Tuesday, October 5, 2010

Auditing Derivatives: Think of what can go wrong Middle Office (Controlling Group)

This article appeared in The Malaysian Accountant journal, Aug&Sep 2010 issue

The previous article showed that thinking of what can go wrong can shape the auditor’s mindset to better detect areas of weakness, more so when reviewing activities in an investment bank trading predominantly derivative instruments. After discussing the trading floor at length in the last article, we now move to the middle office, a significant function of the bank.

The Middle Office
The middle office, being the largest section in the bank comprises the following essential units:
-          Controlling Group (P/L and Risk)
-           Risk Management
-          Model Validation
-           Independent Price Valuation (IPV)
Different banks tend to organise the individual units differently. In some banks, the Controlling Group is combined with the IPV group to be more effective in controlling traders’ activities and to avoid redundancy. Other banks may prefer the IPV group to maintain its independence and report directly to management.  Some banks also tend to have the Risk Management and Model Validation in the same unit. Nevertheless, we will explore the above units individually. This article will focus on the first unit, the Controlling Group.

The Controlling Group (P/L and Risk)
This group, as the name suggests, acts as controller for the bank. It controls what goes into the P/L trading book and reports daily profit and risk on trades. Besides channelling trades to relevant books, it also ensures that the systems adequately capture trades for computation of profit and risk.
The Controlling Group deals with almost every department in the bank. For that purpose, this group is not to be underestimated. Many potential problems can be unearthed and solved right here, before they get further.  
But what could go wrong?
Risky and problem trades go pass the Controlling Group undetected
In the “good” times of the early 2000s, a risky or unusual trade that should have concerned some individuals often just passed through the Controlling Group as usual. To illustrate with an example, Lehman Brothers was holding some very risky equity sub- prime mortgage CDO tranches that ultimately led to its demise. Surely at some point someone had seen the risk growing and could sense some danger. But the controlling employee would not probe further probably because he did not feel it was his responsibility. The situation worsens if employees do not have the necessary experience and knowledge. With structured and exotic trades, controllers may also struggle with unnecessary jargon used by traders. Traders on the other hand feel that the profit of the bank is entirely theirs and should not be questioned.
Management is slowly realising the above situation to be an issue. In many organisations, routine work is being sourced out to allow key staff to spend more time on review and analysis work. For this to work, control staff should have full management backing and must be encouraged to question traders or anyone in the bank on any problematic transaction. Most of all, the whole Controlling Group must be convinced that they are “stakeholders” of the bank and not merely a cost centre. This effectively comprises a culture change for the whole bank.
It is useful for the auditor to get a sense of the culture and attitude of the Controlling Group. A questioning session with the management and the group will indicate the general atmosphere. If change is not underway, the auditor will want to be more cautious in his work. Reviewing for a presence of a skilled team (for example a Modelling team) within the Control Group to whom exotic and complex trades are referred will help reduce the inherent risk. This team will be more prepared to question transactions and modelling issues with the traders and Quantitative Analysis team.
Huge Day 1 Profits unexplained
In the present day, considerable effort is put to assess if Day 1 profits can be recognised upfront in the financial statements. Upfront profit recognition is only allowed when the attributed profit comes from movement in observed market parameters. For example the USD currency, LIBOR or Ford credit spread shifts. Less obvious parameters like credit correlation of a bespoke first-to-default basket will not make it to the P/L on Day 1.
However with all the above checks in place, huge undetected Day 1 profits were still being recognised in the last few years. The Greek tragedy involving cross currency swaps were manipulated such that the Day 1 profit served as an additional USD 1 billion debt to the country. In the Milan Council Derivative case, interest rate collars used as hedges for municipal debt contained banker’s fees masked as Day1 profits so that the amount was undetectable through normal channels.
The Control Group’s task is to explain the Day 1 profits before booking them in the books. When trades keep pouring down the pipeline as they did in the high times of the 1999-2008 credit decade, this can be quite a struggle.
It is dangerous for banks to recognise certain types of Day 1 profits upfront although it may be legally permissible. It encouragers traders to use aggressive models, take their bonuses and leave before the model and the value of the deal come under close scrutiny.
Therefore it is imperative that auditors review all trades’ Day 1 profit of trades, in order to gauge and identify unusually large profits. All profits must be logical, based on parameter shifts and not contain any unexplained elements.
Extremely risky structures go through the New Product Approval (NPA) Process unscathed
Some new and significant trades that were approved by the NPA committee some time back are proving problematic now. The same problems probably also surfaced during the NPA process but went unnoticed with the pressure this team was under.
It is a norm that banks have a special NPA committee which approves new trades within the bank. This committee primarily consists of the heads of Risk Management, Model Validation, Quants Team and Control Group. The main objective is to ensure that the new trade carries profit and risk that are justified and can be booked in the systems accurately. In short, it is hoped that this committee thinks of all potential problems upfront.
The system worked well in the past, despite traders feeling that the process slows things down. At times, by the time everyone’s approval is sought in the committee, the trade would have gone into competitor’s hands.
As a result, there is tremendous pressure on the NPA committee, amongst their daily responsibilities, to avoid erroneous approval of new trades. Moreover, these trades are almost always complexly modelled and carry multi-faced risk. Reserving for these trades can be very challenging. The result:  NPA approval is often not considered thoroughly resulting in potential excessive risk taking by the bank.
For the above reasons, the auditor may not want to place utmost reliance on the NPA approvals unless there is other evidence to support their judgement.
In the midst of the overflowing information that the auditor receives from the Controlling Group, it is extremely difficult to pick up all the right areas of weakness. In this regard, the ability to see the trees from the forest when thinking of things that can go wrong is imperative.   
In the next article we will progress to more technical areas within the Middle Office. The Risk Management function, one of the most important functions in the bank will be examined.
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