DYK, Circuit Judge.

Appellant Ohifuemeh Peter Ayewoh ("Ayewoh") was convicted of bank fraud in the United States District Court for the District of Puerto Rico. On appeal Ayewoh contends that the government provided insufficient evidence that the defrauded bank, Banco Popular de Puerto Rico ("BPPR"), was insured by the Federal Deposit Insurance Corporation ("FDIC") at the time of the crimes; that Ayewoh knowingly defrauded BPPR; or that Ayewoh made misrepresentations to BPPR. Ayewoh further contends that the prosecutor violated his Fifth Amendment rights by referring to his decision not to testify. We reject these contentions and affirm Ayewoh's conviction and sentence.

I.

In 2005, Ayewoh established OIPA, Inc., as a for profit corporation in Puerto Rico to provide inspection services at public housing projects for the U.S. Department of Housing and Urban Development ("HUD"). Ayewoh was the President of the corporation, and his wife was the Secretary. In OIPA's name, Ayewoh acquired a credit card point-of-sale device ("POS device") and a merchant's account at BPPR. A POS device allows a merchant to accept credit card payments either electronically (by swiping a card's magnetic strip through the device) or manually (by entering a card's number with a keypad). HUD's method of payment for OIPA's inspection services included providing Ayewoh with a credit card number and authorizing him to perform manual transactions. However, the government presented evidence at trial that Ayewoh used the POS device to execute numerous unauthorized manual transactions using other credit card numbers, many of which belonged to cards issued by foreign banks. The defense's theory was that Ayewoh innocently believed that the credit card charges made to the POS device were authorized by the cardholders as legitimate investments in a new real estate venture. In order to understand the respective theories, some understanding of industry practice and Ayewoh's own activities is required.

When a credit card transaction executed on a POS device is accepted by the bank that issued the card ("issuing bank"), the payment is deposited directly into the merchant's bank account (at the "acquiring bank"). In the event that a cardholder disputes a transaction, contractual agreements with credit card providers such as Visa and MasterCard permit the issuing bank to initiate a "chargeback" whereby the acquiring bank is liable to the issuing bank for the full amount of the disputed charge. The acquiring bank, in turn, looks to the merchant for reimbursement.

Ayewoh was the only person with access to OIPA's POS device. For several months, Ayewoh, using the POS device, processed relatively small charges, with monthly totals ranging from $1,045 in February 2006, to $9,135 in March 2006. However, in April 2006, OIPA's account received $113,070 in payments from at least sixteen different credit cards. BPPR's records showed numerous instances in which OIPA's POS device attempted to run a large transaction, which was declined, and then it immediately attempted several additional transactions with the same credit card for progressively smaller amounts. Upon detecting this suspicious activity, BPPR froze OIPA's account and sent investigators to question Ayewoh. When asked how he obtained the credit card numbers at issue, Ayewoh claimed he received a list of numbers from a "friend of a friend" in Europe who wanted to invest in a new housing development project Ayewoh had planned for OIPA. Though repeatedly asked, Ayewoh declined to provide information about the project and stated that he did not know the identity of the alleged investor.

Soon thereafter the banks that issued the credit cards in question began notifying BPPR that Ayewoh's March and April charges were unauthorized. BPPR subsequently returned $122,104 in chargebacks to the issuing banks. Of this $122,104, BPPR was only able to recover about $100,000 from Ayewoh, who had withdrawn large amounts before his account was frozen. BPPR suffered a net loss of $19,644.

Ayewoh was indicted by a grand jury in November 2007 on one count of bank fraud, charging a violation of 18 U.S.C. § 1344, and one forfeiture count, charging a violation of 18 U.S.C. §§ 981(a)(1)(C) and 982(a)(2). After a jury trial in October 2008, Ayewoh was convicted and sentenced to thirty months in prison to be followed by five years of supervised release. He was further ordered to pay restitution to BPPR in the amount of its losses. Ayewoh timely appealed.

II.

