By the end of March, the median EDF™ (Expected Default Frequency) value of North American corporate ﬁrms climbed to 2.31%, more than double its level last year. This increase marks the ﬁrst time the median EDF value has doubled year-over-year since September 2008.
The elevated credit risk over the past year is mostly attributable to substantially increased market leverage, stemming from a severe drop in equity markets during February and March.
We also observe a similar magnitude of increasing credit risk in both bond spread and credit default swap (CDS) spread. More speciﬁcally, the spread between high-yield and investment grade widened signiﬁcantly to historical levels.
When we compare the recent equity volatility and stress current asset volatility to the same levels seen during the last ﬁnancial crisis in 2008−2009, resulting EDF levels climb to all-time highs across all ratings.
This special issue of the EDF report provides an extensive discussion of COVID-19’s effects and the implications for leveraged lending during an unprecedented stressed environment for North American corporates.
The North American corporates experienced a very steep credit quality decline in March 2020, following the global outbreak of COVID-19. By the end of March, the median EDF value of North American corporate ﬁrms reached 2.31%, more than double the prior year’s level. This increase marks the ﬁrst time the median EDF value has doubled year-over-year since September 2008. The elevated credit risk over the past year is mostly attributable to signiﬁcantly increased market leverage, stemming from a severe drop in equity markets during February and March 2020. The median EDF credit measure increased by 124% year to date.
We also observe a similar magnitude of increased risk in the credit markets, with both bond spread and CDS spread surging substantially to ﬁve-year highs. In particular, we observe substantial spread increase in the high-yield space, while investment grade levels remained relatively moderate.
Observed corporate defaults of non-ﬁnancial public ﬁrms still remained low by historical standards, which may not be surprising. The onset of the outbreak was sudden and relatively recent. It will take time for impacts to trickle down and become default events. However, we have started to see a few COVID-19–related defaults in the United States.
A number of risky industry/sectors’ EDF levels increased signiﬁcantly during the last 12 months. Speciﬁcally, oil- and consumer-related industries made up the top of the list when ranking industries by liabilities-weighted EDF measures. These sectors were also among the industries that experienced the largest increase in risk during the prior year.
This report ﬁrst shows the overall credit trend reﬂected through realized default counts, the EDF credit measure and its drivers, and credit spreads for North American non-ﬁnancial ﬁrms. Subsequently, we analyze the riskiest industry sectors using a variety of metrics. We also report credit migration patterns using EDF metrics. As a special issue of the EDF report, we also conduct extensive discussion regarding the impact of the COVID-19 pandemic on credit risk, as well as its implications for leveraged lending for North American corporates.
During the last ﬁnancial crisis, we published periodic EDF reports that many clients found helpful. This report marks the ﬁrst EDF report focused on the current economic impact on EDF credit risk measures, with a series of related studies to come.
2. Credit Measure Trends
This section shows EDF credit measure movement and key drivers for North American non-ﬁnancial companies between January 2000 and March 2020. We report these trends in conjunction with realized default counts and credit spreads.
2.1 Realized Defaults
Figure 1 shows quarterly default counts for North American corporates, classiﬁed by whether the ﬁrm is rated at the time of default. Figure 2 presents quarterly number of defaults, highlighting bankruptcy defaults. Default occurrences reafﬁrmed that the most recent benign credit environment continued at least to the end of 2019. So far, we have seen only a handful of defaults in the sample during Q1 2020. However, given that the onset of the pandemic has been sudden, and it is still relatively recent, we expect to see a signiﬁcant increase in defaults during the second half of 2020 if the credit environment continues to deteriorate. Figures 1 and 2 detail default counts by rating status and default type, respectively. Figure 3 reports monthly default counts in the previous year, during which period the default counts remained low. The 12-month default count rolling average currently stands at 3.92, down from 4.17 one year ago. As a benchmark, during the past 10 years, the average monthly default count is 11.01.
