What decides whether bank-record tracing counts as lay testimony or expert testimony?
Whether the analysis required choosing among competing methods. Federal Rule of Evidence 701(c) limits lay opinion to testimony "not based on scientific, technical, or other specialized knowledge within the scope of Rule 702," and the two federal appellate decisions that have applied that provision to bank-record tracing in the past two years draw the line in the same place: arithmetic performed on documents is lay testimony, and selection among competing techniques is expert testimony. Credentials do not decide it, and the Tenth Circuit said so directly in United States v. Joseph, 108 F.4th 1273 (10th Cir. 2024), noting that merely possessing expert knowledge does not alone require a person to testify as an expert.
That line is uncomfortable for forensic accountants precisely because so much of the work looks like addition. Tracing funds through commingled accounts can be a matter of adding deposits and subtracting withdrawals, or it can require the analyst to choose between first-in-first-out, last-in-first-out, pro rata allocation and the lowest intermediate balance rule. Consulting-firm practitioner literature reports that those conventions produce materially different traceable amounts from identical records and that no professional standard or legal rule prescribes a single one — a practitioner observation rather than a holding, but one no opposing expert seems to dispute. The first operation is arithmetic. The second is methodology. So the question to put to any tracing witness, on either side of the case, is which convention was applied and what the figure becomes under the others.
Volume is a weaker signal than it looks. The Seventh Circuit treated a very large quantity of simple arithmetic as lay testimony, while the Tenth Circuit in United States v. Joseph cited the scale of the analysis alongside the methodology choice, holding that the testimony exceeded what a lay witness may do "in both scope and kind." Neither court made size alone decisive, and neither treated it as irrelevant.
Why did the Seventh Circuit allow an FBI forensic accountant to trace bank records without qualifying her as an expert?
Because the Seventh Circuit found the method simple even though the volume was enormous. In United States v. Fenner, Nos. 23-2177 and 24-1089 (7th Cir. 1 July 2025), FBI forensic accountant Kathryn Kanetzke reviewed "over tens of thousands" of pages of bank records, traced the flow of money through bank accounts, expense sheets and emails, and summed up that the two defendants received "gross receipts of approximately $1 million." The government did not qualify Kathryn Kanetzke as an expert witness. The Seventh Circuit held that admitting her testimony was not an abuse of discretion: "[w]hile the 'volume of math' Kanetzke performed was large, the method was simple," and her testimony "required attention to detail; it did not require expertise beyond the ken of a lay person."
Two features of United States v. Fenner did the work, and both are worth noting before the decision is treated as a general license. Kathryn Kanetzke "confined her testimony to the documents and communications in the closed record of the investigation," which is what keeps a witness inside Rule 701's requirement that lay opinion be rationally based on the witness's perception. And the operations the opinion describes are addition and subtraction: to reach the revenue figure she added up the total reported sales of the 100-plus vehicles sold, and she observed deposits of tens of thousands of dollars into one account followed shortly by checks out of it in similar amounts. The opinion records no choice of tracing convention. It records arithmetic done at scale.
Why did the Tenth Circuit hold that using a certified fraud examiner as a lay witness was error?
Because the witness had to choose a tracing method before he could testify at all. In United States v. Joseph, 108 F.4th 1273 (10th Cir. 2024), the government offered certified fraud examiner Michael Petron as a witness to explain the defendant's banking transactions, and the district court permitted it without an expert designation provided he limited his explanations to relatively rudimentary arithmetic. Michael Petron selected what the opinion calls a "conservative" methodology for tracking funds, then gave an in-depth portrait of "eight or nine" bank accounts based on analysis of "thousands of transactions." The Tenth Circuit held that admitting him as a lay witness was error: his testimony went beyond what is permissible for lay witnesses "in both scope and kind," and "to conduct his testimony he had to choose between multiple possible methodologies for tracing bank funds," which "appears to be exactly the type of 'technical, or other specialized knowledge' that is reserved for expert witnesses."
