USSG § 2B1.1 loss-amount disputes: actual vs. intended loss in the Fifth Circuit

What § 2B1.1 actually requires

USSG § 2B1.1 governs fraud, theft, embezzlement, and related property-loss offenses. The base offense level under § 2B1.1(a) is generally 6 or 7. The most consequential variable is the loss-amount enhancement at § 2B1.1(b)(1), which adds offense levels in steps based on the dollar amount of loss.¹

The loss table at § 2B1.1(b)(1) runs from +2 levels for loss over $6,500 up to +30 levels for loss over $550 million. Each enhancement step compounds the guideline range — a difference of $50,000 can shift the offense level by two and the guideline range by months.

Application Note 3 to § 2B1.1 defines loss as the greater of actual loss or intended loss. That definition is where most loss-amount disputes happen.²

Actual loss vs. intended loss

The Application Note 3 definitions:

Actual loss

The reasonably foreseeable pecuniary harm that resulted from the offense.

Intended loss

The pecuniary harm that the defendant purposely sought to inflict; intended loss includes intended pecuniary harm that would have been impossible or unlikely to occur (e.g., an attempted fraud that would have failed in any event).

The guideline uses the greater of the two. In a routine fraud case where actual losses were less than what the defendant attempted, the intended-loss figure controls. In a case where the defendant succeeded in causing more harm than initially intended (e.g., a scheme that snowballed), the actual-loss figure controls.

The mismatch between actual and intended loss is the focal point of many sentencing disputes. PPP and EIDL fraud cases provide a recent example: a defendant who applied for a $300,000 PPP loan and received $150,000 has actual loss of $150,000 and intended loss of $300,000. The intended loss controls.

Fifth Circuit doctrine on loss

The Fifth Circuit has developed substantial doctrine on § 2B1.1 loss calculations. Recurring themes include:

1. Reasonable foreseeability. The reasonably-foreseeable-pecuniary-harm test for actual loss focuses on what was foreseeable at the time of the offense, not what actually happened.
2. Intended-loss inference. The defendant’s intent at the time of the conduct controls the intended-loss calculation. Documentary evidence — submitted loan applications, invoices, contracts — typically establishes intent.
3. Credits against loss. Application Note 3(E) provides for certain credits against loss, including returned money or property recovered before the offense was detected. The credit analysis is fact-intensive.
4. Burden of proof. The government bears the burden of proving loss by a preponderance of the evidence at sentencing.

United States v. Rao, 123 F.4th 270, 285–88 (5th Cir. 2024) (the term "loss" in § 2B1.1(b)(1) continues to mean the greater of actual or intended loss; Kisor v. Wilkie did not abrogate Stinson's rule deferring to Guidelines commentary); United States v. Vargas, 74 F.4th 673, 680 (5th Cir. 2023) (en banc) (Stinson "has not been overruled or modified" by Kisor and still controls Fifth Circuit deference to Guidelines commentary). On the contrary-out-of-circuit view, see United States v. Banks, 55 F.4th 246 (3d Cir. 2022).

Recent guideline amendments affecting loss

The U.S. Sentencing Commission has amended the loss table and the Application Note 3 definitions several times in the past decade. The 2015 amendments restructured the loss table; subsequent amendments have addressed specific contexts including identity theft, securities fraud, and pandemic-relief fraud.

The 2023 amendments addressed § 1B1.13 (compassionate release) and some § 2B1.1 specifics. Counsel should always work from the current Guidelines Manual at ussc.gov, including the supplemental amendments, rather than from an older edition.

Common loss-amount disputes

The recurring litigation flashpoints in § 2B1.1 cases:

Aggregation

Whether multiple transactions should be aggregated into one loss calculation. The general rule is that all transactions within the scheme are aggregated, but the scope of the scheme can be litigated.

Relevant conduct

Whether transactions outside the count of conviction should be included under USSG § 1B1.3 (relevant conduct). Relevant conduct can substantially increase loss amount, particularly in plea cases with limited counts of conviction.

Restitution offset

Whether the loss amount should be reduced by money the defendant has returned or that the victim has recovered from third parties (insurance, recovery efforts).

Intended-loss inference

In fraud schemes, the government often presents inflated intended-loss figures. The defense can challenge the basis for the inference and the assumptions underlying the intended-loss calculation.

Number-of-victims enhancement

§ 2B1.1(b)(2) adds offense levels based on number of victims. The definition of “victim” under Application Note 1 has been litigated extensively.

Sophisticated-means enhancement

§ 2B1.1(b)(10)(C) adds two levels for sophisticated means. The application of this enhancement turns on whether the conduct involved special skill, planning, or technology.

Defense strategy at sentencing

The defense move in a § 2B1.1 case is to litigate every loss-driving variable at the sentencing hearing.