2.2 EDF Credit Measures and Drivers
Figure 4 reports quartiles of the EDF credit measure and its drivers, including the underlying asset volatility and market leverage, from January 2000 through March 2020. Recently, we have seen a substantial rise in EDF values as a result of substantial increases in market leverage. During the past 12 months, the median EDF value rose dramatically, from 0.90% to 2.31%, more than a factor of two, to its ﬁve-year high. Median leverage increased from 21.86% to 30.67%, due to a signiﬁcant equity market sell-off. Median asset volatility increased from 38.10% to 38.61%. As asset volatility typically lags equity volatility, we expect asset volatilities to continue to rise, which would result in further EDF value increases.
2.3 EDF Credit Measures and Drivers: Rated Firms
Figure 5 presents EDF measures and drivers for ﬁrms holding a Senior Estimated Rating assignment by Moody’s Investors Service. To ensure each rating group has a sufﬁciently large number of ﬁrms, we analyze only groups rated A through B. In the cross-section, the median EDF values for these rating groups are consistent with the risk order indicated by agency ratings. For EDF value drivers, B-rated ﬁrms are riskiest in terms of both business risk (that is, asset volatility) and ﬁnancial risk (that is, leverage). The EDF value difference between Baa-rated and Ba-rated ﬁrms is primarily driven by asset volatilities, while leverage is the main driver for the EDF value difference between Baa-rated and A-rated ﬁrms.
Currently, Ba and B-rated ﬁrm EDF credit measures are approaching the historical highs seen during 2008−2009. These levels show a stark contrast with the much more benign credit environment observed during the past few years. For example, a Ba-rated ﬁrm is now almost as risky as a B-rated ﬁrm in early 2019. However, despite the recent uptick, the EDF credit measures for investment grade ﬁrms are still relatively moderate.
We observed a moderate increase in median asset volatilities across all rating groups, up from historic lows at the second half of 2019. The bottom panel of Figure 5 shows that the market leverage — the ratio of the default point over the market value of assets — presented a sharp increase from their historic lows during 2019. The magnitude of the recent increase in leverage is almost as steep as the one observed during October and November 2008. Currently, B-rated corporate leverage is approximately 55.10%, quickly approaching the peak reached during the 2008−2009 crisis. This recent sharp increase in market leverage reﬂects the dramatic drop in market capitalization for North American corporates across all rating groups. As the market cap reﬂects expectations regarding future corporate earnings, this sharp drop points to the quickly deteriorating debt service capacity of North American corporates.
2.4 Credit Spreads
Figure 6 presents the time series of median values for the EDF credit measure, the ﬁve-year CDS spread, and the option-adjusted spread (OAS) of investment grade and high-yield North American public ﬁrms from January 2014 to March 2020. Median bond option-adjusted spreads are derived from the sample of bonds in coverage by the Moody’s Analytics EDF-based bond valuation model. Generally, all three measures have been moving in tandem during the long history. The recent dramatic change is readily apparent for all three measures, especially for the high-yield space. Table 1 shows the median credit metrics by broad rating group as of March 31, 2020 and one year ago. As oil and gas-related ﬁrms account for a large portion of high-yield bonds, we see the high-yield space was essentially doubly hit by both the COVID-19 pandemic and the recent oil price shock, causing signiﬁcant distress to speculative-grade debt.
Similar to the pattern observed for EDF credit measures, both OAS and ﬁve-year CDS spreads present a substantially steeper credit risk increase in the high-yield space than in the investment grade space, suggested by the increasingly widening gap between the two broad rating classes, shown as the dashed blue line in Figure 6.
3. Industry Analysis
This section applies two measures that capture different aspects of industry risk. The ﬁrst measure is based on the average EDF value weighted by total liabilities of each company in a given industry. Therefore, this measure is dominated by the risks of ﬁrms with large amounts of liabilities. The second measure is the percentage of ﬁrms with EDF values greater than the 90th percentile of the entire population of North American non-ﬁnancial companies. Consequently, this measure tends to be more represented by the risks of smaller companies, which are likely to be riskier in most industries. With both measures, the analysis in this section helps paint a relatively complete picture of industry-level credit risk for North American corporates. To avoid small sample bias, we examine only industries with more than 20 ﬁrms as of March 31, 2020.
3.1 Riskiest Industries
The entire North American corporate population has a 2.49% liabilities-weighted EDF credit measure, up from 1.27% last year. Figure 8 shows the riskiest industries and the distribution of liabilities-weighted EDF credit measures across all industries. The chart on the left side displays the top-ten riskiest industries and their EDF measures; the chart on the right displays the EDF measure distribution of all industries.
