Risk Tactics for Turbulent Times
Introduction
“There is a tide in the affairs of men, Which taken at the flood, leads on to fortune. Omitted, all the voyage of their life is bound in shallows and in miseries. On such a full sea are we now afloat. And we must take the current when it serves or lose our ventures.” William Shakespeare, Julius Caesar (1599).
The previous articles on ‘Managing Originations in Turbulent Times’ resulted in a lot of feedback and questions around what can be done to address the current challenges that most lenders around the world are currently facing.
In light of the feedback, I thought it would be useful to create an article describing some of the credit risk tactics that I would recommend risk managers consider and implement in order to better manage their portfolios in this period of uncertainty.
Test, Test, Test!
At the heart of credit risk management is the concept of running A/B tests for originations and account management strategies. Champion/Challenger testing functionality is standard in all modern decision engines, enabling risk teams to ‘test and learn’ using statistically valid test groups of applications or accounts.
If for some reason your current decision engine lacks testing capabilities or you do not have a decision engine available for use, the strong recommendation is that you budget for an upgrade of your existing decisioning software, or for the purchase of a new decision engine.
A risk team trying to manage originations or account management strategies without a decision engine that has A/B testing is akin to ‘taking a knife to a gun fight.’
However, having access to champion/challenger testing functionality does not necessarily assure success. The functionality needs to be used and multiple strategy tests designed and implemented and even more importantly, monitored.
If, like many risk managers, you are only currently running a small number of tests, I would urge you to start planning for the roll out of a lot more! Only by testing can you establish what the impact of the global recession is having on your new and established accounts.
A lot of credit grantors follow the micro economics tenet of Profit Maximisation is where marginal revenue is equal to marginal costs. This means that profits are maximised up to the point where the marginal revenue of an account is less than the associated marginal cost of that account.
In an era of high inflation, interest rate increases, increased bad debts etc. the established point of profit maximisation will need to be re-examined, as all of these factors have a significant impact on this. By running multiple A/B tests, credit risk managers can establish what the new point of profit maximisation is for applications, along with all of the associated new strategy cut-offs.
Examples of tests that should be run in originations include:
- Application score cut-offs
- Affordability calculations
- Loan amounts
- Instalment periods
Application score cut-offs
Lots of tests should be run around the cut-off scores to establish what the new cut-off should be in the new paradigm that we now find ourselves in. For example, if the current application score cut-off is 200, then lenders should examine stepped increases of 205, 210, 215, 220 etc. Clearly there is a trade-off between increased rejections and lower bad debt write-off and hence why testing new cut-offs on a small tranche of the applications is so important.
The same is also true for credit bureau scores and any other scores that a lender may be using.
Affordability Calculations
If a lender is using a custom affordability calculation, this should be examined and tested against, due to the impact of inflation on many of the inputs. Typically, during a recession, lenders will test stricter affordability calculations in order to reduce write-offs from over-indebted applicants.
Loan Amounts
This is another area of focus for testing, as lenders will often rein in the limits that are granted to applicants in order to minimise their overall risk exposure. Testing is vital in this arena, and the caveat is that a very low risk account is still a very low risk account, hence why across the board ‘limit slashing’ is not recommended.
The recommendation is for testing to focus on the high risk, marginal applicants and perhaps also the medium risk, who could be in danger of losing their jobs.
Instalment Periods
Associated with limit amounts are loan periods, where lenders often want to reduce the overall periods offered to applicants in order to reduce risk. Once again, the caveat is that testing should focus on the appropriate applications, not the very low risk customers.
Testing enables lenders to ‘test and learn’ before they make sweeping changes to the champion strategies running on the bulk of the portfolio. By adopting multiple A/B testing, lenders can avoid a knee-jerk reaction to the economic crisis.
Improve Internal Cooperation
It has never ceased to amaze me how many lenders work in silos and how different departments work independently of each other, often pulling in different directions. This dysfunctional set-up has been referred to as the ‘credit tug of war,’ where the credit risk team is often accused of wanting to minimise risk at all costs and the marketing team is accused of wanting to book every application, regardless of risk!
This dysfunctional situation is often seen in legacy banks, where the various departments are often located in different buildings, even in different cities.
The four departments that should ideally be working closely together are:
- Credit Risk
- Marketing
- Fraud
- Collections
Credit Risk
This department is the lynchpin of all functions, as it decides which applicants are approved for credit and then how accounts are subsequently managed. When the economy is booming and credit policies are relaxed there is little critique, apart perhaps from the fraud and collections teams.
However, during an economic downturn, such as is occurring now, tensions arise from all areas. This needs to be managed carefully, in order to maintain the best interests for the organisation.
Marketing Team
The credit risk and marketing departments should ideally work very closely to ensure consistency and avoid any dysfunctional or high risk campaigns. This is even more important when the economic conditions are adverse, as it is pointless to run campaigns that either end up with high rejection campaigns or result in significant adverse selection of sub-prime applicants.
Fraud Team
The fraud department is often the ‘poor cousin’ department, as it is viewed as a cost of business, rather than a revenue generator. In economic downturns, fraud rates typically increase and so the credit risk, as well as the marketing teams should work closely with fraud to try and mitigate against this.
Collections Team
Collections is where the ‘buck stops here’ and is often the victim of poor marketing campaigns and credit risk policies, as well as lax fraud controls. There is often a lot of friction, as the collections department is held accountable for decisions that were made months ago. Working closer with the collections team will identify areas that can be improved across the entire bank, from marketing campaigns to risk management policies and also fraud prevention. (Collections is often where fraudulent accounts are finally identified).
Focus on Scorecards
Associated with the requirement to run multiple test strategies is the need to ensure that the application scorecards that are being used are the most accurate and up to date. You need to ask yourself:
- When were these application scorecards last developed?
- When were these application scorecards last re-aligned?
If the answer to the first question was more than 3-4 years ago, then re-developments need to be considered and budgeted for.
If the answer to the second question was more than 12-18 months ago, then scorecard re-alignments should be scheduled.
If the A/B testing is to use fine scorecard breaks of 5 points, then obviously the scorecards need to be accurate and the score to odds relationship should be as it was intended in the original scorecard development.
Reports are your new best friend!
Key to all of these tactics is the requirement to monitor the results and the only way that this can be achieved is through the frequent use of reports.
The 3 most important reports for strategy testing are:
- Strategy Outcomes Reports – what outcomes did the strategy produce this day/week/month?
- Strategy Performance Reports – how is the strategy performing this day/week/month?
- Strategy Difference Reports – how is the strategy performing compared to the champion strategy, as well as other challenger strategies?
Without a focus on using these reports to the maximum effect, everything else is a waste of time.
Summary
There is no doubt that 2022-2023 will present significant challenges for all of us in the lending industry. These challenges will be addressed more successfully by companies that follow these tactics, than lenders which take ‘knee-jerk’ reactions, or even worse, bury their heads in the sand!
As Shakespeare wrote over 400 years ago, do you want to make the best of the flood, or wallow in the shallows, with all of its associated miseries?
About the Author
Stephen Leonard is the founder of ADEPT Decisions (www.adeptdecisions.com) and has held a wide range of roles in the banking and risk industry since 1985.
