Materiality Limits on Late-Disclosed Interview Memos, Tailored Post‑Trial Discovery, and Conservative Sampling-Based Loss in Tax-Refund Check‑Cashing Fraud

1. Introduction

United States v. Thomas Addaquay (11th Cir. Sept. 9, 2026) arises from a tax-refund-fraud prosecution centered on United Consolidated Accounting and Business Services (“UC”), a business the government characterized as a conduit for converting fraud-generated tax refund checks into spendable funds through a third-party processor, ReliaFund. A jury convicted Thomas Addaquay of conspiracy to commit wire fraud, multiple substantive wire-fraud counts, aggravated identity theft, and money-laundering-related offenses. Separately, he later pleaded guilty to structuring transactions to evade federal reporting requirements, and the district court sentenced him in the tax-refund-fraud and structuring cases together.

On appeal, Addaquay challenged: (i) the sufficiency of evidence for certain wire-fraud and aggravated-identity-theft counts; (ii) alleged Brady v. Maryland / Giglio v. United States violations based on four investigative memoranda disclosed post-trial; (iii) denial of broader post-trial motions to compel discovery; and (iv) the district court’s loss and restitution calculations using sampling (including reliance on IRS Transaction Code 841, median-based extrapolation, and an additional 25% downward adjustment). The Eleventh Circuit affirmed across the board.

2. Summary of the Opinion

  • Sufficiency of evidence: The court held the evidence was sufficient to support wire-fraud convictions tied to (a) ReliaFund-to-UC settlement transfers and (b) electronically filed unauthorized returns that generated refund checks later processed through UC. It also upheld related aggravated identity theft counts.
  • No Brady/Giglio violation: The late-disclosed interview/proffer memoranda (Awiti and Liady materials) were not material and did not establish perjury.
  • Discovery rulings affirmed: The district court permissibly ordered production of tax returns already in the government’s possession relevant to loss, but denied speculative broader requests (investigative files and K1 client list).
  • Loss/restitution affirmed: The court upheld a conservative loss estimate derived from sampling, the median (rather than mean), and a further 25% reduction “in the abundance of caution,” resulting in $4,123,474.55 loss and restitution to the IRS in that case.

3. Analysis

3.1 Precedents Cited

A. Sufficiency review and inferences favoring the verdict

  • United States v. Jiminez, 564 F.3d 1280 (11th Cir. 2009) (citing United States v. Williams, 144 F.3d 1397 (11th Cir. 1998)): used for the evidentiary lens—viewing proof “in the light most favorable to the verdict.” This framing underlies the court’s rejection of Addaquay’s attempt to disaggregate each transaction from the broader scheme.
  • United States v. Deason, 965 F.3d 1252 (11th Cir. 2020) and United States v. Merrill, 513 F.3d 1293 (11th Cir. 2008): reinforce the “reasonable construction of the evidence” standard and the deference owed to jury credibility determinations.

B. Elements of wire fraud and proving participation via scheme membership

  • United States v. Machado, 886 F.3d 1070 (11th Cir. 2018): supplied the elements of 18 U.S.C. § 1343 and confirmed intent may be inferred from conduct and circumstantial evidence.
  • United States v. Hewitt, 663 F.2d 1381 (11th Cir. 1981) (quoting United States v. Martinez, 555 F.2d 1269 (5th Cir. 1977)): provided the classic aiding-and-abetting formulation—association with the venture, participation to bring it about, and actions to make it succeed. The court relied on this to explain why the government need not prove Addaquay personally filed each return or personally initiated each wire.
  • United States v. Schwartz, 666 F.2d 461 (11th Cir. 1982): used to emphasize that guilt requires more than “association”—the evidence must support knowing participation.
  • United States v. Ward, 486 F.3d 1212 (11th Cir. 2007) and United States v. Watkins, 42 F.4th 1278 (11th Cir. 2022): central to the court’s treatment of substantive wire fraud—membership in the scheme plus a co-schemer’s use of wires in execution can sustain conviction.
  • United States v. Williams, 527 F.3d 1235 (11th Cir. 2008): invoked for the proposition that the government must prove the charged wires furthered the scheme, not that every business transaction was fraudulent.
  • United States v. Reed, 887 F.2d 1398 (11th Cir. 1989): supported the “on or about” flexibility as to dates, addressing a one-day discrepancy in a charged settlement date.