Ayewoh does not dispute that the evidence established that he made unauthorized charges to the credit cards at issue. However, Ayewoh contends on appeal that the government failed to prove all of the essential elements of the bank fraud offense. See 18 U.S.C. § 1344. Specifically, he claims that the government failed to present sufficient evidence to permit a rational jury to find beyond a reasonable doubt (1) that BPPR was insured by the FDIC at the time of his crimes; (2) that Ayewoh had the requisite knowledge that he was defrauding BPPR (as opposed to the individual credit card owners); or (3) that he made misrepresentations to BPPR. We review sufficiency claims de novo, "affirm[ing] the conviction if, after assaying all the evidence in the light most amiable to the government, and taking all reasonable inferences in its favor, a rational factfinder could find, beyond a reasonable doubt, that the prosecution successfully proved the elements of the crime." United States v. Wilder, 526 F.3d 1, 7-8 (1st Cir. 2008) (quoting United States v. Connolly, 341 F.3d 16, 22 (1st Cir. 2003)). We address each claim in turn.

A.

Ayewoh first contends that the government failed to present sufficient evidence at trial to establish that BPPR was insured by the FDIC at the time of his crimes. The bank fraud statute at issue punishes the

knowing[] execut[ion] . . . [of] a scheme or artifice — (1) to defraud a financial institution; or (2) to obtain any of the moneys . . . owned by, or under the custody or control of, a financial institution, by means of false or fraudulent pretenses, representations, or promises.

18 U.S.C. § 1344. The term "financial institution" as used in Title 18 is defined in relevant part as "an insured depository institution (as defined in section 3(c)(2) of the Federal Deposit Insurance Act)." 18 U.S.C. § 20. Section 3(c)(2) in turn defines "insured depository institution" as "any bank or savings association the deposits of which are insured by the [FDIC] pursuant to [the Federal Deposit Insurance Act]." 12 U.S.C. § 1813(c)(2). Thus, under § 1344, the defrauded financial institution's federally-insured status is "a jurisdictional prerequisite as well as a substantive element of the crime," and as such must be proven by the prosecution beyond a reasonable doubt. United States v. Key, 76 F.3d 350, 353 (11th Cir. 1996); see United States v. Brandon, 17 F.3d 409, 424 (1st Cir. 1994).

The government presented two items of evidence at trial to establish BPPR's federally-insured status: (1) a certificate of FDIC insurance issued to BPPR on January 2, 1999; and (2) testimony by BPPR's record custodian regarding the certificate. The record custodian's testimony was as follows:

Q I would like to show you government's identification number 1. . . . Without discussing the substance, do you recognize that document?

A Yes, I do.

Q What do you recognize that to be?

A This is the Federal Deposit Insurance Corporation certificate issued to Banco Popular Puerto Rico on January 1999.

Q How do you know it is the same one at Banco Popular?

A My initials are on the back.

Appellant's App. 131-32 (emphasis added).

Ayewoh did not object to admission of the certificate or to the record custodian's testimony, and he did not cross examine the witness. Nonetheless, Ayewoh questioned in closing arguments whether the evidence was sufficient to establish that BPPR was federally insured during March and April of 2006, the time frame of his criminal acts, and he subsequently moved for dismissal on these grounds. The district court treated Ayewoh's argument as a motion for acquittal under Fed.R.Crim.P. 29 and denied the motion. The court concluded that, while an FDIC certificate issued in 1999 might by itself be insufficient to permit the jury to find that BPPR was federally insured in 2006, the addition of the record custodian's unobjected "present tense" testimony that the certificate "is" BPPR's FDIC certificate rendered the evidence sufficient. United States v. Ayewoh, 587 F.Supp.2d 378, 381 (D.P.R. 2008). That is, the court interpreted the record custodian's statement as "oral testimony that the bank was insured at the time of trial," which, when coupled with an FDIC certificate that predated the offense, could permit a rational jury to infer that BPPR was insured at the time of Ayewoh's crimes. Id.

The FDIC certificate presented in this case, which predated Ayewoh's offenses by seven years, states: "[FDIC] Washington, D.C., Hereby certifies that the deposits of each depositor in Banco Popular de Puerto Rico . . . are insured to the maximum amount provided by the Federal Deposit Insurance Act." It further includes the official seal of the FDIC, signatures from both the FDIC's Executive Secretary and Chairman of the Board, BPPR's unique FDIC certificate number, and the date of January 2, 1999. No expiration date is listed because FDIC insurance never expires, but rather continues indefinitely until terminated voluntarily or involuntarily under 12 U.S.C. § 1818(a) [hereinafter Section 8(a)].

