- We defined a stress test that assumes equity and credit markets drop as much as they did in the 2008 global financial crisis.
- Our scenario shows a hypothetical portfolio of global equities and U.S. bonds could suffer a 15% loss. Further, our scenario shows global equities might lose around 22%; Treasurys might gain 2%, while investment-grade and high-yield bonds could sink a further 8% and 19%, respectively.
- Sector performance would vary under this scenario, with spreads in the energy sector impacted most and those in health care impacted least.
While newspaper headlines are focused on volatile stock markets stemming from the COVID-19 pandemic, credit markets are not immune. By drawing a parallel with the 2008 global financial crisis, our hypothetical analysis shows that investment-grade and high-yield bonds could drop a further 8% and 19%, respectively. These top-level numbers hide sector differences, however. For investment-grade bonds, the total spread widening since mid-February could range from 280 basis points (bps) for the health-care sector to 660 bps for bonds issued by companies in the energy sector.
Credit under pressure
COVID-19’s market impact has not yet led to losses in some credit markets as steep as those during the depths of the 2008 global financial crisis. Nevertheless, the exhibit below shows that the trajectory is broadly similar, with one important difference: The dramatic spread widening in 2008 had been preceded by a more gradual, yet significant, spread widening that began in August 2007, as credit conditions began to deteriorate. In contrast, the recent sudden widening of credit spreads followed an economic environment that was generally described as positive and that kept spreads at relatively tight levels in 2019 and early 2020.1
Many investors are now concerned that the crisis could worsen. Spreads on U.S. investment-grade bonds have widened by over 230 bps since the onset of the coronavirus outbreak. At their widest spreads in 2008, investment-grade bond spreads widened by approximately 400 bps from their previous lows established in July 2007.
COVID-19’s impact on credit markets is not yet as large as in the 2008 financial crisis
High-yield bonds are represented by Markit iBoxx indexes. The equity market is represented by the MSCI ACWI Index and U.S. investment-grade corporate bonds by the MSCI USD Investment Grade Corporate Bond Index. Based on March 20, 2020, market data. The vertical line marks the Lehman Brothers collapse (Sept. 15, 2008) for the financial crisis and the pre-COVID-19-crisis market peak (Feb. 19). Source: IHS Markit, MSCI.
What if this crisis becomes as bad as the global financial crisis?
We defined a stress test assuming that equity and credit markets drop as much as they did in 2008 and early 2009.2 Under such a scenario, U.S. equities could lose another 25% while the 10-year U.S. Treasury yield could decrease by another 30 bps. U.S. investment-grade and high-yield credit spreads could widen to levels similar to those in 2008, which would imply spread shocks of 50% and 70%, respectively.3
Propagating these shocks to a hypothetical portfolio of global equities and U.S. bonds could result in a further potential 15% loss.4 While global equities would lose around 22%, partly offset by a potential 2% gain in Treasurys, the investment-grade and high-yield bonds could sink a further 8% and 19%, respectively. We emphasize that this scenario is not a forecast — markets could drop more or less — but a “what-if” analysis: What would it mean for portfolios if the losses reached 2008 levels?
Impact of the stress test: Returns across asset classes
U.S. Treasurys and high-yield bonds are represented by Markit iBoxx indexes. The equity market is represented by the MSCI ACWI Index and U.S. investment-grade corporate bonds by the MSCI USD Investment Grade Corporate Bond Index. Based on March 20, 2020, market data. Source: IHS Markit, MSCI.
The above hypothetical asset-class-level losses hide how various credit sectors could be impacted differently by this crisis. The exhibit below combines the realized impact on specific sectors so far with a hypothetical further spread widening based on the results of our stress test. According to this analysis, the spreads of companies in information technology and health care could be impacted least, while energy and financial companies could see their spreads widen the most since the mid-February onset of the COVID-19 crisis.
Impact of the stress test: Spread changes within credit sectors
The U.S. investment-grade corporate bonds are represented by the MSCI USD Investment Grade Corporate Bond Index, broken down into GICS® sectors. Based on March 20, 2020, market data. Source: MSCI.
Thus, based on our scenario, credit investors may not be out of the woods yet if the coronavirus pandemic worsens. Our stress-test results are based on broad asset classes performing similarly to how they did in the depths of the 2008 financial crisis, although various sectors could react differently than in 2008. While a hypothetical portfolio of global equities and U.S. bonds could record additional losses under this scenario, our results highlight the disparate impact on specific credit sectors.
1Wigglesworth, R. “US stocks’ record bull run brought to abrupt end by coronavirus.” Financial Times, March 12, 2020.
2Credit spreads hit their highs in December 2008, and the equity market reached its low in March 2009.
3Our stress test starts with making assumptions about broad market aggregates. We then apply MSCI’s correlated stress-testing framework to propagate the main assumptions to all other risk factors impacting returns. The results shown in the second and third exhibits are generated based on this methodology, using MSCI's BarraOne®. Our assumptions of broad market aggregates have been informed by our analysis of the 2008 financial crisis, but there is no historical period that exactly corresponds to our assumptions. MSCI clients can access this scenario on MSCI's client-support site. This scenario can then be used in MSCI's BarraOne® and RiskMetrics® RiskManager®.
4Our stress-test results are based on factor correlations as estimated by the MAC.S model to propagate the asset-class-level spread shock to various sectors. When users have views on specific sectors, the correlations can be overridden by having an explicit shock on these sectors. For the real estate and information technology sectors, we have imposed the hypothetical future spread changes used in the analysis to more closely correspond to recent spread behavior in those sectors. Those sectors’ spreads were shocked by 50% and 45%, respectively.