1. PSR objections. The presentence investigation report (PSR) drives the court’s loss calculation. Specific objections to the PSR’s loss figures, with supporting evidence, are the starting point.
2. Sentencing memorandum. A loss-focused sentencing memorandum can be 20+ pages and can include charts, alternative calculations, and case citations.
3. Evidentiary submissions. Documents that contradict the government’s loss figure — corrected invoices, recovery records, scope-of-conspiracy evidence — should be submitted as exhibits.
4. Live testimony. In high-stakes loss disputes, live testimony from forensic accountants, victim representatives, or co-defendants can shift the calculation. The hearing on loss is often the most important hearing in the case.
5. § 3553(a) variance arguments. Even if the guideline calculation stands, the court can vary under § 3553(a). A variance argument premised on the unfairness of the loss calculation — for instance, where intended loss vastly exceeds any plausible actual harm — can drive a below-guideline sentence.

Pandemic-relief fraud — the recent litigation surge

PPP and EIDL fraud prosecutions have produced a recent surge of § 2B1.1 loss litigation in the Fifth Circuit. The recurring loss-calculation issues in these cases:

Loan amount vs. loan-forgiveness amount

The intended loss in a PPP-fraud case is typically the total loan amount, not just the portion the defendant ultimately retained. A defendant who applied for $200,000 and received $150,000 generally faces a $200,000 intended-loss calculation.

Forgiveness applications

A subsequent fraudulent forgiveness application can add a separate count and a separate loss computation. The aggregation analysis can include both the loan and the forgiveness.

EIDL grants

EIDL advances (the targeted grant portion) are calculated as part of intended loss even though they were structured as grants rather than loans.

Multiple-PPP loans

A defendant who applied to multiple banks for multiple PPP loans has each loan counted toward the loss figure. The total can rapidly exceed $1 million, with corresponding offense-level enhancements.

Number-of-victims enhancement

For PPP/EIDL cases, the victim question is litigated. The U.S. government may be a single victim or, in some readings, the participating banks and the Small Business Administration may be multiple victims.

The defense response often focuses on what the defendant actually retained (the actual-loss number) versus the program-funded amount (the intended-loss number), and on the § 3553(a) variance arguments around the policy rationale for treating intended loss as actual harm.

Mitigation strategies in loss-driven cases

Beyond technical disputes over the loss calculation, several mitigation strategies can reduce sentencing exposure.

1. Pre-sentencing restitution payments. Voluntary restitution paid before sentencing is sometimes credited as a mitigating factor under § 5K2.0 or supports a § 3553(a) variance.
2. Cooperation under § 5K1.1. If the defendant has provided substantial assistance to the government, a § 5K1.1 motion can produce a substantial departure below the loss-driven guideline range.
3. Acceptance of responsibility under § 3E1.1. The full three-level acceptance reduction is typically available for early-pleading defendants who admit conduct fully.
4. Role adjustment under § 3B1.2. In multi-defendant fraud schemes, the defendant’s role can support a minor or minimal role adjustment that reduces the offense level.
5. Aberrant-behavior departure under § 5K2.20 or related grounds. Some single-instance fraud cases qualify if the conduct represents a marked departure from the defendant’s normal behavior.
6. Compassionate-release framing. For sentences imposed before the 2023 § 1B1.13 amendments, post-sentencing compassionate-release motions can reduce sentences based on extraordinary and compelling reasons.

Plea strategy in § 2B1.1 cases

Loss-amount negotiation is one of the most consequential pre-plea conversations in a federal fraud case. The recurring negotiation moves:

1. Pre-indictment loss stipulation. In cases where the U.S. Attorney’s Office is open to pre-indictment negotiation, the loss amount can be specified in the plea agreement at a level lower than what the government might prove at sentencing.
2. Stipulated count of conviction limiting relevant conduct. A plea to a narrower count of conviction, with a specific stipulation that relevant conduct will not be expanded beyond the count, can substantially reduce loss exposure.
3. Joint sentencing recommendation. A negotiated joint recommendation for a specific sentence (with §3553(a) factors and variance grounds identified) can constrain the court’s analysis without binding the court.
4. Restitution agreements. The plea agreement can specify the restitution figure, which may differ from the loss-calculation figure for sentencing-guidelines purposes.
5. Cooperation agreement. A § 5K1.1 cooperation agreement can produce a departure motion at sentencing that effectively overrides the loss-driven guideline range.
6. Specific stipulation on enhancement factors. The plea agreement can stipulate or contest specific guideline enhancements (number of victims, sophisticated means, abuse of trust) at the time of plea.

What to do if you are facing a § 2B1.1 case

The pre-sentencing tasks are document-intensive.

1. Obtain the PSR draft and identify every loss-driving variable.
2. Run the government’s loss calculation independently. Identify discrepancies, double-counting, or transactions that should not be in the calculation.
3. Identify credits — money returned, insurance recoveries, third-party recoveries.
4. Identify alternative theories of loss that produce a lower figure.
5. Identify potential witnesses for the sentencing hearing — accountants, co-defendants, victims who have been made whole.
6. Draft a sentencing memorandum that addresses every loss-driving variable and proposes a specific lower loss figure with supporting evidence.
7. Identify variance grounds under § 3553(a) for a below-guideline outcome even if the guideline calculation is upheld.