Voluntary termination is not permissible for most banks. 12 U.S.C. § 1818(a)(1). Even involuntary termination for failure to pay the annual assessments (premiums) required by 12 U.S.C. § 1817(b), or for any other reason, is not automatic; it requires notice to the bank and a pre-termination hearing. 12 U.S.C. § 1818(a)(2)(B), (a)(3). Furthermore, for any termination under Section 8(a) — whether voluntary or involuntary — banks are required to give individual notice to each of their depositors, 12 U.S.C. § 1818(a)(6), 12 C.F.R. § 307.3(a), and all deposits on the date of termination will continue to be insured for at least six months and up to two years, 12 U.S.C. § 1818(a)(7). In 2006, the year of Ayewoh's crimes, there were 8,755 banks insured by the FDIC. In the ten years from 1997 through the end of 2006, only two banks had their FDIC insurance terminated under Section 8(a) — one voluntarily and one involuntarily. Clearly, termination of FDIC insurance is an exceedingly rare event in the banking industry.

Nevertheless, there is a serious question as to whether an FDIC certificate that predates the offense by seven years is alone sufficient evidence of federal insurance. The Fifth and Seventh Circuits have overturned convictions where the only evidence of federal insurance was a single FDIC certificate which predated the offense, see United States v. Platenburg, 657 F.2d 797, 799 (5th Cir. 1981) (overturning conviction where only evidence of federal insurance was an FDIC certificate predating the offense by seven years); United States v. Shively, 715 F.2d 260, 265 (7th Cir. 1983) (where only evidence was an FDIC certificate which predated the offense by nine years), and the Ninth and Tenth circuits have indicated their support for this view, see United States v. Chapel, 41 F.3d 1338, 1340-41 (9th Cir. 1994); United States v. Darrell, 828 F.2d 644, 648 (10th Cir. 1987).

We need not decide in this case whether the FDIC certificate would alone be sufficient proof of federal insurance, for here there was also the testimony of BPPR's record custodian who testified that the certificate offered into evidence "is the [FDIC] certificate issued to [BPPR] on January 1999." Appellant's App. 132. Here, the time between the alleged offenses and this trial testimony was approximately two and a half years.

Generally, courts have found that testimony that a bank is insured at the time of trial can alone permit a reasonable jury to infer that the bank was insured at the time of the crime. For example, when presented with a bank vice president's testimony that the bank's deposits "'are' FDIC-insured," the Second Circuit held

that where . . . the evidence is oral testimony that the bank is insured, and the interval between the crime and the trial is not great, it is reasonable to conclude that "viewed in context, the jury could draw the inference that the bank was insured at the time of the [crime]," — in other words, that th[e] jury could take "is" to mean "is and has been."

United States v. Sliker, 751 F.2d 477, 484-85 (2d Cir. 1984) (internal citation omitted). This approach is widely accepted by the circuits. See, e.g., United States v. Ware, 416 F.3d 1118, 1121-23 (9th Cir. 2005); United States v. Lewis, 260 F.3d 855, 855-56 (8th Cir. 2001); United States v. Knop, 701 F.2d 670, 672-73 (7th Cir. 1983); United States v. Safley, 408 F.2d 603, 605 (4th Cir. 1969); Cook v. United States, 320 F.2d 258, 259 (5th Cir. 1963). But see United States v. Ali, 266 F.3d 1242, 1244 n. 3 (9th Cir. 2001) (finding testimony insufficient where time between trial and crime was too lengthy). Furthermore, courts have tended to accept any bank employee's testimony as sufficient, regardless of whether that employee was in a managerial position. See, e.g., United States v. Bindley, 157 F.3d 1235, 1239 (10th Cir. 1998).

However, Ayewoh contends in this case that the district court erred in interpreting the testimony of BPPR's record custodian as "oral testimony that the bank was insured at the time of trial." Ayewoh, 587 F.Supp.2d at 381. Rather, Ayewoh argues that the custodian's testimony that the certificate offered by the government "is the [FDIC] certificate issued to [BPPR] on January 1999" should be construed as nothing more than merely attesting to the document's authenticity.

We conclude that the bank employee's testimony is ambiguous and that reasonable minds could differ as to its interpretation. On one hand, the testimony could be read as meaning "This is the [current FDIC] certificate [which was] issued to [BPPR] on January 1999." Alternately, as Ayewoh contends, the statement could reasonably be interpreted as mere authentication testimony — that is, to mean "This is the [FDIC] certificate [which was] issued to [BPPR] on January 1999."