United States v. Joseph rests on a line of Tenth Circuit authority that is more precise than "it depends." Taking the median of several numbers is basic arithmetic and so is taking "a simple average of 103 numbers, though technically a statistical determination." Calculating depreciation is not, because "[t]echnical judgment is required in choosing among different types of depreciation," and "moving averages, compounded growth rates, and S-curves" have been classified as too technical for lay testimony. The court's own formulation is that testimony stops being lay testimony when it moves beyond basic mathematics either by considering advanced topics or by requiring the selection of a mathematical methodology. Tracing across commingled accounts sits with depreciation rather than with the average, for the same structural reason: more than one defensible convention exists, and somebody has to pick one.
What happens if the lay-versus-expert call is wrong?
The party who offered the witness has usually served no expert disclosure, and that is a separate problem from whether the analysis was reliable. The Advisory Committee said as much when Rule 701(c) was added in 2000: channelling testimony that is actually expert testimony to Rule 702 "ensures that a party will not evade the expert witness disclosure requirements set forth in Fed.R.Civ.P. 26 and Fed.R.Crim.P. 16 by simply calling an expert witness in the guise of a layperson." A witness misclassified as lay is a witness for whom nothing was disclosed, so the exposure is procedural before it is evidentiary.
In civil matters a forensic accountant retained or specially employed to give expert testimony owes a written report under Federal Rule of Civil Procedure 26(a)(2)(B) containing all opinions and the basis and reasons for them, the facts or data considered, any exhibits used to summarize or support them, qualifications including all publications from the previous ten years, a list of other cases in which the witness testified as an expert in the previous four years, and a statement of compensation. A witness who is neither retained for the case nor an employee whose duties regularly involve giving expert testimony falls under Rule 26(a)(2)(C) and owes only a statement of the subject matter and a summary of the facts and opinions. Federal Rule of Civil Procedure 37(c)(1) then does the damage without anyone filing a Rule 702 motion: a party that fails to disclose "is not allowed to use that information or witness to supply evidence on a motion, at a hearing, or at a trial, unless the failure was substantially justified or is harmless." In criminal matters the parallel obligation is Federal Rule of Criminal Procedure 16(a)(1)(G), amended effective 1 December 2022 to require a complete statement of all opinions rather than a summary, the bases and reasons, qualifications including a ten-year publications list, a four-year testimony list, and the witness's own approval and signature unless the government explains why it could not obtain one.
The same "simple math" characterisation runs the other way in civil litigation, where the party opposing a forensic accountant argues that the work is ordinary arithmetic and therefore adds nothing the trier of fact needs. In Metropolitan Design & Development, LLC v. Frankenmuth Mutual Insurance Co., No. 1:25-cv-38 (S.D. Ohio 29 Apr. 2026), the insurer made exactly that argument about a certified public accountant specializing in forensic accounting, and the court disagreed: because that type of financial analysis was grounded in her specialized knowledge as a forensic accountant, it qualified as expert testimony. Her opinions were nonetheless excluded in part, on relevance rather than reliability, to the extent they reached persons who could not recover under Ohio bad-faith law. Two honest caveats belong with all of this. The Tenth Circuit found the error in United States v. Joseph harmless because admitted bank records and a separate agent's detailed account had already put the same movement of funds before the jury, and in United States v. Fenner the Seventh Circuit found no abuse of discretion as to the forensic accountant and held that any error in a second witness's testimony was harmless. Harmlessness is a poor thing to plan around when the question is whether your own tracing evidence gets in.
Does calling the tracing schedule a Rule 1006 summary avoid the problem?
No. Federal Rule of Evidence 1006 is a different vehicle with its own limits, not an exemption from the lay-versus-expert line. Rule 1006 lets a court admit as evidence "a summary, chart, or calculation offered to prove the content of voluminous admissible writings, recordings, or photographs that cannot be conveniently examined in court," and the proponent must make the underlying originals or duplicates available to other parties for examination or copying. Since the amendments effective 1 December 2024, Rule 1006 governs only summaries admitted as evidence, while "[a] summary, chart, or calculation that functions only as an illustrative aid is governed by Rule 107" — under which an illustrative aid "is not evidence and must not be provided to the jury during deliberations" unless all parties consent or the court orders otherwise for good cause. A great many tracing schedules ride in under Rule 1006, which is why counsel who prepare only a Rule 702 motion sometimes find they have aimed at the wrong target.
