order is cancelled”)44 because those orders are designed to be
executed upon the arrival of certain subsequent events. Spoof‐
ing, on the other hand, requires, an intent to cancel the order
at the time it was placed.45 The fundamental difference is that
legal trades are cancelled only following a condition subse‐
quent to placing the order, whereas orders placed in a spoof‐
ing scheme are never intended to be filled at all.
At bottom, Mr. Coscia’s vagueness challenge fails. The
statute clearly defines the term spoofing, providing sufficient
notice. Moreover, Mr. Coscia’s actions fall well within the core
of the anti‐spoofing provision’s prohibited conduct, preclud‐
ing any claim that he was subject to arbitrary enforcement.
Furthermore, even if his behavior were not well within the
core of the anti‐spoofing provision’s prohibited conduct, the
statute’s intent requirement clearly suggests that the statute
does not allow for ad hoc or subjective prosecution.
43 Government’s Br. 36.
44 Id.
45 Mr. Coscia’s contention that “the Government perceives a distinction
between orders placed with intent to fill under certain circumstances and
those placed with intent to cancel under certain circumstances” is thus
wholly inaccurate. Reply Br. 19 (emphasis in original). Mr. Coscia did not
place orders with the intent to cancel under certain circumstances—he
placed orders with the present intent to always cancel the large orders. His
purpose was not to trade on those orders, but rather to use them to shift
the market up or down.
No. 16‐3017 25
B.
Having determined that the anti‐spoofing provision is not
void for vagueness, we next address Mr. Coscia’s contention
that the evidence of record does not support his spoofing con‐
viction. “In reviewing a challenge to the sufficiency of the ev‐
idence, we view all the evidence and draw all reasonable in‐
ferences in the light most favorable to the prosecution and up‐
hold the verdict if any rational trier of fact could have found
the essential elements of the crime beyond a reasonable
doubt.” United States v. Khattab, 536 F.3d 765, 769 (7th Cir.
2008) (internal quotation marks omitted). “[We] will not …
weigh the evidence or second‐guess the jury’s credibility de‐
terminations.” United States v. Stevens, 453 F.3d 963, 965 (7th
Cir. 2006) (citation omitted). Recognizing that “it is usually
difficult or impossible to provide direct evidence of a defend‐
ant’s mental state,” we allow for criminal intent to be proven
through circumstantial evidence. United States v. Morris, 576
F.3d 661, 674 (7th Cir. 2009).
As we have noted earlier, a conviction for spoofing re‐
quires that the prosecution prove beyond a reasonable doubt
that Mr. Coscia knowingly entered bids or offers with the pre‐
sent intent to cancel the bid or offer prior to execution.
Mr. Coscia’s trading history clearly indicates that he cancelled
the vast majority of his large orders. Accordingly, the only is‐
sue is whether a rational trier of fact could have found that
Mr. Coscia possessed an intent to cancel the large orders at
the time he placed them.
