Harry Cassin
Publisher and Editor

Andy Spalding
Senior Editor

Jessica Tillipman
Senior Editor

Richard L. Cassin
Editor at Large

Elizabeth K. Spahn
Editor Emeritus

Cody Worthington
Contributing Editor

Julie DiMauro
Contributing Editor

Thomas Fox
Contributing Editor

Marc Alain Bohn
Contributing Editor

Bill Waite
Contributing Editor

Shruti J. Shah
Contributing Editor

Russell A. Stamets
Contributing Editor

Richard Bistrong
Contributing Editor

Eric Carlson
Contributing Editor

Bill Steinman
Contributing Editor

Daniel Patrick Wendt: Did the DOJ give SocGen an unofficial discount?

On June 5, Société Générale SA (or SocGen) agreed to settle Libya-related corruption allegations in a coordinated resolution with U.S. and French authorities. As part of the settlement, the company entered into a Deferred Prosecution Agreement (DPA) with the Department of Justice over FCPA charges, while one of its subsidiaries, SGA Société Générale Acceptance NV, pled guilty on a count of conspiring to violate the FCPA. 

A closer review of the DPA (pdf) in this case reveals an interesting wrinkle in the application of the U.S. Federal Sentencing Guidelines against SocGen.

The DPA is somewhat unique because it covers two distinct activities with separate criminal fine calculations for each activity — first, transactions with an intermediary linked to the Libyan government during the end of the Gaddafi regime (giving rise to FCPA violations); and second, separate SocGen activities intended to manipulate the LIBOR exchange rate. In particular, there is a notable difference between the separate FCPA and LIBOR-related criminal fine calculations in the DPA.

As background, the process for calculating a criminal fine range under the Sentencing Guidelines involves three phases. First, the Sentencing Guidelines provide guidance on establishing a “base fine.” In practice, however, this step is usually moot because the Sentencing Guidelines allows the DOJ to set a company’s base fine using the “pecuniary gain” from FCPA violations, and the pecuniary gain often exceeds any base fines otherwise recommended under the Sentencing Guidelines (at least for large resolutions comparable to the SocGen DPA).

Second, the Sentencing Guidelines provide instructions for calculating a “culpability score” and multiplier range (the highest range is 2.00 to 4.00). The culpability score takes into account whether the business at issue was substantial, whether senior management was involved, whether the government official who received the bribe was high ranking, whether the company had an effective compliance program, whether the company cooperated with the investigation, and more.

Third, the Sentencing Guidelines call for applying the culpability multipliers to the base fine. For example, if the base fine is $1 million, and the multipliers are 2.00 and 4.00, then the Sentencing Guidelines criminal fine range would be $2 million to $4 million. In practice, in negotiating DPAs, the DOJ typically uses the lower end of the Sentencing Guidelines range to set the criminal fine and reduces it further through the application of any discounts for voluntary disclosures, cooperation, etc.

In Para. 7 of the SocGen DPA, the DOJ calculates the fine for the FCPA charge and fixes the Sentencing Guidelines multiplier range at 1.4-2.8. In doing so, the DOJ stated that “the organization had 1,000 or more employees and an individual within the high-level personnel of the organization participated in … the offense,” which added 4 points to the culpability score under Sentencing Guidelines § 8C2.5(2)(A). After reducing the fine by 20 percent to recognize SocGen’s cooperation, the parties agreed on an FCPA fine of $585.5 million. 

In Para. 8, the DOJ calculates the fine for the LIBOR-related charge and fixes the Sentencing Guidelines multiplier range at 1.6-3.2. Unlike Para. 7, the DOJ stated that the “the organization had 5,000 or more employees and an individual within the high-level personnel of the organization participated in … the offense,” which added 5 points to the culpability score under Sentencing Guidelines § 8C2.5(1)(A).

Reducing even one point in a culpability score in a large resolution like this leads to a significant difference. If the DOJ would have used the higher multiplier range for the FCPA charge, it would have increased the criminal penalty by more than $83 million, or by approximately 14 percent. Put differently, by using the lower Sentencing Guidelines range, the DOJ reduced the potential fine from $669 million to $585.5 million, or by 12.5 percent (assuming the DOJ would apply the same 20 percent reduction for SocGen’s cooperation).

Based on the papers, it is unclear how to explain the difference between paras. 7 and 8, without attributing it to negotiation between the parties. The parent company SocGen is the same “organization” involved in both the FCPA and LIBOR-related violations, and the Société Générale Group as a whole is a large organization with more than 100,000 employees overall. It seems safe to assume that SocGen employs far more than 5,000 people.

One potential explanation is that the smaller employee total used for Para. 7 is based on the unit involved in the FCPA charges (rather than the organization), indicating that the unit involved had between 1,000 and 4,999 employees. This is allowed under the Sentencing Guidelines, and in fact the DPA Statement of Facts focuses on one division of SocGen for the FCPA charge (namely, Société Générale Corporate and Investment Bank).  See, for example, Paras. 1, 2, 51-55, and 88-92, for mentions of this division.

However, if the DOJ were relying on the use of a separate division (or “unit,” as defined under the Sentencing Guidelines), the DOJ would have used different citations and language than what is used in Para. 7. It should have said that a “unit of the organization” had more than 1,000 employees and cited § 8C2.5(2)(B) rather than just referring to the “organization” and citing § 8C2.5(2)(A). Nonetheless, the DPA — as filed with the court — includes handwritten edits to other parts of the Sentencing Guidelines calculations and citations, indicating last second corrections before filing with the court. So it is possible that the citation in Para. 7 may have been incorrect as well, but neither the DOJ or SocGen’s counsel deemed it worthy of a correction.

Another potential explanation is that the DPA is using the lower range based on the employee totals for the SocGen subsidiary that pled guilty to the related conspiracy count (namely, SGA Société Générale Acceptance, NV). In other words, the “organization” referenced in the DPA for the FCPA fine is not SocGen, but the subsidiary pleading guilty. However, this seems unlikely.

First, much of the conduct at issue was performed by SocGen employees, not employees of the subsidiary.

Second, the subsidiary that pleaded guilty has a very different calculation of its culpability score under the Sentencing Guidelines, without any additional points based on the number of its employees (which indicates to me that the DPA did not lift its calculation from the guilty plea).

For companies like SocGen that do not voluntarily disclose an FCPA issue, the maximum discount that the DOJ will bestow in a DPA is 25 percent, based on current DOJ policy. Here, SocGen received a 20 percent reduction from the lower end of the Sentencing Guidelines criminal fine calculation, for a total fine of $585.5 million. I initially wondered whether SocGen had negotiated an additional, unofficial 13 percent discount through the negotiation of a lower Sentencing Guidelines culpability score. After reviewing the papers, however, I suspect instead that the DOJ agreed that the conduct at issue was limited to a specific unit within SocGen, which would justify the lower culpability score for the FCPA fine. 


Daniel Patrick Wendt, pictured above, is a Member in Miller & Chevalier’s International Department. He focuses on matters involving the FCPA and U.S. customs laws. He can be contacted here.

Share this post


Comments are closed for this article!