C. Identity theft aiding-and-abetting knowledge

  • United States v. Barrington, 648 F.3d 1178 (11th Cir. 2011): supplied the basic aggravated identity theft framework under 18 U.S.C. § 1028A in relation to the predicate fraud offense.
  • Rosemond v. United States, 572 U.S. 65 (2014): used to articulate that an aider and abettor must intend to facilitate the “specific and entire crime charged” and must have the requisite knowledge at a time when he can still choose to withdraw. The Eleventh Circuit held the record permitted the inference that Addaquay knew about identity theft during the scheme (e.g., PII files, runners, forged endorsements), not merely after funds arrived.

D. Brady/Giglio materiality and perjury requirements

  • Brady v. Maryland, 373 U.S. 83 (1963) and Giglio v. United States, 405 U.S. 150 (1972): provided the constitutional baselines; the court treated Giglio as a “species” of Brady requiring proof of perjury and prosecutorial knowledge.
  • United States v. Noriega, 117 F.3d 1206 (11th Cir. 1997) (quoting Kyles v. Whitley, 514 U.S. 419 (1995)): supplied the key materiality test—reasonable probability of a different result and cumulative assessment of withheld items.
  • United States v. Vallejo, 297 F.3d 1154 (11th Cir. 2002): provided the Eleventh Circuit’s four-part Brady test used to structure the analysis.
  • Ford v. Hall, 546 F.3d 1326 (11th Cir. 2008): defined Giglio error (perjury and prosecutorial knowledge) and its “reasonable likelihood” materiality standard.
  • United States v. Michael, 17 F.3d 1383 (11th Cir. 1994): underscored that Giglio requires actual falsity, not simply testimony that could have been impeached or supplemented.
  • United States v. Jones, 601 F.3d 1247 (11th Cir. 2010): supported the conclusion that additional impeachment is often not material where the jury already knows the witness’s incentives and credibility issues.
  • United States v. Jordan, 316 F.3d 1215 (11th Cir. 2003) and United States v. Gallardo, 977 F.3d 1126 (11th Cir. 2020): supplied standards of review for Brady/Giglio issues and motions to dismiss/new trial.

E. Post-trial discovery limits

  • United States v. Espinosa-Hernandez, 918 F.2d 911 (11th Cir. 1990): supplied abuse-of-discretion review for post-trial discovery rulings.
  • United States v. Arias-Izquierdo, 449 F.3d 1168 (11th Cir. 2006): emphasized Brady is not a discovery device and courts should not order discovery based on speculation; also recognized Brady’s post-trial posture.

F. Loss and restitution estimation in large frauds

  • United States v. Medina, 485 F.3d 1291 (11th Cir. 2007), United States v. Bradley, 644 F.3d 1213 (11th Cir. 2011), and United States v. Willis, 560 F.3d 1246 (11th Cir. 2009): grounded the principle that loss need only be a “reasonable estimate” supported by specific circumstantial evidence, not precision.
  • United States v. Sepulveda, 115 F.3d 882 (11th Cir. 1997): cautioned sentencing courts against speculation that increases Guidelines exposure.
  • United States v. Campbell, 765 F.3d 1291 (11th Cir. 2014): supported using amounts transferred to a fraudulent enterprise as a starting point when conduct is “permeated with fraud” (though the district court here still adopted a conservative approach).
  • United States v. Foley, 508 F.3d 627 (11th Cir. 2007) and United States v. Stein, 846 F.3d 1135 (11th Cir. 2017): reiterated the government’s burden (preponderance) to prove loss.
  • United States v. Presendieu, 880 F.3d 1228 (11th Cir. 2018) and United States v. McCrimmon, 362 F.3d 725 (11th Cir. 2004): framed relevant conduct and jointly undertaken criminal activity under U.S.S.G. § 1B1.3(a)(1)(B).
  • United States v. Whitman, 887 F.3d 1240 (11th Cir. 2018) (quoting United States v. Moran, 778 F.3d 942 (11th Cir. 2015)): justified appellate deference to district courts’ loss findings.
  • United States v. Gupta, 572 F.3d 878 (11th Cir. 2009): used as a contrast point—here the district court articulated an evidentiary basis for its adjustments.
  • United States v. Baldwin, 774 F.3d 711 (11th Cir. 2014) and United States v. Moss, 34 F.4th 1176 (11th Cir. 2022): supported that restitution must reflect actual loss but may be reasonably estimated when precision is impracticable.