A review of the trial evidence reveals the following. First,
Mr. Coscia’s cancellations represented 96% of all Brent fu‐
tures cancellations on the Intercontinental Exchange during
26 No. 16‐3017
the two‐month period in which he employed his software.46
Second, on the Chicago Mercantile Exchange, 35.61% of his
small orders were filled, whereas only 0.08% of his large or‐
ders were filled.47 Similarly, only 0.5% of his large orders were
filled on the Intercontinental Exchange.48 Third, the designer
of the programs, Jeremiah Park, testified that the programs
were designed to avoid large orders being filled.49 Fourth,
Park further testified that the “quote orders” were “[u]sed to
pump [the] market,” suggesting that they were designed to
inflate prices through illusory orders.50 Fifth, according to one
study, only 0.57% of Coscia’s large orders were on the market
for more than one second, whereas 65% of large orders en‐
tered by other high‐frequency traders were open for more
than a second.51 Finally, Mathew Evans, the senior vice presi‐
dent of NERA Economic Consulting, testified that Coscia’s or‐
der‐to‐trade ratio was 1,592%, whereas the order‐to‐trade ra‐
tio for other market participants ranged from 91% to 264%.52
As explained at trial, these figures “mean[] that Michael
Coscia’s average order [was] much larger than his average
trade”—i.e., it further suggests that the large orders were
46 R.86 at 41 (Tr. 308).
47 Id. at 127 (Tr. 394).
48 Id. at 22 (Tr. 289).
49 See id. at 198 (Tr. 465), at 231–32 (Tr. 498–99).
50 Id. at 235 (Tr. 502).
51 R.91 at 35–36 (Tr. 1281–82).
52 Id. at 41 (Tr. 1287).
No. 16‐3017 27
placed, not with the intent to actually consummate the trans‐
action, but rather to shift the market toward the artificial price
at which the small orders were ultimately traded.53
We believe that, given this evidence, a rational trier of fact
easily could have found that, at the time he placed his orders,
Mr. Coscia had the “intent to cancel before execution.” As in
all cases based upon circumstantial evidence, no single piece
of evidence necessarily establishes spoofing. Nonetheless,
when evaluated in its totality, the cumulative evidence cer‐
tainly allowed a rational trier of fact to determine that
Mr. Coscia entered his orders with the intent to cancel them
before their execution.
C.
Mr. Coscia also challenges his conviction for commodities
fraud under 18 U.S.C. § 1348(1). This statute makes it a crime
“to defraud any person in connection with any commodity for
future delivery.” Id. The elements54 of this crime are (1) fraud‐
ulent intent, (2) a scheme or artifice to defraud, and (3) a nexus
53 Id.
54 Mr. Coscia proposes a different formulation of these elements, stating
that “there must be (a) proof of deceptive conduct, and (b) proof that the
deception is ‘material.’” Appellant’s Br. 26. Nonetheless, the case that he
cites in support of this formulation actually employs the more widely ac‐
cepted formulation that we have articulated above. See United States v. Hat‐
field, 724 F. Supp. 2d 321, 324 (E.D.N.Y. 2010) (“Under § 1348(1), the Gov‐
ernment must provide sufficient evidence to establish that Mr. Brooks had
(1) ‘fraudulent intent’; (2) ‘a scheme or artifice to defraud’; and (3) ‘a nexus
with a security.’”).
28 No. 16‐3017
with a security.55 United States v. Mahaffy, 693 F.3d 113, 125 (2d
Cir. 2012) (citing United States v. Motz, 652 F. Supp. 2d 284, 294
(E.D.N.Y. 2009)). “False representations or material omissions
are not required” for conviction under this provision. Id.
Mr. Coscia contends that the jury could not reasonably
have found that he had a fraudulent intent because his con‐
duct was not fraudulent as a matter of law. He also contends
that the court applied an incorrect materiality standard. We
now turn to an examination of each of these submissions.
1.
We first address Mr. Coscia’s view that the jury’s finding
of fraudulent intent was not supported by the evidence be‐
cause his conduct was, as a matter of law, not deceptive. In
reviewing challenges to the sufficiency of the evidence, we
“uphold the verdict if any rational trier of fact could have
found the essential elements of the crime beyond a reasonable
doubt.” Khattab, 536 F.3d at 769 (internal quotation marks
omitted).
