What is the key issue facing global banking industry today? Ask this question to the CEO’s of Top 50 global banks and it is highly likely that their answer will be the same – managing balance sheet risks and effective utilization of Economic Capital! While this problem is not new now and banking industry has already seen the need for more stringent capital management standards, there is no practical solution in sight! As a banker or a senior leader in a bank, you may have provisioned for enough capital or your bank may ace the stress testing requirements on annual basis, but does that mean you are managing your risks and capital effectively? Is your portfolio risk optimized? Are you recovering the cost of your risks from your customers or are you passing on the risks that you should not have on your balance sheet effectively?
Answers to all these questions are unknown (or at least most bankers are not confident about them) today because of many reasons. Firstly, it is not possible to measure risk from different asset classes on a common platform. Secondly, the complexities of banking transactions are ever increasing and with that increases the model risks or the risks of going wrong with your assumptions in quantifying the risk! While all these problems continue to persist, one thing that has become extremely important for banking industry is timeliness of information regarding risk and capital. How soon can a Chief Risk Officer (CRO) act on the available information? How soon can a CRO find out that a certain part of his bank’s portfolio is fast becoming toxic asset? How soon can a bank leadership team find out that some of their top performing loans are quickly turning into a nightmare? How soon can they act on hedging a certain loan? And how soon can they recover the cost of hedging such risk from the customer?
Any possible answer to any of these questions would have been a post-mortem 2 years back. It would have required tons of analysis from the Risk Management gurus of the bank and would have taken them ages to find out the right path of action. By the time they take a decision and act, the losses would have already hit their balance sheet. However, it is possible to visualize such scenarios today – with help of Big Data Analytics. It is possible to create a risk (by lending money or investing in certain securities) and hedging them (with almost no incremental cost to the bank) on real-time basis using Big Data Analytics.
Risk Adjusted Performance Management – a tool that many banks wish they had before 2008 recession, is available today. With Big Data Analytics, banks can build platforms that give the power to their front-line staff to manage the balance sheet risk. The classic paradigm in banking is when a loan sales officer has to balance between the customer requirements and the risks the bank can take. In past, if sanctioning a loan meant the bank increased its portfolio risk and in turn its economic capital requirements, it either chose to not issue the loan or hedge it afterwards – eventually putting pressure on its margins due to increasing costs of hedging.
However, using a Big Data platform, banks can adapt to Risk Adjusted Performance Measurement for their front-line sales or underwriting officers. Consider this scenario:
A loan sales officer of a large bank is sitting in front of an existing client. This client – an automobile giant – has already borrowed a couple of billion dollars from the bank. This exposure has already increased the concentration or correlation risk for the bank’s portfolio (because of many such loans issued to other automobile companies too) and is pressurizing the margins of the bank. The client is asking for an additional loan facility of half a billion dollars more for expansion of its business in Asian markets. The sales officer knows that the credit team of the bank will not approve this because of the incremental capital requirement to match the incremental risk or the hedging costs of the loan would reduce the net spread significantly. He pulls out his handheld phone and opens an application provided by Office of the CRO of the bank. He feeds in the details of the client and his requirements in the application and hits a button to run simulations. His phone connects to the servers of the bank located some 1,000 miles away and in less than a minute returns multiple scenarios. One of the scenario tells the sales officer that the bank will require x million dollars as incremental economic capital and will reduce the net spread of the transaction by a certain basis points. The second scenario tells him that the loan can be hedged using a Credit Default Swap (CDS) at a cost of y%. However, if he chooses to hedge the loan, the total spread (or interest rate in simple terms) that he should charge the client is z% instead of the normal r% interest offered by the bank.
If the bank policy requires the sales officer to generate a certain margin on a risk-adjusted basis (performance management for staff) and the z% interest and the net spread of the transaction falls within this threshold, he can go ahead with the loan. However, if the threshold is not met even after hedging the transaction, he can let go the deal. While banks and bankers may not like to let go deals, there is no choice. A risk adjusted performance management is the need of the hour – and it can be implemented with Big Data solutions.
This will help the banks in improving:
- their credit quality or asset quality
- their net spread on assets (net profit of the bank in other words)
- reduce their balance sheet risk
- manage their economic capital effectively
- ace the stress testing requirements (not only statutory but also internal) and
- give confidence to the leadership team of the bank that they are far away from the risk of a sudden shock in their asset book
All these on real-time basis – possible! Big Data Analytics can make it possible – and it’s not a big change management exercise. It’s about deploying the front-end technology using the back-end risk engines that banks already use today. The business case for the investment may show a net loss for first couple of years, but will benefit the banks over period and will certainly improve the functioning of overall banking industry.
Risk Adjusted Performance Management is the only way out – and it is possible – with Big Data!
Try deep learning using MATLAB