The top-two riskiest industries are Oil & Gas Reﬁning and Mining-related, probably not surprising. These industries were hit twice in rapid succession — decreased demand due to the COVID-19 pandemic and a massive equity price drop due to the fall in oil prices. In particular, the liabilities-weighted EDF value for the Oil, Gas & Coal Exploration/Production industry is 12%, presenting alarming levels of default risk. Other industries topping the list tend to be non-essentials, including Consumer Services as well as Consumer Goods Retailers and Wholesalers, Entertainment & Leisure, and Construction-related. All of these industries have above a 3% liabilities-weighted EDF measure. Medical Services, considered a “mission-critical” industry in light of the COVID-19 pandemic, is still relatively risky and ranked fourth. However, as shown in Table 3, the Medical Services industry is doing relatively well compared to other industries, given the moderate increase in its median EDF measure.
3.2 Industries with the Largest Credit Risk Increase
The sharp elevation in credit risk is more apparent when we look at year-over-year increases in EDF values within each industry. From March 2019 to March 2020, most industries in the North American corporate space experienced substantial increases in credit risk: the median EDF value for all industries has almost doubled from a year ago. The industries with the largest percentage increases are reported on the chart to the left in Figure 10, where we ﬁnd industries heavily engaged in travel and leisure (for example, Hotels & Restaurants, Entertainment & Leisure), as well as international trade (Apparel & Shoes, Electronic Equipment, Furniture & Appliances).
Among these industries, the median EDF value for Apparel & Shoes increased almost by ten-fold in light of the COVID-19 pandemic. The chart on the right side of Figure 10 presents the distribution of changes across all industries, which is skewed, with a median percentage increase of 147%.
4. Credit Migration
This section analyzes credit quality shift from both the change in agency rating and the change in EDF-implied rating. The agency rating is more latent, reﬂecting the credit risk ranking of a ﬁrm over a long period of time, and is thus, more stable. On the other hand, the EDF-Implied Rating (EIR) is calibrated monthly, and is, therefore, more point-in-time, but also more volatile.
Table 2 shows the mapping between EDF values and the corresponding implied ratings for North American non-ﬁnancials as of March 2020. EDF values within the lower bound and upper bound are mapped to the corresponding rating class. The resulting EIRs are then compared to their counterparts in March 2019. Figure 11 shows the differences, along with the notch differences in agency ratings.
Looking at the histogram of notch differences, even though the agency ratings suggest that the longer-term credit risk for the majority of the ﬁrms have changed little, we see a positive skewness for the EIR, suggesting clear, immediate credit deterioration, in line with other credit metrics.
5. Risk Highlight
March 2020 EDF values were highly affected by COVID-19. Emerging as a surprise wildcard risk, the pandemic will likely continue to have a profound risk impact as the situation evolves. Because of the recent outbreak, ﬁnancial institutions are likely to re-assess expected losses, while portfolio managers will likely adjust their credit exposures.
Among a variety of asset classes impacted signiﬁcantly by COVID-19, leveraged lending is certainly one of the most concerning. The dramatic decline in equity markets reﬂects pessimistic expectations for future corporate earnings. This issue could lead to deteriorating debt servicing capability, which would be especially problematic for highly leveraged, speculative-grade companies.
We next highlight speciﬁc risks associated with the immediate impact of COVID-19 across industries, as well as the implications for leveraged lending.
On an aggregate level, March 2020 has seen one of the steepest one-year EDF measure increases during the past 20 years, shown in Figure 12. This metric has almost doubled, compared to where it has been during the past three years. However, this measure has yet to reach the level last seen during the 2008−2009 ﬁnancial crisis. Given a ﬁxed loss given default (LGD) and exposure, this observation suggests that a well-diversiﬁed portfolio currently has an expected loss more than double its level at the beginning of the year.