Mr. Coscia contends that because “his orders were fully
executable and subject to legitimate market risk,” they were
not, as a matter of law, fraudulent.56 In particular, he main‐
tains that his “orders were left open in the market long
enough that other traders could—and often did—trade
55 The parties do not contest the presence of this element.
56 Appellant’s Br. 27.
No. 16‐3017 29
against them, leading to thousands of completed transac‐
tions.”57 He accordingly concludes that his “orders were not
fraudulent or ‘illusory’ as a matter of law.”58
We cannot accept this argument. At bottom, Mr. Coscia
“confuses illusory orders with an illusion of market move‐
ment.”59 The evidence of record supports the conclusion that
Mr. Coscia designed a scheme to pump and deflate the mar‐
ket through the placement of large orders. His scheme was
deceitful because, at the time he placed the large orders, he
intended to cancel the orders. As the district court correctly
noted, Mr. Coscia’s argument “ignores the substantial evi‐
dence suggesting that [he] never intended to fill [his] large or‐
ders and thus sought to manipulate the market for his own
financial gain.”60
The evidence supporting the existence of a fraudulent in‐
tent is substantial. Jeremiah Park, who designed the computer
program at Mr. Coscia’s behest, explained that the objective
of the computer program was “to pump [the] market”61 and
act “[l]ike a decoy.”62 It was intended to create the illusion of
market movement. With Park, Mr. Coscia designed a system
that used large orders to inflate or deflate prices, while also
57 Id. at 27–28.
58 Id. at 28.
59 Government’s Br. 43 (emphasis in original).
60 R.124 at 4.
61 See R.86 at 235 (Tr. 502).
62 Id. at 231 (Tr. 498).
30 No. 16‐3017
structuring that system to avoid the filling of large orders. The spe‐
cific parameters of Mr. Coscia’s programs, which were de‐
signed to cancel orders (1) based on the passage of time (usu‐
ally measured in milliseconds), (2) following the partial filling
of the large orders, or (3) following the complete filling of the
small orders, suggests, strongly, fraudulent intent. The pro‐
grams facilitated the consummation of small orders and ac‐
tively avoided the completion of large orders.63 That 0.08% of
his large Chicago Mercantile Exchange orders were filled
does not make his scheme to shift artificially the market any
less fraudulent.64
63 R.87 at 72 (Tr. 578).
64 See R.86 at 127 (Tr. 394). Mr. Coscia additionally invites our attention to
United States v. Radley, 659 F. Supp. 2d 803 (S.D. Tex. 2009), and CP Stone
Fort Holdings, LLC v. Doe(s), No. 16 C 4991, 2016 WL 5934096 (N.D. Ill. Oct.
11, 2016). Both are inapposite.
Radley involved a prosecution under 7 U.S.C. § 13(a)(2), which prohib‐
its price manipulation and cornering of commodities in interstate com‐
merce. In that case, the defendants were charged with conspiring to ma‐
nipulate the price of TET propane by misleading “the market about the
true supply of … TET propane.” 659 F. Supp. 2d at 807. Ultimately, the
court held that “even if [the bids] were higher than any others, [they] were
actually bids, and when they were accepted, defendants actually went
through with the transactions.” Id. at 815. Accordingly, “[s]ince defend‐
ants were willing and able to follow through on all of the bids, they were
not misleading.” Id. CP Stone Fort Holdings, LLC similarly rejected a theory
that the defendants’ orders could have “creat[ed] the false appearance of
… a change in the supply and demand for the securities[]” in light of the
fact that “all of the offers or bids were legitimate and could have been
matched at any time by a willing participant placing an aggressive order.”
CP Stone Fort Holdings, LLC, 2016 WL 5934096, at *6.
No. 16‐3017 31
The evidence contrasting Mr. Coscia’s trading patterns
and those of legitimate traders was striking and also supports
the jury’s conclusion of fraudulent intent. For example, John
Redman, the director of compliance for Intercontinental Ex‐
change, testified that Mr. Coscia
would place a small buy or sell order in the mar‐
ket, and then immediately after that, he would
place a series of much larger opposite orders in
the market, progressively improving price lev‐
els toward the previous order that he placed.
That small initial order would trade, and then
the large order would be canceled and be re‐
placed by a small order, and the large orders in
the opposite direction will have previously
taken place.[65]
Neither case provides an apt analogy. Neither of these cases involved,
as did this case, the development of a specific program to create the illu‐
sion of artificial market movement that included the use of large orders to
inflate the price while also taking steps to avoid transactions in the large
orders. Indeed, in Radley, the court specifically noted that the alleged facts
fell “short of alleging an artificial price because none of these bidding tac‐
tics is anything other than legitimate forces of supply and demand.”
Radley, 659 F. Supp. 2d at 815 (emphasis in original). Similarly, CP Stone
Fort Holdings rejected a theory of securities fraud rooted in the proposition
that, “if a subset of orders was ultimately cancelled, those orders, in hind‐
sight, must never have been intended to be executed.” 2016 WL 5934096,
at *6 (internal quotation marks omitted). Here, however, Mr. Coscia arti‐
ficially moved the market by cancelling all but 0.08% of his large Chicago
Mercantile Exchange orders. R.86 at 127 (Tr. 394).