G. Out-of-circuit authority addressed and distinguished

  • United States v. Ragan, 24 F.3d 657 (5th Cir. 1994): invoked by Addaquay for reversal, but the Eleventh Circuit distinguished it because in Ragan the proof linked the defendant to uncharged trades, whereas here each charged transmission was tied to unauthorized returns/checks and the UC–ReliaFund payment channel.
  • Bonner v. City of Prichard, 661 F.2d 1206 (11th Cir. 1981) (en banc): cited to emphasize that decisions of other circuits (like Ragan) are not binding.

3.2 Legal Reasoning

A. Wire fraud: scheme participation plus charged wires in execution

The court’s sufficiency analysis proceeds in two steps: (1) whether a rational jury could find Addaquay knowingly joined a fraudulent venture; and (2) whether the charged wire transmissions were used to execute that venture. Relying on United States v. Machado for elements and United States v. Ward/United States v. Watkins for scheme-membership liability, the court held the government was not required to prove that Addaquay personally filed each fraudulent return or initiated each wire transfer.

Critically, the panel treated UC’s operational features (overwhelmingly tax-refund checks, minimal ordinary customer traffic, employee testimony about forging endorsements, and Addaquay’s representations to ReliaFund) as evidence from which the jury could infer knowledge and intent. It then matched each challenged count to specific proof: a ReliaFund settlement or an electronically filed unauthorized return, coupled with victim testimony of non-authorization and records tracing checks into UC accounts.

B. Aggravated identity theft: Rosemond knowledge satisfied by end-to-end participation indicators

For Counts 13–16, the court linked identity theft to the same electronically filed returns underpinning the challenged wire-fraud counts and stressed that Addaquay’s role was not limited to “after-the-fact” receipt of money. Invoking Rosemond v. United States, the panel concluded the jury could infer advance or contemporaneous knowledge because the evidence placed Addaquay at multiple points where identity theft was inherent: possession of PII, dealing with runners, and directing forged endorsements on checks payable to victims.

C. Brady/Giglio: late disclosure is not enough—materiality and falsity are required

The panel assumed arguendo that the government “suppressed” the memoranda but found no materiality. The court emphasized: (i) the Awiti memorandum did not contradict Edwards’s core testimony that Edwards cashed checks for Addaquay; (ii) the jury already knew Edwards’s key credibility problems (immunity, participation, financial motive, and personal animus); (iii) Awiti’s statements were partly inculpatory of Addaquay (knowledge that checks were tied to stolen identities and help disguising proceeds); and (iv) independent evidence supported guilt apart from Edwards.

For Giglio, the panel treated United States v. Michael as controlling in principle: “more detail” or an “incomplete” story does not equal perjury. The Awiti memo, in the court’s view, suggested at most a different perspective on relationships and side-arrangements, not proof that Edwards testified falsely.

The Liady documents were deemed too attenuated: they described separate fraud by a different preparer and did not connect to charged victims, transmissions, or specific checks. Considering the items cumulatively under Kyles v. Whitley, the court held they did not undermine confidence in the verdict.

D. Post-trial motions to compel: a “reasoned line” between material and speculative requests

Echoing United States v. Arias-Izquierdo, the court approved the district court’s refusal to compel broad investigative files and a client list based on speculation, while affirming the order compelling production of tax returns already in the government’s possession because those documents directly tested the loss methodology. The court rejected the “burden shifting” argument: requiring a defendant to articulate a concrete basis for compelled discovery is not the same as requiring him to disprove loss.