However, as shown in previous sections, the current asset volatility level remains relatively tame, especially for investment grade. We estimate asset volatility using a three- to ﬁve-year rolling window, designed to be less volatile. Given that equity volatility, often proxied by the volatility index (VIX), reached a historical high in March 2020, it would be reasonable to think that asset volatility would follow the same path, given the uncertainties ahead. Figure 13 presents the stressed EDF value calculated with the current level of leverage, but assuming the asset volatility in the 2008−2009 high range (shown as the triangular points). With such heightened volatility, we would see EDF values across all rating grades, rather than just with high-yield, climbing back to their respective all-time highs.
The ﬁrst immediate impact of COVID-19 is the re-shufﬂing of industry risks. With lockdowns enforced across many regions and countries, the universal stay-at-home order has become the most devastating economic force in modern history. A variety of Main Street industries have been signiﬁcantly affected. However, some industries seem to be coping better than others.
Table 3 compares median EDF values on March 31, 2020 to respective values on December 31, 2019. The table ranks industries from top to bottom, by the year-to-date percentage change. From the year-to-date perspective, industries related to travel, transportation, entertainment, and leisure were among those hardest hit by the pandemic, perhaps not surprising, as consumers signiﬁcantly cut travel and entertainment needs. In addition, ﬁrms in Apparel and Shoes, as well as Consumer Goods, found themselves caught between slashed demand and disrupted supply chains. Meanwhile, oil reﬁners fought hard to maneuver through the pandemic, in addition to feeling the effects of a profound oil price shock. On the other end of the spectrum, industries deemed “essential,” such as Utilities and Agriculture, as well as “mission-critical” industries related to Medicals Services and Pharmaceuticals, performed relatively well.
Such “relativeness” can be better seen by analyzing the EIRs. While most industries experienced an EIR downgrade, some in the bottom-10 impacted list are associated with EIR upgrades, despite increased in EDF values, suggesting relatively better performance and less risk than other industries, year-to-date.
Although most main street industries have been negatively impacted by COVID-19, some companies in certain industries appear to have managed well. For example, Table 5 lists the top-10 companies with the most EDF value decrease in the Pharmaceutical industry. Some companies that were among the riskiest in the industry have experienced more than an 80% drop in EDF value and substantial upgrades in EIR while looking at soaring market capitalizations. Although certain companies on this list achieved substantial market cap increases due to mergers and acquisitions and positive news on other fronts, others managed to turn the tide, potentially because of being perceived as mission-critical during the pandemic. For example, this includes a number of companies associated with coronavirus vaccine development, testing, and treatment (Vir Biotechnology, IBIO, Co-Diagnostic, and so on).
5.2 Leveraged Lending
Highly-leveraged speculative companies are mostly likely to fall victim to the COVID-19 pandemic’s ramiﬁcations, given their vulnerability to falling equity and earnings. In particular, the oil and gas companies that account for a large portion of the high-yield space are navigating a double-hit, from both the pandemic and the oil price shock, which makes these companies especially vulnerable.
As shown in previous sections, we have already seen a signiﬁcant divergence in credit spreads between investment grade and high-yield. Figure 14 shows the historical breakdown of total liabilities by rating for North American non-ﬁnancials. While high-yield has been steady at around 25% of the entire sample, the sheer amount has been fast-increasing, almost tripling during the past 20 years, to about $3 trillion USD today.
Taking a closer look at the high-yield space, Figure 16 presents the top-10 industries with the highest expected losses as of March 31, 2020, as well as breakdowns by rating. The distribution of expected losses among high-yield is extremely skewed, where the top-10 industries amount to 72% of the total. While the current expected loss for investment grade is relatively stable compared to its level a year ago, expected loss for high-yield has increased substantially, and its distribution has become much more extreme, with the risk increasingly concentrated in a handful of industries. For leveraged lending, the Oil and Gas Exploration/Production industry shows an alarming increase in expected loss, rising from about $5 billion USD one year ago to $27 billion USD currently.
With the Federal Reserve’s current market operation continuously purchasing investment grade instruments on the line, this gap between investment grade and high-yield may further expand in the near future. Previously, during the dot.com bubble and the 2008−2009 ﬁnancial crisis, when the total expected loss for all ratings reached similar levels, we saw increases in both asset correlations and systematic risks. The current situation calls for precaution and possible government intervention, as we continue to navigate the turbulent COVID-19 pandemic.