65 R.82 at 254 (Tr. 254).
32 No. 16‐3017
Mr. Redman made clear that this was highly unusual,66 spe‐
cifically explaining that
What we normally see is people placing or‐
ders of roughly the same size most of the time
and, therefore, there aren’t two order sizes in
use with a different cancellation rate between
them. There’s just one order size in use and the
cancellation rate is, there’s just one.[67]
Similar evidence was presented regarding Mr. Coscia’s trad‐
ing on the Chicago Mercantile Exchange, where 35.61% of his
small orders were filled, whereas only 0.08% of his large or‐
ders were filled. In other words, Mr. Coscia’s trading patterns
clearly indicated a desire to use the large orders as a means of
shifting the market equilibrium toward his desired price,
while avoiding the actual completion of those large transac‐
tions.
2.
Mr. Coscia also submits that the district court applied an
incorrect standard of materiality when it instructed the jury
that the alleged wrongdoing had to be “capable of influencing
the decision of the person to whom it is addressed.”68 In his
66 R.86 at 85 (Tr. 352) (“Mr. Coscia was the only person we looked at in
this time frame who would put in small orders with one cancellation rate
and big orders with a completely different cancellation rate. That was un‐
usual.”).
67 Id. at 25 (Tr. 292).
68 R.92 at 177 (Tr. 1590).
No. 16‐3017 33
view, the district court should have told the jury that the al‐
leged scheme had to be “reasonably calculated to deceive per‐
sons of ordinary prudence” and that “there is a substantial
likelihood that a reasonable investor [or trader] would con‐
sider [the deceptive conduct] important in making a deci‐
sion.”69
We review challenges to jury instructions de novo. United
States v. Marr, 760 F.3d 733, 743 (7th Cir. 2014). Nevertheless,
“[t]he district court is afforded substantial discretion with re‐
spect to the precise wording of instructions so long as the final
result, read as a whole, completely and correctly states the
law.” Id. (internal quotation marks omitted). “We reverse only
if the instructions as a whole do not correctly inform the jury
of the applicable law and the jury is misled.” Id.
Our circuit does not have a specific pattern jury instruc‐
tion for commodities fraud. The district court therefore
adopted the jury instruction in our pattern jury instructions
for mail, wire, and carrier fraud.70 This was an entirely rea‐
sonable decision. District courts often must craft instructions
69 Appellant’s Br. 30 (alterations in original) (internal quotation marks
omitted).
70 R.124 at 9–10. The instructions at trial read, in relevant part, as follows:
Counts One through Six of the indictment charge
Mr. Coscia with commodities fraud.
In order for you to find Mr. Coscia guilty of this
charge, the government must prove each of the four fol‐
lowing elements beyond a reasonable doubt:
1. there was a scheme to defraud any person as
charged in the indictment; and
34 No. 16‐3017
for areas of law for which there is no pattern jury instruction.
In such situations, borrowing from the jury instructions gov‐
erning analogous areas of law is entirely appropriate. See Chi‐
cago Coll. of Osteopathic Med. v. George A. Fuller Co., 719 F.2d
1335, 1345 (7th Cir. 1983) (approving a jury instruction for the
standard of care owed by architects based on the pattern jury
instructions outlining the standard of care owed by physi‐
cians). Because section 1348 was modeled on the federal mail
and wire fraud statutes, the district court certainly was on
solid ground in looking to the pattern jury instruction for
those offenses. See United States v. Wey, No. 15‐CR‐611 (AJN),
2017 WL 237651, at *9 n.6 (S.D.N.Y. Jan. 18, 2017) (“Several
courts have recognized that ‘because the text and legislative
history of 18 U.S.C. § 1348 clearly establish that it was mod‐
2. Mr. Coscia knowingly executed the scheme; and
3. Mr. Coscia acted with the intent to defraud; and
4. the scheme was in connection with any commod‐
ity for future delivery.
….
A scheme to defraud any person means a plan or
course of action intended to deceive or cheat another. A
scheme to defraud need not involve any false statement
or misrepresentation of fact. A scheme to defraud must be
material, which means it is capable of influencing the decision
of the person to whom it is addressed.
R.85 at 20–21 (emphasis added); see also R.92 at 169–70 (Tr. 1582–
83).