This Court has previously stated that, "[i]n reviewing a sufficiency-of-the-evidence claim, the Court must view the facts in the light most favorable to the Government, deferring to the jury's verdict if the evidence can support varying interpretations, at least one of which is consistent with the defendant's guilt." United States v. Neal, 36 F.3d 1190, 1203 (1st Cir. 1994) (emphasis added). This analysis "does not require a court to ask itself whether it believes that the evidence at the trial established guilt beyond a reasonable doubt," but rather whether " any rational trier of fact could have found the essential elements of the crime beyond a reasonable doubt." Jackson v. Virginia, 443 U.S. 307, 318-19, 99 S.Ct. 2781, 61 L.Ed.2d 560 (1979) (internal citations and quotation marks omitted). Therefore, we "ought not to disturb, on the ground of insufficient evidence, a jury verdict that is supported by a plausible rendition of the record." Wilder, 526 F.3d at 8 (quoting Ortiz, 966 F.2d at 711) (emphasis added).

Thus, when reviewing evidence for sufficiency and considering ambiguous testimony that is susceptible to more than one reasonable interpretation, we must construe the testimony in the light most favorable to the prosecution. We must therefore presume that the jury interpreted the record custodian's testimony as a statement that the FDIC certificate offered into evidence indicated BPPR's presently-insured status as of the date of trial.

As such, the ultimate question with respect to sufficiency in this case is whether a rational jury could have found beyond a reasonable doubt that BPPR was federally-insured during March and April of 2006 (the time frame of Ayewoh's crimes), given the evidence of: (1) an FDIC insurance certificate issued to BPPR in 1999; and (2) testimony by BPPR's record custodian which, interpreted in the light most favorable to the government, asserted that the certificate remained a valid symbol of BPPR's federally-insured status at the time of trial. We join the vast majority of circuits in holding that such evidence is sufficient proof of FDIC insurance. That is, with evidence of FDIC insurance both at a time predating the offense and at the time of trial, a reasonable jury could infer that, absent evidence to the contrary, the bank was insured on the date of the crime.

Consequently, we reject Ayewoh's contention that the government failed to offer sufficient proof of FDIC insurance. This is not to say that we endorse the prosecutor's handling of the case. The prosecutor assuredly could — and should — have done better. Federally insured status is easily proved in a garden-variety bank fraud case, and it is not too much to expect that the government will take the need to do so seriously. That said, however, we find the proof in this case adequate, if barely, as to that element of the offense.

B.

Ayewoh next contends that he lacked the requisite mens rea for the crime charged. Under the bank fraud statute, the alleged scheme or artifice to defraud a financial institution must have been undertaken knowingly by the defendant. 18 U.S.C. § 1344; see also Brandon, 17 F.3d at 424. Ayewoh argues that the government presented insufficient evidence to prove he had knowledge that his actions would defraud BPPR itself, as opposed to the individual credit card holders. We disagree that the evidence was insufficient.

When interpreting § 1344 in this circuit, "[i]t has been established that the government does not have to show the alleged scheme was directed solely toward a particular institution; it is sufficient to show that [the] defendant knowingly . . . exposed a . . . bank to a risk of loss." Brandon, 17 F.3d at 426 (emphasis added); see also United States v. Moran, 312 F.3d 480, 489 (1st Cir. 2002). "'Intent to harm' [a bank] is not required," United States v. Kenrick, 221 F.3d 19, 29 (1st Cir. 2000) (en banc), so "[e]ven proof of an extremely remote risk will suffice." United States v. Barakett, 994 F.2d 1107, 1111 n. 15 (5th Cir. 1993). Moreover, it is unnecessary that a defendant know "which particular bank will be victimized by his fraud," so long as he "knows that a [bank] will be defrauded" — even a bank that is not federally insured. Brandon, 17 F.3d at 426. As this Court stated in Brandon,

Congress intended to criminalize bank frauds that harm federally insured banks, not just bank frauds directed specifically toward federally insured banks. As other courts have noted, the legislative history supports a broad construction of the statutory language of the bank fraud statute.

Defendants [cannot] . . . sanitize their fraud by interposing an intermediary or an additional victim between their fraud and the federally insured bank. . . . The fact that it should turn out that [a] financial institution actually defrauded was federally insured is a fortuitous stroke of bad luck for the defendants but does not make it any less of a federal crime.

Id. (internal citations and quotation marks omitted). Thus, if a defendant merely knows that his fraudulent actions will expose some bank (whether or not federally insured) to a risk of loss, the mens rea requirement of § 1344 is satisfied.

