Manoeuvring credit portfolios beyond Covid-19
Over the last months, we have been reminded that during times of heightened volatility or generic instability, markets have a tendency to run synchronously. As a consequence, assets become more correlated, even across different sectors. It was indeed striking to observe that in this respect the Covid-19 crisis even outpaced the global financial crisis from 2008.
The video below shows rolling windows of 1-year pairwise correlations of S&P sector indices, comparing them for the two crises: the crisis of 2008 and the current one.
Correlations can serve as a good measure to gauge the overall development of the larger market. We observe that during both crises multiple sectors start moving essentially in lockstep with correlations above 90%. However, the impact of the current crisis affected even more sectors in this respect and more sudden than the global financial crisis from 2008. In this crisis more than 10 pairs of sectors started moving synchronously!
How will the world look in the aftermath of the current crisis? Will markets continue to run synchronously for a longer while like they did after 2008–2009?
After the financial crisis of 2008, increased defaults were observed over the following years, most prominently between 2009 and 2010. While first defaults as a consequence of the Covid-19 lockdowns are unfolding already now, many people fear default scenarios well beyond those in 2009–2010, reaching out for several years. In some of its most recent statements, the Fed announced that its own loss projections suggested that in a bad scenario the eventual hit to loans could be far worse than in the aftermath of the crisis of 2008.
The just published Fed’s CCAR stress testing results for the US banking sector are still to be updated this year, asking banks to re-run parts of the stress tests with updated scenarios.
How can financial institutions react to this situation? How to update models in a market environment changing at an unprecedented speed?
Especially from a Risk Management perspective, it will be a challenge to prevent severe downside scenarios in the portfolios and corresponding instabilities in financial and capital planning.
When the Fed announced its CCAR scenarios for the 2020 stress testing exercise in February, they were largely outdated a month later and not just by small figures. Also input parameters to credit impairment and portfolio level models, can hardly be updated in a meaningful way under the standard approaches to react to the changing market environment and allow for stable planning.
How to differentiate between industries in scenarios that usually build on high-level macro-economic variables? How are they impacted by the crisis and governmental protective measures?
Forward-looking measures as used in IFRS 9/CECL or regulatory stress testing like CCAR are typically not adapted to the current situation. Their underlying scenarios rarely embed pandemic events and there has not been any precedent case for gauging them. The macro-economic variables usually have modest power to differentiate between industries. Here not only Risk Management gets into the spotlight. IFRS 9/CECL impairment calculations will substantially be impacted by the crisis. This will lead to earnings volatility, induced by higher default probabilities and also correlated credit migrations in the portfolios.
How do you approach manoeuvring your credit models and their calibration through and beyond Covid-19 ?
SigmaQ Analytics is a specialist Lab for Quantitative Analysis/Risk Management and Data Analytics. As one of our main focuses, we develop cutting edge models and use advanced data analytics to closely look at future scenarios and how they impact IFRS 9/CECL impairment, regulatory stress testing and credit portfolio models. Our team has worked on CCAR stress testing, IFRS 9 impairment and credit portfolio models for a wide range of clients, from G-SIBs to leading Reinsurance and Insurance companies.
If your institution faces challenges with one or more of the above topics, we will be glad to share our expertise with you. Get in touch with us to receive more information: email@example.com.