Anshuman, can you please tell the Center for Financial Professionals’ audience about yourself and your professional experience?
I am currently a Managing Director at Moody’s Copal Amba and head their Risk Management Services Practice. In my current role, I look after important consulting engagements for a number of strategic clients cutting across various risk and regulatory issues. However, given my prior background I tend to specially focus on regulations in Market Risk and Stress Testing projects. The clients I work with range from large global banks with substantial multi-asset class exposures and sizeable trading portfolios, to more localized banks with limited trading exposure. Prior to this I was the Head of Risk Modeling at CRISIL. Before I got into Risk, which as is the case for many of us, happened only post-crisis, I was involved on the derivatives trading support side of the business, managing projects in various areas including structuring, product control, and derivatives technology.
We are looking forward to you presenting at the Risk EMEA Summit where you will be focusing on the trading book and banking book. Can you give us a very brief overview of the trading book and banking book revisions within the revised Basel Market risk framework?
The trading book should be used by banks ideally only to incorporate assets which are held for trading as opposed to being held till maturity (for example, any derivative instrument used as a speculative or arbitrage tool to earn profits). The banks accept market risk for any adverse movement in market value of assets and accordingly calculate the Value At Risk at 99% confidence, for a 10-day horizon to account for market risk capital. The trading book assets are valued at their market values.
In contrast – the banking book is an accounting tool for banks to incorporate assets which are held to maturity (for example, corporate/retails loans). Here the banks typically accept credit risk and interest rate risk. However, the assets are valued at their book value and a write down in asset value happens only in case of defaults. Hence, a 99.9 % confidence level 1-year horizon calculation of risk is deemed sufficient enough to capture regulatory banking book capital requirements.
The primary issue with the above framework has been a lack of clear demarcation of boundary between the two books, which in the past enabled banks to shift assets to the trading book from the banking book (prior to the financial crisis) due to lower capital requirements and then do the reverse (post financial crisis) due to massive loss in market values and illiquidity of these ‘held for sale instruments’ in trading books, resulting in higher capital requirement for trading books as compared to banking books.
In order to prevent such capital arbitrage by banks, the revised market risk framework has set out to define clear demarcation between the two books and frame stringent rules in order to prevent transfer of risk between the two books (with exceptional cases requiring regulatory approval).
The BCBS committee has come up with a presumptive list of securities to be included for each of these books. Any re-designation of assets from one book to other must be approved by the senior management, after thorough internal review for compliance with internal policies; subject to prior approval by regulatory authorities. Even in such special cases the difference in capital is accounted for through an additional capital surcharge.
What implementation approach does Copal Amba advocate?
Given that the revised Market Risk framework has now been finalized, we advocate that banks start moving forward quickly to address the changes that have been proposed by the Basel Committee. We believe that in order to properly implement the changes suggested in the framework there will need to be coordinated effort across Middle Office, Risk, Finance, Trading and IT functions. Banks that invested early in having that right IT and data infrastructures are poised to reap benefits in the implementation phase. In our experience, change management projects around creating golden data sources or overhauling risk reporting and organizing risk data infrastructure, take up a bulk of the resources and time. If banks don’t have this in place by now then we suggest that they try and supplement internal resources/platforms by engaging external vendors and looking at off-the-shelf platforms. The modelling challenges, although many, are not insurmountable and are hugely dependent on having good quality data at the right level of granularity available. Apart from the tangible aspects there are also organizational challenges given the desk-level approval process that is now going to be the standard. This may potentially involve some internal reorganization which could be painful.
What are the key modelling challenges for banks within the trading book and banking book?
On the modeling side there are several challenges. One challenge is to calibrate the credit risk capital charge for a particular instrument recognized in the banking book to a corresponding default risk charge for a similar instrument recognized in the trading book. This is especially more complex for equity default risk charge because of lower correlation between bond market default spreads and equity returns. Another challenge is that due to varying liquidity horizons there are “cliff” effects which not only impacts modeling but has real-life considerations as well. There was initially quite a lot of debate in the industry around the introduction of expected shortfall as the preferred measure. However, as organizations have gone through multiple QIS exercises, it has become less of an issue
How do you see the role of the market risk professional changing over the next 6-12 months?
The role of market risk professional has been evolving fast post the financial crisis in accordance with Basel 2.5 regulations. We see this evolution pick up more steam as the framework takes a concrete shape.
While we believe that banks and financial institutions will eventually come to grips with understanding the framework itself, the real challenge seems to be in implementing this framework.
The demand for high quality data and increased processing power may arise from calculation of multiple market risk measures at various liquidity horizons and asset classes, Default Risk Charge, and Non Modellable risk factors. Hence a market risk professional will need to adapt and update his skills to not only meet the modeling challenges but also the software implementation challenges related to storage and complex processing of such large scale data. This may even demand use of cloud computing and non-relational data bases frameworks to process unstructured data.