No. 16‐3017 35
eled on the mail and wire fraud statutes,’ an analysis of Sec‐
tion 1348 ‘should be guided by the caselaw construing those
statutes.’”).71
Moreover, Mr. Coscia’s conduct certainly was material
even under his own formulation of materiality.72 The evi‐
dence at trial showed that his course of action was not only
“reasonably calculated to deceive” but also that actual inves‐
tors did find his actions “important in making a decision.” Jer‐
emiah Park clearly related that Mr. Coscia had expressed a
desire to “pump” the market, and thus deceive market partic‐
ipants by creating illusory depth, satisfying the first of his
new definitions. Moreover, market participants testified that
71 See also United States v. Motz, 652 F. Supp. 2d 284, 296 (E.D.N.Y. 2009)
(explaining that “the text and legislative history of 18 U.S.C. § 1348 clearly
establish that it was modeled on the mail and wire fraud statutes, [and
that] the Court’s analysis should be guided by the caselaw construing
those statutes”) (internal quotations omitted); United States v. Mahaffy, No.
05‐CR‐613, 2006 WL 2224518, at *11 (E.D.N.Y. Aug. 2, 2006) (explaining
that “the text and legislative history of 18 U.S.C. § 1348 clearly establish
that it was modeled on the mail and wire fraud statutes”).
72 We are unpersuaded by the Government’s contention that this line of
argument is waived. Although Mr. Coscia initially proposed a jury in‐
struction similar to that adopted by the district judge, R.59 at 4–6, he pre‐
served his objection by later seeking to amend that instruction, R.74. Cf.
Wilson v. Kelkhoff, 86 F.3d 1438, 1442 (7th Cir. 1996) (“A party waives an
argument on appeal if that argument related to a jury instruction and he
failed to object to the relevant jury instruction below.”); United States v.
DiSantis, 565 F.3d 354, 361 (7th Cir. 2009) (“The ‘touchstone’ of the waiver
inquiry is ‘whether and to what extent the defendant ha[s] actually ap‐
proved of the jury instructions assigned as error on appeal.’”) (alteration
in original).
36 No. 16‐3017
(1) large orders induced firms to fill small orders,73 (2) algo‐
rithms were tricked by large orders, creating the illusion of an
oversaturated market,74 and (3) Mr. Coscia’s actions even in‐
duced certain traders to leave the market altogether.75 In sum,
Mr. Coscia’s actions were material regardless of whether we
apply his standard or the district court’s.
In this respect, Mr. Coscia’s invocation of United States v.
Finnerty, 474 F. Supp. 2d 530 (S.D.N.Y. 2007), aff’d, 553 F.3d
143 (2d Cir. 2008), and Sullivan & Long, Inc. v. Scattered Corp.,
47 F.3d 857 (7th Cir. 1995), is unpersuasive. Mr. Coscia be‐
lieves that “the core principle arising from these decisions” is
that “there can be no fraud where the underlying conduct is
not contrary to reasonable expectations.”76 Although a trader
may not have expected any given trade to remain on the mar‐
ket for any particular period of time,77 no trader expected a
73 R.88 at 31 (Tr. 636).
74 Id. at 44–50 (Tr. 649–55).
75 Id. at 59 (Tr. 664), at 105 (Tr. 710), at 137 (Tr. 742).
76 Reply Br. 13.
77 The cases more readily stand for the unremarkable rule that fraud re‐
quires deception. See United States v. Finnerty, 474 F. Supp. 2d 530, 542
(S.D.N.Y. 2007) (holding that the Government “failed to show that inter‐
positioning constituted a deceptive act within the meaning of the federal
securities laws because it did not provide proof of customer expecta‐
tions”); Sullivan & Long, Inc. v. Scattered Corp., 47 F.3d 857, 864 (7th Cir.
1995) (explaining that “there was no deception”).
No. 16‐3017 37
complex, concerted effort not only to pump the market but
also to create a totally non‐existent market.78
Mr. Coscia’s arguments related to “fill‐or‐kill orders” and
“iceberg orders” are also unpersuasive. Fill‐or‐kill orders,
“which are programmed to cancel if not filled immediately,”79
and iceberg orders, which “are designed to obscure the true
extent of supply or demand that lurks beneath the order
book,”80 are both different from the instant conduct because
they are designed to be executed under certain conditions,
whereas Mr. Coscia’s large orders were designed to evade ex‐
ecution.