E. Loss and restitution: conservative sampling, defense-favorable median, plus an additional discount

The opinion affirms a layered approach to estimation in large-scale fraud:

  1. Sampling and extrapolation were acceptable given impracticality of check-by-check proof, consistent with “reasonable estimate” doctrine from United States v. Medina and United States v. Bradley.
  2. The district court did not treat IRS Transaction Code 841 as a definitive “fraud” label; it recognized error rates in both directions.
  3. The court credited Addaquay’s expert critique of the mean (outlier distortion) and adopted a median-based estimate.
  4. The additional 25% reduction was upheld because the district court articulated why it was applied—accounting for legitimate checks, Code 841 limits, and attribution uncertainty—distinguishing United States v. Gupta.

For restitution, the court relied on United States v. Baldwin and United States v. Moss to hold that actual loss can be reasonably estimated where exact calculation is infeasible, and it found no separate defect beyond the loss objections already rejected.

3.3 Impact

  • Scheme-based proof in payment-processing fraud: The decision reinforces that when a defendant is proven to control the “monetization” channel (here, converting refund checks into deposits via ReliaFund), the government can sustain substantive wire-fraud counts without proving the defendant filed each underlying return— so long as the charged wires are concretely tied to unauthorized returns/checks and the defendant’s knowing participation is supported by operational evidence.
  • Brady/Giglio discipline for post-trial memo disclosures: The opinion illustrates a common appellate stance: late disclosure does not equal reversal absent (i) a true contradiction or perjury, and (ii) materiality in light of cumulative trial evidence and existing impeachment.
  • Tailored post-trial discovery: The affirmed “line-drawing” signals that courts may compel production directly relevant to loss testing (e.g., returns in possession) while denying broad fishing expeditions into investigative files and third-party client lists unless tied to particular losses or checks.
  • Loss methodology in high-volume refund fraud: The case supports the acceptability of conservative statistical approaches—sampling, use of robust statistics (median), and explicit downward adjustments to avoid over-attribution—when the court explains the evidentiary basis and errs on the side of caution.

4. Complex Concepts Simplified

  • Wire fraud (18 U.S.C. § 1343): A plan to cheat someone, done intentionally, that uses electronic communications (internet transmissions, bank wires) as part of carrying out the plan.
  • Aiding and abetting (18 U.S.C. § 2): You can be guilty as a “principal” if you knowingly help the crime succeed—even if you did not personally do every step.
  • Aggravated identity theft (18 U.S.C. § 1028A): Using a real person’s identifying information (like a Social Security number) without permission during and in relation to certain felonies (here, wire fraud). Under Rosemond v. United States, an aider/abettor must have the required knowledge in time to choose not to participate.
  • Brady v. Maryland: The government must disclose evidence favorable to the defense if it is material—meaning there is a reasonable probability it would change the outcome.
  • Giglio v. United States: A Brady-type rule focused on false testimony: if the prosecution knowingly uses perjury (or fails to correct it) and it could affect the verdict, a new trial may be required.
  • Materiality: Not “could be useful,” but “could realistically change the result,” considering the entire trial record.
  • Sampling and extrapolation: Reviewing a random subset of transactions and using it to estimate totals for a huge population. Courts allow this if the method is reasonable and the estimate is grounded in evidence rather than guesswork.
  • Mean vs. median: The mean (average) can be skewed by a very large outlier; the median (middle value) is often more stable for fraud-loss estimation when data are uneven.
  • Restitution: Money the defendant must pay back to victims for actual losses caused. It is not intended to punish; it is intended to compensate.

5. Conclusion

The Eleventh Circuit’s decision in United States v. Thomas Addaquay affirms three practical points in complex, high-volume financial fraud cases: (1) scheme participation plus proof that the charged wires executed the scheme can sustain substantive wire-fraud and related identity-theft convictions, even absent proof the defendant personally filed each return; (2) post-trial disclosure of interview memoranda does not establish Brady v. Maryland or Giglio v. United States violations without materiality (and, for Giglio, proof of falsity/perjury); and (3) district courts have latitude to estimate loss and restitution through conservative, explained sampling methods—especially when they adopt defense-favorable statistical corrections and further discount uncertain portions of the estimate.