Is Trader Fraud Finished or has it just been automated?
I don't think anyone is ready to attest that going the best part of six years since UBS admitted that Kweku Abdoli had lost them billions in unauthorised risk taking indicates that financial markets have cracked the rogue trader risk for good. But I think it is worth reviewing changes over the last decade which do suggest the industry is now less likely to suffer incidents of this nature.
Cultural Issues
· The major banks boards and senior management are now expected to have far better operational understanding of their risks and issues
· Traders are not seen as the top dogs who cannot be challenged. It's clearly understood that such attitudes helped conceal frauds in the past.
· The demise of prop desks and the introduction of Volcker Rule monitoring mean that proprietary gains can be problematical rather than a benefit for an organisation.
Practical Changes
· The necessary controls around a trading operation are much better understood, and not something for a debate with the traders. The necessary controls are now much more documented science than an art. Strange trade, cancel and amend scrutiny are much tighter. Macro data analysis is applied by some organisations although I've yet to hear of an example of it detecting a fraud.
· Banks' capital requirements for trading positions are now much higher and so there is less incentive for management to endorse opening a new trading activity without considering all the consequent requirements and controls.
· Similarly new liquidity rules mean that banks have to have a much better understanding of the detailed positions requiring funding.
· Bonus claw-back rules for traders, their supervisors and senior control function managers have made the penalties for failing to supervise and control traders much more significant.
· Most large banks have centralised and off-shored their key control support junior functions to the cheapest locations they can find worldwide. These staff now have a detached relationship with the traders and the likelihood of them turning a blind eye to a trader's transgression is very small.
· Operational risk functions are vastly larger with more clout than previously. ORF frameworks make it easier to highlight key control weaknesses.
· The growth of electronic confirm/affirm of OTC trading via services like MarkitWire reduce the scope for fictitious OTC trade booking.
But Some things haven't improved much
· Business supervision of individual trader activity still feels like lip service.
· Internal audit and operational risk rarely have the detailed knowledge of large organisations workflows to effectively identify all the weaknesses.
· Confirmation controls still struggle to prioritise existence and accuracy over complete affirmation.
· I doubt any major organisation is in a position to analyse its cashflow requirements to detect a fraud, other than like Barings when it threatened the institution itself.
Today
Recent years have focussed much more on traders colluding to rig indices and markets. Equally, as the trading job becomes fully digitised perhaps we would be better tracking and preventing rogue trading programmes like the incident at Knight Capital in 2012 that cost the firm $440m. AI trading algorithms using complex pattern analysis will not be able to explain their strategies to their peers, just as the world beating new Go programmes develop moves never seen before in human or computerised games. One of the more ironic consequences in some institutions of creating programmes to perform human functions is the extreme difficulty in getting them to comply or receive exemption from the myriad of mandatory compliance training courses their human colleagues must complete to receive their trading login.
Major Trader Frauds
|
Company |
Individual |
Year |
Location |
Loss |
Business |
Financial Instrument |
Head Office |
Technique |
Fraudster Sanctions |
Company Sanctions |
Discovery Method |
Comment |
|
Knight Capital |
Power Peg |
2012 |
Jersey City |
$440m |
Equity trading |
Equities |
Yes |
Market simulation test programme which moved stock prices, erroneously operated in production |
NA |
$1.5m fines and compensation |
Market disruption. |
Capital injection required. Error took 45 minutes to identify and close. |
|
UBS |
Kweku Abdoli |
2011 |
London |
₤2.3Bn |
Delta1 Equity trading |
Equity futures |
Yes |
Prop position concealed by fictitious OTC trades |
7 years prison |
₤29.7m fine |
Forward settlement confirmation controls reinstated. |
Co-traders knew about some of the concealment but kept quiet. |
|
Society Generale |
Jerome Kerviel |
2008 |
Paris |
$7.2Bn |
Delta 1 Equity trading |
Equity futures |
Yes |
Prop position concealed by fictitious OTC trades |
5 years prison €4.9Bn fine |
None |
Fictitious trades caused capital calculation problems which were investigated. |
Exchange wrote querying activity, management failed to consider the query. |
|
Credit Suisse |
Kareem Serageldin |
2008 |
London/ New York |
$2.65Bn |
Structured Credit trading |
Mortgage Backed Securities |
Yes |
Distressed positions falsely marked |
30 months prison $26.6m fine |
Accounts restated |
A departing trader pointing out valuation inconsistencies |
|
|
Goldman Sachs |
Matthew Marshall Taylor |
2007 |
New York |
$118m |
Equity derivatives |
Equity futures |
Yes |
Prop position concealed by fictitious trades |
9 months prison |
$1.5m fine |
Standard position controls within days. |
|
|
National Australia Bank |
Luke Edward Duffy |
2004 |
Melbourne |
$360m |
FX Derivatives |
FX spot and derivatives |
Yes |
Trade alteration to move profits plus fictitious trades |
29 months prison |
|
Whistleblower |
Aggressive trader responses to any control challenges was tolerated by management. |
|
Allied Irish Bank |
John Rusnak |
2002 |
Baltimore |
$750m |
FX spot and derivatives |
FX forwards and options |
No |
Prop position concealed by fictitious OTC trades |
7.5 years prison |
|
OTC confirmation controls improved. |
Managers and internal auditors dismissed. Operation sold and downsized. |
|
Nat West |
Kyriacos Papouis |
1997 |
London |
£91m |
Interest rate derivative trading |
Swaptions |
Yes |
OTC Swaptions mis-priced |
|
£0.4m fine |
A new trader taking over the portfolio identified the mis-marking of implied volatility |
The IPV function was not technically up to the job on the non-linear derivatives. |
|
Sumitomo Corporation |
Yasuo Hamanaka |
1996 |
London/ Tokyo |
$1.68 |
Commodities |
Copper futures |
Yes |
Prop positions concealed in unauthorised accounts |
8 years prison |
$150m fine |
Unauthorised accounts discovered by company. |
Several of their trading counterparties were also fined for failing to verify the unauthorised accounts correctly. |
|
Barings Bank |
Nick Leeson |
1995 |
Singapore |
$1.4Bn |
Listed derivative client sales |
Equity futures |
No |
Prop position concealed as client position |
6.5 years prison |
Barings went bust, sold to ING for ₤1 |
Treasury investigation of large funding requirement |
Multi-year fraud. |
|
Daiwa Bank |
Toshihide Iguchi |
1995 |
New York |
$1.1Bn |
US government bond trading |
Government bonds |
No |
Client assets stolen to cover trading losses as trader was also client asset back office manager. |
4 years prison |
US banking licence withdrawn. |
Confession |
Fed audits had uncovered lack of segregation, Daiwa failed to correct issue. After discovery Daiwa were still seeking to conceal event and delayed informing Fed. |
|
Kidder Peabody |
Joe Jett |
1994 |
New York |
$350m (false profits) |
US Treasury trading |
Treasury strips |
Yes |
Forward settled trades were not present valued correctly . |
$8.4m fine |
Operation was closed. |
Management slowly realised the nature and size of the profit recognition error. |
|
© Greg Stevens 2017