D.
We address now Mr. Coscia’s argument that the district
court erred in applying a fourteen‐point loss enhancement.
“We review a district court’s interpretation and application of
the guidelines de novo and its findings of fact for clear error.”
United States v. White, 737 F.3d 1121, 1139 (7th Cir. 2013).
Mr. Coscia urges that the district court erroneously em‐
ployed his gain as a measure of loss in determining his sen‐
tence. It is clear that the defendant’s gain may be substituted
for loss if there were losses that cannot reasonably be calcu‐
78 See, e.g., R.88 at 50 (Tr. 655) (explaining the import of actual supply and
demand in accurately pricing commodities).
79 Appellant’s Br. 6.
80 Id. at 8. Iceberg orders accomplish their goal of obscuring supply and
demand by segmenting large orders into smaller orders. Michael Durbin,
All About High‐Frequency Trading 54–55 (2010).
38 No. 16‐3017
lated. See U.S.S.G. § 2B1.1 cmt. n. 3(B); cf. United States v. An‐
dersen, 45 F.3d 217, 221 (7th Cir. 1995) (“Generally the defend‐
ant’s gain may provide a reasonable approximation of a vic‐
tim’s loss, and may be used when more precise means of
measuring loss are unavailable. The Application Notes …
specifically allow the defendants’ gain to be used as a basis
for calculating an approximate loss when evidence of the ex‐
act amount of loss is not available.”) (interpreting predecessor
guideline, § 2F1.1). Nonetheless, we will not substitute gain
as a proxy for loss where there is “no means of determining
whether [the defendant’s] gain is a reasonable estimate of [the
victim’s] loss.” United States v. Vitek Supply Corp., 144 F.3d 476,
490 (7th Cir. 1998) (interpreting predecessor guideline,
§ 2F1.1).
After reviewing the record, we are satisfied that the dis‐
trict court did not err in applying the loss enhancement. The
nature of Mr. Coscia’s trading made determining the when,
where, and with whom of his transactions almost impossible;
using his programs, he executed thousands of trades over a
ten‐week period with innumerable counterparties. In such sit‐
uations, where the loss is not easily ascertainable, we have
held that “probable loss” “is ‘loss’ within the meaning of the
guideline.” United States v. Vrdolyak, 593 F.3d 676, 681 (7th Cir.
2010) (emphasis removed).
Applying this rule, the testimony presented at trial sup‐
ports a finding of probable loss. Some particular losses were
documented and before the court. Twells testified that he lost
$480 on trades with Mr. Coscia;81 Dermenchyan stated that he
81 R.88 at 30 (Tr. 635).
No. 16‐3017 39
“lost $10,000 over the course of an hour;”82 Gerko stated that
“we probably lost low hundreds of thousands of dollars.”83
Applying our deferential standard of review, we find that the
district court did not err in finding loss.
The district court also was correct in concluding that all
losses could not be calculated reasonably. Mr. Coscia’s
scheme was complex, involving thousands of anonymous
trades executed across multiple exchanges with numerous
counterparties. Consequently, the hours of labor required to
collect, collate, and analyze the relevant trading logs would
have imposed an insurmountable logistical burden on the
prosecution. This case exemplifies the type of logistical bur‐
dens the gain‐for‐loss approach was designed to alleviate.
The district court therefore did not err in concluding that sub‐
stituting gain for loss was reasonable. Mr. Coscia made
money by artificially inflating and deflating prices. Every
time he did so, he inflicted a loss.84
82 Id. at 51 (Tr. 656).
83 Id. at 105 (Tr. 710).
84 We recognize that these traders did not independently testify as to the
identity of the counterparty in each of their losing transactions or the iden‐
tity of the spoofer; indeed, the anonymous nature of commodities trading
would have prevented them from reasonably doing so. Nonetheless, there
was substantial support establishing a connection between Mr. Coscia’s
trades and the losses suffered by other market participants.
First, the parties stipulated to the user identities employed by
Mr. Coscia and the traders who worked for him. See R.86 at 88 (Tr. 355).