Ayewoh claims he "had no way of knowing that [BPPR] would be held responsible to the issuing banks for chargebacks" resulting from his unauthorized credit card transactions. Appellant's Br. 27. "[T]he target of the scheme was not [BPPR]," he contends, "but [was] rather the credit card holders. . . ." Appellant's Br. 22. This argument lacks merit. Whether Ayewoh specifically knew BPPR would be liable to the issuing banks for chargebacks is irrelevant, for all § 1344 requires is knowledge that his fraudulent actions would expose some bank to a risk of loss. It is common knowledge that cardholders may generally disavow unauthorized charges on their credit card statements and avoid personal losses from fraud, thereby placing the risk of loss on a bank. A reasonable jury therefore could have concluded that Ayewoh knew his fraudulent actions would expose at least some bank — whether BPPR or even the foreign issuing banks — to a risk of loss. For mens rea, § 1344 requires nothing more. Thus, there was sufficient evidence to find that Ayewoh defrauded BPPR while possessing the requisite mental state under the bank fraud statute.

C.

Ayewoh urges that there was not sufficient evidence for the jury to conclude that he violated § 1344 by knowingly executing a scheme to obtain money from the custody of BPPR "by means of false or fraudulent pretenses, representations, or promises." 18 U.S.C. § 1344. Ayewoh contends that the act of entering a credit card number into a POS device is not a "representation" of anything. We disagree. The majority of our sister circuits have broadly held that the misrepresentation element of § 1344 is fulfilled by any intentional act or statement by an individual that falsely indicates, explicitly or implicitly, that he has authority to withdraw money from a bank. We find that, by running credit card numbers through his POS device, Ayewoh at least implicitly represented to BPPR that he had the cardholders' authorization to make the charges. Here, as in other contexts, we are reluctant to adopt a construction of the statute that would exclude "new financial instruments." United States v. Agne, 214 F.3d 47, 54 (1st Cir. 2000). Moreover, it makes no difference that the agent of the bank defrauded by Ayewoh's deceit was a computer as opposed to a human teller. Without the implied representation that he was authorized to charge the credit cards at issue, BPPR assuredly would not have paid money into OIPA's account. Consequently, there was sufficient evidence for the jury to conclude that Ayewoh acquired funds from BPPR through fraudulent misrepresentations.

III.

We turn finally to Ayewoh's contention that the prosecutor violated his Fifth Amendment rights by referring during closing arguments to Ayewoh's decision not to testify. The closing argument is "an especially delicate point in the trial process," and we scrutinize the prosecutor's comments accordingly. United States v. Glover, 558 F.3d 71, 77 (1st Cir. 2009) (quoting United States v. Taylor, 54 F.3d 967, 977 (1st Cir. 1995)). In Griffin v. California, 380 U.S. 609, 614, 85 S.Ct. 1229, 14 L.Ed.2d 106 (1965), the Supreme Court held that "comment on the [defendant's] refusal to testify" burdens the constitutional privilege against compelled self-incrimination and is therefore prohibited. To decide whether Ayewoh is entitled to relief, we engage in a two-pronged analysis: First, we conduct a de novo review to determine whether the prosecutor's comments offended the Fifth Amendment by improperly insinuating that Ayewoh's failure to testify constituted evidence of guilt. Gomes v. Brady, 564 F.3d 532, 537 (1st Cir. 2009). To that end, "we consider 'whether, in the circumstances of the particular case, the language used was manifestly intended or was of such character that the jury would naturally and necessarily take it to be a comment on the failure of the accused to testify.'" Id. at 538 (quoting United States v. Glantz, 810 F.2d 316, 322 (1st Cir. 1987)). Second, if we determine the comments were improper and the objection was preserved, we review for harmless error. Glover, 558 F.3d at 76.

At trial, Ayewoh's counsel urged that Ayewoh did not knowingly defraud anyone because he believed at the time that he had obtained the credit cards legitimately from a "friend of a friend" who sought to invest in his company OIPA, Inc. Under this theory, the unidentified "friend of a friend" was the real criminal who defrauded innocent Ayewoh by purporting to be a legitimate investor. Though Ayewoh did not testify at trial, he proffered this defense through the testimony of his wife. BPPR's investigator Mr. Ponce also testified that Ayewoh stated during an interview that he had received the credit cards from a "friend of a friend" who was an investor in Europe and that the transaction had been done entirely through his friend. Appellant's App. 211. Finally, in closing arguments, Ayewoh's counsel repeatedly asserted that the purported "friend of a friend" was entirely to blame and that Ayewoh was himself an innocent victim of fraud. For context, we quote the relevant portions of defense counsel's closing argument:

[T]he money was invested in OIPA, Inc., by this friend of a friend.

