Sunday, September 19, 2010

Auditing Derivatives:Think of what can go wrong- The Trading Floor


This article appeared in The Malaysian Accountant journal , June-July 2010 issue

In the current crisis of derivatives, the job of auditors, both internal and external is being put to the test. There is always the danger that even the most thorough job can leave some holes uncovered, which may (and normally does) blow up one day. While it is imperative to understand the business and how derivative products work, a crucial prerequisite is for auditors to have the right mindset.
Thinking of what can go wrong is a sound foundation to detect potential weaknesses. This is even more so for derivatives in the investment banking industry, which are constantly increasing in complexity.
Countless recent case studies can point us to potential areas of weakness. This series of articles will explore the things that can go wrong by analysing the various functions within an investment bank.  To start, let us begin where all trades originate and profits (and losses) are generated – the trading floor.

Traders’ risk limits increased when he continuously outguesses the market
Trading limits are sometimes breached denoting excessive risk taking. But a bigger worry is when
management increases a trader’s limit when he continues to bring in profit. No trader gets it right all the time. Nick Leeson of Barings brought in huge profits at the beginning but his luck did not continue.  Increasing the trading limit gives the trader a higher limit indefinitely (as seldom is the limit lowered). When things go wrong, the excessive risk will lead to an even bigger loss.
Besides looking for the usual trading limits breaches, the auditor should also watch for significant increases in trading limits. This may be a more useful approach in detecting potential problem areas.

High performing traders viewed as “untouchable”
Somehow there is always a tendency for a handful of traders who are bringing in the profit for the firm to be treated in a superior manner. Their activities may not be subjected to the same scrutiny as other traders.  
There must be no exceptions. For example, putting new products through the mill in seeking approval from all departments cannot be compromised.
A trader who is too “busy” to answer questions all the time indicates something suspicious. At the same time, knowing which trading team contributes to most of the firm’s profits will help the auditor to keep his eyes and ears open to budding issues.

Fake trades
It is surprising that the Societe Generale’s trader was able to allegedly make-up fake trades in today’s control environment where double confirmations are a norm.
It is highly challenging for the auditor to capture fake trades. Besides sending end-of-period balance confirmations, specific trade confirmations (possibly abnormally large trades or randomly chosen) may point to hidden issues or agendas.  In addition, specific health checks which actively search for signs of fraud could be designed.

Trader changed from hedger or arbitrageur to speculator
A hedger, arbitrageur and speculator have different trading methods and objectives in achieving revenue for the firm. Naturally the speculator is given the responsibility to take the greatest risk on behalf of the firm. A hedger or arbitrageur transforming into a speculator should only happen with proper authorisation. What should ensue then is proper control of the size of positions and reporting and monitoring of the risk being taken.
Unauthorised cases are of course tricky to detect. An effective approach could be to look for risk reports that point to unusually high risk by certain traders and follow up by asking probing questions.

Blindly trusting models and frequent changing of models
Incorrect models have caused many losses, the latest being the underestimation of remote default events in the valuation of collateral debt obligations (CDO). When different models are used by different banks to value the same product (like the CDO), it signals a huge mispricing in the market.
Frequent changing of models for the valuation of a product often implies the abuse of models to fit the valuation needed.
Besides ensuring ample buffers, review of the models is paramount in identifying where grey areas might lie. In most cases this is when the parameters are not marked-to-market and cannot be independently verified. The next step will be to observe how sensitive these parameter are by performing scenario analysis and stress tests.

Selling client inappropriate products
We have been too aware of clients with large appetite for big risks and rewards in the 1990s and early 2000s. The Milan interest rate derivatives and Goldman Sachs Abacus deal fits both.  Investment banks were selling clients inappropriate products which had excessive underlying risk that they did not realise.
As auditors, the interest of clients is probably not of utmost interest. But the repercussions tend to hit back at the banks when something goes wrong and the client takes action.  With this in mind, auditors could pick some big and (usually) complex trades and review all paperwork and correspondences for any possible alarms.

Everyone following the same tactic
When everyone follows the same trading strategy, there is bound to be some major market moves. In May 2005, in the midst of the correlation product era, the downgrading of Ford Motors and General Motors caused the market to panic resulting in big moves in spreads and mispricing in the correlation market. To a larger degree, the credit derivatives crisis in 2008 caused a major flight to liquidity.
Entrepreneurs believe that too much trading of the same product is just as worrying as too little trading of the same product. For the auditor, when market participants are undertaking the same trading tactic, it is important to see the big picture to understand the risks inherent in these situations. If not anything else, at least a warning signal could be given to the management.

Conclusion
It is always challenging for the auditor to catch a problem before it occurs, what more in good times when banks are competing for hefty profits and traders for big bonuses. Perhaps by thinking of what can go wrong from all angles, the auditor has a better chance of detecting potential trouble. However, a right balance must be struck by him to be effective and at the same time far-reaching and thorough.
Subsequent to the trading floor, things can equally go wrong in the areas of risk management, modelling and valuation areas in the bank. These will be discussed in future publications.

Jasvin Josen is a specialist in developing methodologies for valuation of various derivative products. She has over 10 years’ experience in investment banking and the financial industry in Europe and Asia.

No comments:

Post a Comment