These user identities were then used to collect relevant trading data and
create summary charts. See, e.g., id. at 114 (Tr. 381) (“This chart represents
40 No. 16‐3017
Mr. Coscia disagrees. In his view, the district court “fun‐
damentally misapphrend[ed] the nature of futures markets”
and unrealistically viewed the commodities market as “‘zero
sum.’”85 He proceeds along this line of argument in three
parts. First, he notes that the ultimate gain or loss enjoyed by
a trader can be evaluated only after that commodity has been
both purchased and sold. He then highlights that participants
in futures markets established hedge positions and that “par‐
ties submit their orders not to individual counterparties but
to the entire market simultaneously.”86 Ultimately, he con‐
tends that “the District Court’s decision to apply a 14‐point
loss enhancement at sentencing was predicated on erroneous
various summary statistics surrounding a large order entry fill and can‐
cellations engaged by various Panther Tag 50s.”); see also R.86 at 91–92 (Tr.
358–59), R.89 at 19–69 (Tr. 776–826). The summary charts, associated data,
and derivative charts were in turn used to establish Mr. Coscia’s use of
large orders to shift the market and, thus, the losses suffered by the other
market participants. See, e.g., R.88 at 28–29 (Tr. 633–34) (testimony of Mr.
Twells); id. at 102–06 (Tr. 707–11) (testimony of Mr. Gerko).
At bottom, the Government identified Mr. Coscia’s user identities,
and collected trading records related to those user identities, which
showed the use of large orders to shift the market. The counterparties (i.e.,
the victims) then confirmed, based on their own records and recollections,
that they had been involved in those trades and suffered a loss. Nothing
else was required because any trade executed in Mr. Coscia’s artificial
market involved a transaction at a skewed price—i.e., any party trading
on the opposite side of the market from his small orders necessarily lost
money even though it was impossible to say with any accuracy how much
money.
85 Appellant’s Br. 52.
86 Id. at 53.
No. 16‐3017 41
findings concerning the reasonableness of using Coscia’s al‐
leged US $1.4 million gain as a proxy for losses and the proof
of loss adduced at trial.”87
We do not think that Mr. Coscia’s arguments rebut ade‐
quately the proposition that, in the environment of high speed
trading, gain is a reasonable proxy for loss. Although a single
trade cannot be viewed in isolation, the fact remains that a
loss resulting from a trade with Mr. Coscia could not be
purged entirely by a profit on any subsequent sale, even
where the latter sale resulted in a net profit. That profit neces‐
sarily would be less than the proceeds earned in a series of
transactions absent Mr. Coscia’s artificial prices.
We also believe that Mr. Coscia’s contention that gains or
losses must be evaluated in relation to hedge positions in cash
markets does not survive scrutiny. In particular, it seems to
suggest that a loss in the futures markets may not actually be
a loss due to positions in cash markets designed to set off any
such financial hardship. This theory essentially would ab‐
solve Mr. Coscia from the damage he inflicted on the market
and on those with whom he traded simply because at least
some victims had taken steps to insure themselves and their
clients. The fact remains that Mr. Coscia’s illegal actions
caused damage. His victims’ prudence in attempting to miti‐
gate such a loss does not require that the law ignore the initial
damage caused by his actions.
Nor does it make a difference that orders initially were
made to the market as a whole. The reality remains that his
87 Id. at 51.
42 No. 16‐3017
trades injured those who traded with him; these parties were
always harmed by the artificial shift in market price.
Finally, we note that Mr. Coscia’s conduct caused the
losses incurred; without his spoofing the price of the affected
commodities would not have risen or fallen, and his counter‐
parties would not have overpaid or received less than the
price their commodity would otherwise have been worth. In
the end, due to the complexity and nature of the crime, gain
was a reasonable substitute for loss.
Conclusion
Mr. Coscia engaged in ten weeks of trading during which
he placed orders with the clear intent to cancel those orders
prior to execution. As a result, Mr. Coscia violated the plain
wording of the Dodd‐Frank Act’s anti‐spoofing provision.
Mr. Coscia engaged in this behavior in order to inflate or de‐
flate the price of certain commodities. His trading accordingly
also constituted commodities fraud. Finally, given the nature
and complexity of his criminal enterprise, the district court
did not err in imposing a fourteen‐point loss enhancement.
For the foregoing reasons, Mr. Coscia’s conviction is affirmed.
AFFIRMED