How to Use ERM, ML and Scenario Analysis to Advance Your Risk Career

Financial risk managers with strong communication skills have a leg up in today’s job market – particularly if they are knowledgeable about risk integration, machine learning and the power of scenarios.

Friday, July 1, 2022

By Tod Ginnis

Financial institutions have been attempting to move away from risk silos for years, but true, integrated enterprise risk management (ERM) remains a work in progress.

Since some on the business side still have a skeptical view of risk managers, perceiving them as purveyors of doom who prevent the firm from achieving its full potential, FRMs need to do a better job of effectively communicating the benefits of ERM and its potential return on investment (ROI).

In the past, some firms may have done a poor job of selling this ROI because they were looking at ERM through a short-term lens. Scenario-based stress testing partly addresses this shortcoming, but it’s limited in scope, since it considers only certain scenarios.

Alla Gil – CEO and co-founder of the financial modeling and risk management software provider Straterix – says that the only way to uncover “hidden risk concentration” across the enterprise is to deploy fully-integrated ERM. Through this approach, she elaborates, firms can gain a better long-term perspective of their risks while also altering the view of senior management on the benefits of ERM – but only if risk managers make proper use of machine-learning (ML) technology and scenario analysis, and clearly communicate all the steps they are taking.

What’s the role of ML, and why are scenarios so critical to the explanation of risks and opportunities to senior management? Moreover, how can FRMs use ML and scenario analysis to advance their careers?

Machine Learning to the Rescue

Risk managers rely on ML algorithms to comb through massive amounts of data to find all relevant risks and to project how they might impact the enterprise. While you don’t need the expertise of a technologist, Gil recommends studying statistics and probability theory to anyone interested in a ML-heavy risk management job.

She also stresses financial risk managers should learn enough about ML to handle the initial legwork for any projects they want to pitch. “Being hands-on and testing your ideas before approaching your technology team or senior staff is always helpful,” she counsels.

Once you isolate some exposures and opportunities that could improve the company’s risk profile and financial prospects, what’s the next step? Gil believes effective communication is the key to success as a risk manager. You’re not going to make any allies if you simply tell, say, the head of trading that your algorithms “show it would be best if your group changed its strategy.”

The Power of Scenarios

Gil urges early-career practitioners to embrace the power of scenarios. Instead of relying on opaque data from a model, if you want to influence someone, walk them through a plausible scenario, step-by-step, using non-controversial statements. At the end of this process, if your communication is clear and logical, you should stand a good chance at persuading your audience to agree with you on the potential for an outcome that they might have otherwise dismissed out of hand.

alla-gilAlla Gil, CEO and co-founder, Straterix

For example, in 2006-2007, Gil asked bank executives if they believed real estate was a bubble market where prices could fall 10%. Most agreed.

Next, she asked for acknowledgement that if that happens, the existing loan-to-value ratio of 95% would result in negative equity for many, and that negative equity would lead to increased defaults. The final piece of the scenario that Gil described was that foreclosed properties tend to drop further in price, which banking executives readily conceded.

When you walk someone through a scenario using reasonable assumptions, you can, for example, paint a vivid picture of a snowball effect. This can come in particularly handy when you need to convince senior management about the potential of an event to create a severe exposure that needs to be managed.

In contrast, if you lead with a scary prediction from an algorithm, your presentation will be less effective and less likely to gain the backing of executives. Though you may have discovered the consequences of the snowball effect using complex models, your role as a risk manager is to walk others through what you found in plain language, so they can buy into your algorithm’s results.

Make your case to stakeholders in an intuitive way. “We must be able to demonstrate our point of view. The best way to do that is through transparent and explainable models,” Gil says. “Don’t just find the flaws in stakeholders’ thinking. You can also point them to opportunities.”


Tod Ginnis is a content specialist at GARP. He is the author of a GARP blog that is aimed at early-career risk managers and professionals aspiring to earn their Financial Risk Manager (FRM) certification.

BylawsCode of ConductPrivacy NoticeTerms of Use © 2024 Global Association of Risk Professionals