Balancing act

Jodi Alperstein outlines how best practices for improved enterprise risk management start at origination.

“We have several clients who are computing the impact of a new loan or a deal that gets added to their portfolio in real time - and many more are planning to do this. Understanding the Return on Risk Adjusted Capital (RORAC) for any given deal is a huge competitive advantage”
-Jodi Alperstein

It is no secret that very few lending institutions have a comprehensive, accurate and timely view of their enterprise risk. The problem becomes more acute when exposures and portfolios are spread across divisions, geographies and asset classes, which they usually are. No matter how much regulatory pressure has been placed on risk management or how much business benefit can be derived from optimizing risk management processes, the end goal of having a comprehensive view of enterprise risk often remains elusive - in spite of spending a great deal of time and resources on the problem.

One of the root causes of this problem has to do with the way lenders originate and monitor their commercial loans. Why is this? Historically, the commercial loan process has been very paper-intensive and subject to human errors. Most processes were developed within silos separate from the rest of an organization's risk management solutions and lack sufficient controls. Understanding the true impact on the overall portfolio and to the top and bottom lines is done after the loan is issued, not before - and this remains the case today. With all of these manual and disconnected processes, it's almost impossible for most firms to evaluate a single transaction while factoring in the impact to the overall portfolio and the business model of the firm. 

Managing the day-to-day risk decision process is also a problem. Firms are continually challenged to improve corporate governance and optimize the way credit committees manage portfolio risk. Yet, the processes surrounding loan origination, risk rating, underwriting, limit management and portfolio monitoring are complex and often inefficient.

Best Practices

Based on all of my travels and discussions with risk professionals at leading commercial banks, here are the best practices around managing commercial loan origination and the monitoring process:

Standardize the way data is collected – Commercial loan data is often captured in a variety of different formats, using a range of different technology tools across geographies. While this may work at a local level, spreading financials and capturing other data in an inconsistent manner poses problems when it comes time to analyze the bank's overall portfolio. Information needs to be captured and documented in accordance with a financial institution's origination policies - across geographies and departments. 

Consolidate your data – Once data has been captured in a consistent way, it needs to be archived in one central repository, which we like to call the "single source of truth."  By having access to data in one centralized location, banks will have one of the key building blocks in place to develop sound credit practices. 

Keep in mind, however, we see firms spend a lot of money on the risk side of the equation by standardizing the way data is captured and stored. But they neglect to take the extra step and capture data about a deal - such as fees - that can help them better understand returns in their portfolio. The incremental effort to do this is relatively small, and the business benefit in terms of understanding risk and return can be enormous.

Optimized probability of default (PD) and loss given default (LGD) measures for single obligors – The end goal is to get PD and LGD measures correct, without any mistakes, for every deal. This, however, is unrealistic. So, rather than have credit analysts in the same underwriting group use different models and approaches to calculating PDs, or leave it up to the underwriter to select and apply a range of PD models to each prospective loan, institutions need to apply frameworks and enforce bank policies to prevent user error. Lots of banks use standardized PD models and software to pre-determine which Internal Rating Model (IRM) gets assigned to evaluate a deal. These attributes considered are often industry, peer group, region etc. 

Also, don't use static measures for probability of default - especially when looking at cyclical industries. Default rates can change quickly. For example, according to data from Moody's Analytics Credit Research Database, the median probability of default in the retail sector went from 1.59 percent to 2.92 percent in one year, from June 2008 to June 2009. 

Furthermore, some banks - especially smaller banks - are still using static lookup tables to get LGD measures. These measures are often old and outdated. Banks need to use accurate LGD measures as part of their reserve capital calculations. An analysis is incomplete when this data point is based on old information. Even if you already developed your own PD or LGD models, it's important to benchmark against an independent model. 

Tie single-obligor risk into portfolio-level risk assessment – Banks need the ability to evaluate individual deals on a stand-alone basis and how the deal impacts the overall portfolio. This analysis needs to be done at origination, not after a deal is closed. Just because a loan doesn't look good at the individual level doesn't mean it isn't a good deal for your organization when the loan is considered in the context of your overall portfolio. We have several clients who compute the impact of a new loan or a deal that gets added to their portfolio in real time - and many more are planning to do this. Understanding the Return on Risk Adjusted Capital (RORAC) for any prospective deal is a huge competitive advantage for them.

Systematic and Continuous Centralized Limits Management – Lenders need a way to consolidate and view all their exposures worldwide - from subsidiaries, business units, banking books to trading books. Exposures should be compared with pre-defined limits for counterparties, economic sectors, countries or product types. When coupled with a workflow solution, the monitoring system should trigger alerts when limits are breached and when a "watch-list" customer engages with your firm. Limits management should also be embedded into the origination and approval process. 

A workflow solution that integrates with internal loan systems and front-office systems – Throughout the commercial lending and loan origination process, lots of hands touch lots of different processes. This process needs to be more automated to reduce errors and save valuable time. Identify who needs to be involved - and who can make what decisions - when. To do this effectively and efficiently, risk managers should identify workflow solutions that tie into their own internal loan systems. The workflow system needs to enforce consistent lending policies and ensure that the most accurate portfolio information is available in real-time.

Scenario planning and stress-testing – Firms need a way to model different scenarios and view the effect on single-obligors and the overall portfolio. Understanding economic or regulatory capital needs is only the beginning. Proper stress-testing can be used as a competitive advantage. By understanding portfolio performance during different periods and market situations, lending decisions can be made to optimize the portfolio's performance. Clients use our tools regularly to run what-if analyses and assess the impact of changes in their portfolio's risk-return given a particular stressed scenario. 

Bringing all of these best practices together is no easy feat, but the business benefits - such as pricing risk and optimizing risk-return ratios - can be profound. In order to implement standardized loan processing, decision policies and demonstrate compliance to regulators, risk managers really need to adopt a common platform to standardize the commercial loan origination process - from initial assessment and underwriting through monitoring, servicing and reporting. Firms are starting to apply these best practices today, but many still have a long way to go.

Jodi Alperstein is a Managing Director at Moody's Analytics overseeing Product Management and Marketing for the software division. Alperstein has over 15 years' experience developing innovative products and market strategies in the financial technology sector. The article was co-authored by Arik Pelkey, Director of Product Marketing at Moody's Analytics.

---

Steps to better risk management

• Standardize data collection
• Consolidate your data
• Use dynamic and consistent PD and LGD measures
• Tie single-obligor risk into portfolio-level risk
• Systematic and Continuous Centralized Limits Management
• Workflow solution that integrates with internal systems
• Incorporate scenario planning and stress-testing