Federal Rule of Evidence 1006 also carries its own limit on inference. In United States v. Fenner the Seventh Circuit held that summary witnesses "may testify to what 'the Government's evidence shows,' but cannot offer implausible or speculative inferences from those records," and that summary witnesses "should tread carefully when opining on a defendant's intent." The same opinion supplies the affirmative model, in the court's own account of a second summary witness, Special Agent Kathryn Graham: she "never concluded that Fenner inflated the value of his liens," the Seventh Circuit wrote, but "identified specific documents" bearing on what was charged for towing against what was paid for it, and "never opined that these costs were 'fake' or 'illegal'" — leaving that question for the jury. Where her testimony did address motivation the court called it "perhaps a slightly closer call" and held any error harmless. The complication worth stating rather than papering over is that the same opinion notes lay opinion as to the mental state of another is admissible from non-expert witnesses, so for a lay summary witness the rule is caution rather than prohibition. The tie-out discipline is what keeps a schedule a summary rather than an argument: every figure on it should trace to an identified underlying document.
What does correctly scoped tracing testimony actually sound like?
It describes what changed in the records, in dollars and dates, and stops there. In United States v. May, 131 F.4th 633 (8th Cir. 2025), the Eighth Circuit recorded that the FBI's forensic accountant testified to "cash deposits in excess of $15,000 into May's bank accounts in 2015, which was a significant increase over the less than $500 in cash deposits made in 2014," and treated that as circumstantial corroboration of the conduct charged. That is a pattern change in bank records, described in figures and years. The opinion records no testimony from that witness about what anyone intended, and none characterizing the conduct in the vocabulary of the charged offenses.
Two limits sit behind that discipline and they come from different places. Federal Rule of Evidence 704(a) provides that an opinion "is not objectionable just because it embraces an ultimate issue," but in United States v. Scop, 846 F.2d 135 (2d Cir. 1988), an SEC regional office's chief investigator testifying as an expert drew directly on the language of the statutes and regulations charged, and the Second Circuit held that his "repeated statements embodying legal conclusions exceeded the permissible scope of opinion testimony," reversing all but the false-declaration convictions on that ground and on a second, independent objection to opinions resting on his assessment of another witness's credibility. Rule 704(b) adds a separate criminal-case bar that reaches experts only: "an expert witness must not state an opinion about whether the defendant did or did not have a mental state or condition that constitutes an element of the crime charged or of a defense." Which is one more reason the lay-versus-expert call matters — it changes which intent rule applies to the same witness.
Two professional bodies impose a further restraint on their own members. AICPA Statement on Standards for Forensic Services No. 1, paragraph 10, provides that "[t]he ultimate decision regarding the occurrence of fraud is determined by a trier of fact; therefore, a member performing forensic services is prohibited from opining regarding the ultimate conclusion of fraud," while expressly permitting "expert opinions relating to whether evidence is consistent with certain elements of fraud." The ACFE's CFE Code of Professional Standards, section V.B.2, provides that "[n]o opinion shall be expressed regarding the legal guilt or innocence of any person or party." These are membership obligations rather than rules of evidence. SSFS No. 1 does not reach forensic work performed within an attest engagement or under the tax standards, or internal assignments given by an employer to an employee member not in public practice, and a witness who is neither an AICPA member nor a Certified Fraud Examiner is not reached by either document at all.
Neither professional standard stops a forensic accountant at the records. The ACFE's own Interpretation and Guidance to section V.B.2 states that a fraud examination report may conclude that a person misappropriated cash, misrepresented a transaction or concealed funds, and may draw conclusions on every element of a fraud statute, "[b]ut this is where the CFE's conclusions must stop." The line is conduct versus legal guilt, not findings versus silence. The measurement question that often follows a completed trace — what the movement of funds is worth once liability is assumed — runs on different admissibility law and is covered by our Economic Damages Institute, frequently with the same accountant doing both halves of the work. This Institute covers whether the record underneath the number will get in, and in whose hands.