Mr. Ayewoh committed a mistake, but that mistake was not defrauding anybody. That mistake was trying to grow and seeing an opportunity and grabbing it without looking at what was behind it, because somebody who knew that he was in need offered him his dream, I'll give you the money, here, start charging here. And he did.

Trial Tr., Oct. 14, 2008, ECF No. 84, at 28 (emphasis added).

[Ayewoh] got somebody to invest in his company, the company that he, with his wife, was developing so that he could make a better living for himself and also pay back his investors.

But he was defrauded. . . . And you'll say, but look at all those things that the government showed. . . .

. . . .

The government doesn't know if [Ayewoh] called his friend to find out, hey, these [credit cards] are being rejected, what's going on[?] And his friend told him, Oh, I gave you the wrong number. Oh, I need to deposit. Oh, I need to do something.

We don't know that. The government doesn't know that.

Trial Tr., ECF No. 84, at 29 (emphasis added).

Ladies and gentlemen, the mistake Mr. Ayewoh made was that he believed the friend of a friend.

Trial Tr., ECF No. 84, at 35 (emphasis added).

It is a pity that [Ayewoh] didn't have a contract with this gentleman that lent him the money. If he had sought his attorney's advice at the moment, he would have had one.

Trial Tr., ECF No. 84, at 38.

Everything that surrounds this case shows that somebody tried to defraud the bank. . . .

But that somebody is not Mr. Ayewoh. That somebody is the person, the friend of the friend that gave him the credit card numbers. We don't know who he is.

Trial Tr., ECF No. 84, at 40 (emphasis added).

Seeking to cast doubt on Ayewoh's defense theory, the government made several comments in closing regarding the lack of a known identity for Ayewoh's purported "friend of a friend." Specifically, Ayewoh contends that the following three comments made during the government's rebuttal argument violated his Fifth Amendment rights by improperly commenting on his failure to testify as to the alleged investor's identity:

[1] Okay. Let's say the defendant says it is a mistake. Well, you have an instruction on willful blindness. Was he willfully blind that these credit cards were, in fact, stolen[?]

Well, let's look at the defendant's statement.

A friend of a friend. To this day, to this day, you have heard from government witnesses, you have even heard from the defendant's spouse. We still don't know who this friend of a friend is. . . .

You are going to have, you know, an individual make over $100,000 worth of investments in your company and then not know who the person is. Not do an investment contract?

Trial Tr., ECF No. 84, at 46-47 (emphasis added).

[2] It makes absolutely no sense that this defendant got these numbers from a friend of a friend, over 16 credit cards, start[ed] essentially doing the manual POS transactions, multiple times, and not know who that individual is.

We still don't know who that individual is. Gentleman who lent the money, we have no idea. We have no name.

Trial Tr., ECF No. 84, at 47-48 (emphasis added).

[3] But here, in looking at all the evidence — and, you know, the defendant has argued that, now, this was a mistake, that he believed a friend of a friend. That it was a mistake, in believing a friend of a friend, that what he wanted to do was to get a good life.

Well, this is not believable. Use your common sense. Okay. Common sense doesn't — it doesn't make sense. It is not a mistake that you get 17 credit card numbers from a person that you don't know, have no idea and then do multiple transactions over a short period of time in order to figure out what the credit limits of the credit cards are. You just don't do that.

Because if you have a friend of a friend who is actually giving you these numbers and you know who that friend is, you call that friend and you tell the friend, "Listen, the credit card isn't working, what's happening, [is it] being declined for insufficient funds?" But you haven't heard evidence of that.

Trial Tr., ECF No. 84, at 51-52 (emphasis added).

Ayewoh's counsel timely objected to the first of these remarks by requesting to approach the bench, but the court denied the request. Ayewoh subsequently moved for a mistrial "based on the comments made by the prosecution as to the defendant's silence," but the court denied the motion as well. Ayewoh now contends on appeal that the prosecutor's comments were improper because only Ayewoh could have provided the identity of the "friend of a friend," so references to the lack of such evidence necessarily pointed to his failure to testify. We disagree.

