The note on OFAC’s penalty record says the banking sequence largely ends after 2019 and that a few large actions account for most of the money. This note puts numbers on the first half of that, using the same dataset, and then reads what the individual releases say about which programmes the money was for.
The nine
Between December 2009 and October 2015 OFAC published nine settlements of $100 million or more with banks over dollar payments and trade finance involving sanctioned countries:
| Date | Bank | Amount |
|---|---|---|
| Dec 2009 | Credit Suisse | $536.0m |
| Dec 2009 | Lloyds TSB | $217.0m |
| Aug 2010 | Barclays | $176.0m |
| Jun 2012 | ING | $619.0m |
| Dec 2012 | Standard Chartered (the 2012 settlement) | $132.0m |
| Dec 2012 | HSBC | $375.0m |
| Jun 2014 | BNP Paribas | $963.6m |
| Mar 2015 | Commerzbank | $258.7m |
| Oct 2015 | Crédit Agricole CIB | $329.6m |
Together they come to about $3.6 billion, which is 46 per cent of the $7.86 billion in the dataset of 284 published actions. Add Clearstream’s 2014 settlement of $151.9 million, a different case about a custody chain, and the share is 48 per cent. Calendar years 2009 to 2015 hold 53 per cent of all the money, and these ten settlements are 91 per cent of that period’s total.
Seventeen settlements of $100 million or more account for 90 per cent of the whole record. Nine of those seventeen are the banks above. After the 2019 Standard Chartered and UniCredit settlements, none of the remaining large actions in the dataset is a bank.
It was not mostly Iran
Six of the nine releases break the base penalty down by programme. The pattern is the same in most of them and it is not the one the public account suggests.
| Bank | Largest programme by base penalty | Its share of the base penalty |
|---|---|---|
| ING (2012) | Cuba | 96% |
| Standard Chartered (2012) | Sudan | 52% |
| HSBC (2012) | Burma | 36% |
| BNP Paribas (2014) | Sudan | 87% |
| Commerzbank (2015) | Iran | 42% |
| Crédit Agricole CIB (2015) | Sudan | 97% |
Only Commerzbank’s largest component is Iran. The others are Cuba, Burma and, three times, Sudan. The Sudan programme page records why that regime figures so heavily: it was in force in the same years, and the banks’ non-transparent dollar payment practices applied to every sanctioned jurisdiction at once.
Two more releases give counts, not base penalties. By value, 94 per cent of the transfers Barclays counted were Sudanese, and about 3,500 of the 4,200 transfers Lloyds counted were for a single Libyan customer.
A caution on the table. Base penalties are built per violation against a statutory maximum, so a programme with many small transactions can outweigh one with a few large ones. HSBC’s Iran transfers were the largest by value ($164 million) and only third by penalty. The shares measure how OFAC’s formula treated each programme. They do not measure harm or intent.
What the releases say the banks did
The same methods recur across the releases and Treasury’s accompanying press statements:
- Written procedures and instructions. Barclays circulated standing instructions from 1987 and a fax template for re-inputting messages “without the offending text”; Lloyds had, in OFAC’s words, “a policy of intentionally manipulating and deleting information about U.S. sanctioned parties”.
- Cover payments and altered messages. Credit Suisse had “standard procedures for using cover payments to avoid referencing parties subject to U.S. sanctions”; Crédit Agricole’s banks “used cover payments and/or implemented special payment practices in a manner that omitted references to U.S.-sanctioned parties”.
- Removal or substitution of names. HSBC’s messages, according to Treasury, “falsely referenced an HSBC affiliate as the ordering institution”.
- Instructions from above. Treasury says ING’s staff in Curaçao were told by senior management to omit Cuba from payment messages.
- Routing through clients and documents. ING’s Dutch trade-finance business routed Cuban clients’ payments through other corporate clients, and its French office supplied false endorsement stamps for travellers’ cheques.
The pattern is wire stripping, and the techniques page on correspondent banking abuse describes how it is detected.
What the figures do not mean
The OFAC amounts are not additional to other penalties. For ING, HSBC, BNP Paribas, Commerzbank and Crédit Agricole the releases say OFAC’s obligation was deemed satisfied by payments to the Justice Department, the New York County District Attorney or other US authorities for the same conduct. The dataset counts OFAC’s figure, so the dollars here overlap with amounts other agencies collected.
It is not a measure of what banks did. It counts what was found, settled and published by one authority. Anything resolved without publication is absent.
The dataset is not a causal account of the later decline. The note on the whole record offers two readings, that the settlements worked or that the conduct moved, and these numbers do not decide between them.