First, contrary to Ayewoh's contention that only he could provide the identity of the "friend of a friend," Ayewoh himself was not the only person who could identify the alleged investor if he existed. For example, the "friend" who purportedly acted as an intermediary between Ayewoh and the mysterious "friend of a friend" might have testified as to the investor's identity and to their business arrangement. Moreover, in his interview with the bank representative, Ayewoh reportedly stated that he never knew the identity of the "friend of a friend" because the deal was all done through an immediate friend. Appellant's App. at 211. Thus, this case is distinguishable from cases in which "[n]o one but [the] appellant . . . could have [testified]" regarding the issue, see Desmond v. United States, 345 F.2d 225, 227 (1st Cir. 1965), because here it was "apparent on the record that there was someone other than himself whom the defendant could have called." Id. As such, the prosecutor's comments regarding the lack of such evidence in the record did not "naturally and necessarily" point to Ayewoh's failure to testify. Gomes, 564 F.3d at 538 (quoting Glantz, 810 F.2d at 322); see also United States v. Wilkerson, 411 F.3d 1, 8-9 (1st Cir. 2005) (prosecutor's comments that "there's no real evidence" and "pretty much nothing" to support defendant's theory did not "naturally and necessarily" comment on his failure to testify because they likely "referred to [defendant's] failure to produce other evidence supporting his theory," or were at most ambiguous).

Second, even if the prosecutor's remarks could be read as commenting on Ayewoh's failure to testify, the comments were a "fair response" to claims by the defendant. In United States v. Robinson, 485 U.S. 25, 34, 108 S.Ct. 864, 99 L.Ed.2d 23 (1988), the Supreme Court carved out an exception to Griffin for prosecutors who refer to a defendant's failure to testify while "fairly responding to an argument of the defendant." In that case, the defense counsel asserted in closing arguments that the government had not given the defendant a chance to tell his side of the story. Id. at 27, 108 S.Ct. 864. On rebuttal, the prosecutor responded by informing the jury that the defendant "could have taken the stand and explained it to you." Id. at 28, 108 S.Ct. 864. Finding no Fifth Amendment violation, the Supreme Court reasoned that, "[w]here the prosecutor on his own initiative asks the jury to draw an adverse inference from a defendant's silence, Griffin holds that the privilege against compulsory self-incrimination is violated," but where the prosecutor's reference to the defendant's failure to testify is a "fair response to a claim made by defendant or his counsel," the Fifth Amendment is not offended. Id. at 31-32, 108 S.Ct. 864. The Court further explained that "the Fifth Amendment should [not] be converted into a sword that cuts back on the area of legitimate comment by the prosecutor on the weaknesses in the defense['s] case." Id. (quoting United States v. Hasting, 461 U.S. 499, 515, 103 S.Ct. 1974, 76 L.Ed.2d 96 (1983) (Stevens, J., concurring)).

Following the Supreme Court's decision in Robinson, this Court has applied the "fair response" doctrine in numerous cases. Likewise, our sister circuits have invoked the "fair response" doctrine to reject Fifth Amendment claims on similar facts.

We find that Robinson controls in this case and that, viewed in context, the prosecutor's challenged comments were part of a "fair response" to the arguments made by Ayewoh's counsel. The first two challenged comments, which emphasized the lack of evidence regarding the identity of the purported "friend of a friend," were permissible attacks on the plausibility of Ayewoh's defense theory. "Having put forth a theory in [his] defense," Ayewoh "[could not] expect the government to refrain from commenting on its deficiencies" and the lack of evidentiary support for his theory — namely, that the record did not even include the identity of the mysterious person whom Ayewoh claimed he mistakenly believed "invested" over $122,000 in his company. The prosecution was permitted to highlight such a glaring weakness in the defense's case.

Similarly, it is clear that the prosecutor's third challenged comment, viewed in context, was a "fair response" specifically to the defense counsels comment that "[t]he government doesn't know if [Ayewoh] called his friend to find out" why some of the credit cards weren't working. Trial Tr., ECF No. 84, at 29. On rebuttal, the prosecutor's third comment fairly responded to the defense counsel's statement by informing the jury that "you haven't heard evidence" that Ayewoh placed any such calls to his friend. Trial Tr., ECF No. 84, at 52.

Thus, because all of the prosecutor's challenged comments were made in "fair response" to claims by the defense, Ayewoh's Fifth Amendment rights were not violated.

IV.

For the foregoing reasons, we affirm Ayewoh's conviction and sentence on the bank fraud and forfeiture counts.

Affirmed.