US HY turned to suffer from the fall in commodity prices
The dramatic fall suffered by commodity and energy prices in July and August led to an almost automatic renewed underperformance of US HY bonds vs. their European counterparts. In this respect, this summer very much resembles the end of 2014: in our January 2015 Cross Asset issue, we presented and investigated the decoupling of the two asset classes that occurred over the last weeks of 2014 (see: “US and European High Yield: what a difference!”). We showed how bottom-up factors make the EUR-denominated universe of speculative grade bonds more resilient than the US HY universe to violent shifts in energy prices and sudden spread re-pricing. Two elements, in particular, play quite a role: the exposure to energy/commodity sectors and the weight of low quality issuers. In both respects, USD-denominated speculative grade bonds look more vulnerable, as energy plus metals & mining exposure is substantial and much higher than in the Eurozone (a combined 17% of the universe vs. a much lower 5% for EUR-denominated bonds). At the same time, 14% of US HY debt is low-rated (CCC or lower) and 40% is mid-rated (Bs) respectively vs. corresponding 5% and 28% weights in Europe. Furthermore, the European HY’s exposure to the energy sector is almost entirely represented by two large and highly-rated (BB1) integrated energy groups which tend to be much less impacted by price swings than many small and low-rated US issuers. Last but not least, European HY outperformance was also supported by the resolution of the Greek issue, which eased systemic risk fears and led to a recovery of euro-denominated risky assets.
Performance over the last year: US vs. EUR HY
The reported graphs and tables show the numbers of the decoupling on the two sides of the Atlantic: in a nutshell, over the 12 months ending in July 2015, US HY suffered a cumulative –3% excess return over government bonds vs. a +2% gain by EUR-denominated HY bonds. At the same time, if energy and commodity related sectors are excluded, US HY moved into slightly positive territory, as the entire cumulative negative return was produced by the oil and metals & mining sectors. In particular, what we noted is that commodity sectors mostly contributed for example to December 2014’s and July 2015’s poor performances as these months saw dramatic falls in energy prices. But, at the same time, these two sectors failed to lead a comparable recovery in months like February or April, which delivered positive returns. What is common to the US and European experience is the differentiation made by credit quality: CCCs underperformed significantly in the US but also in Europe. BBs outperformed, especially in the US, and were in line with Bs in Europe in leading performance. As the interest-rate normalisation cycle is likely to start soon in the US, quality, which also means better liquidity, is going to remain quite a relevant factor on the other side of the Atlantic, but is likely to also remain a driver of performance in Europe.
US HY default rate still low, increasingly driven by the energy and metals & mining sectors
In the first half of 2015, the US HY default rate remained relatively stable at low historical levels: according to Moody’s data, the US HY 12-month rolling default rate started the year at 2.1% in January before moving slightly higher to 2.3% in June. However, though the default rate did not diverge significantly from H1 or H2 2014, it became increasingly concentrated in just two sectors. In fact, according to data published by the major rating agencies, the energy and the metals & mining sectors respectively accounted for 10 and three issuers, out of the 22 companies that defaulted in the first seven months of the year: this means around 70% of the total. If we also take into consideration distressed debt exchanges (transactions in which credit holders receive less than par), credit events look even more concentrated: out of a total of 10 distressed exchanges, seven companies were from the energy sector and one from the metals and mining sector. Furthermore, the number of credit events tended to accelerate over recent months, increasing from just three and four respectively in January and February, to six in both March and April, before jumping to nine in both May and June. According to Fitch, as we are writing, US energy speculative grade issuers show a 2.3% default rate, while metals & mining even reached 10% on the back of the most recent credit event.
Distress ratios and default rates: the first lead the second
Different parameters are used to define debt in “distress”, but usually it is a matter of spread levels above government bond yields or of the bond prices’ levels. In the first case the threshold is usually fixed at an OAS at or over 1,000 b.p. above Treasury bond yields, in the second case a bond is considered distressed if it trades below 75 cents of par. Our preference between the two alternative approaches goes to the first, as it doesn’t depend on coupon and yield levels, but just refers to the risk premium paid by the issuer over risk-free assets. As such, if spreads at or above 1,000 b.p. are considered, we come up with the reported graph which shows how default rates tend to lag the proportion of issues trading at distressed levels by around four quarters. Trading above certain spread levels means a very high probability of default implied by investors and consequently almost unsuitable conditions for meeting refinancing needs through the bond market: in a nutshell, the market is practically closed for these issuers. The distress ratio finds its exact parallel within the other major financing channel, namely the traditional bank loan market, in the percentage of banks tightening lending standards to companies. Our default rate regression model is actually mainly based on these two inputs, both leading defaults by around one year. As such, within our model the regression coefficient for the distress ratio is close to 10%, simply meaning that a 10% distress ratio usually results in a 1% default ratio in one year’s time. The graph shows that the US HY distress ratio started to rise by the last quarter of 2014, increasing from 5% to the low double digit area: by the end of July it had risen to almost 14%. This means, all other things being equal, that a rise in expected default rates in one year’s time by around 1%, was promptly captured by the regression. Most of the rise in the distress ratio was driven by energy and, more recently, also by the metals & mining sector. The energy-sector distress ratio was 12% in July 2014; one year later it stands at 30%. The smaller metals and mining sector is in an even worse position with a 50% distress ratio, currently. Therefore, not only is 70% of the current default rate represented by just two sectors, but also at least half of the overall distress ratio of US HY belongs to these two sectors.
Let’s take a deeper look into the US HY energy sector: are all energy companies equally vulnerable to oil prices?
Over the last five years, mainly thanks to extraordinary primary market activity, the energy sector has become the largest sector in the US speculative grade universe: the number of issues jumped from 242 as of August 2010 to the current 367, while debt face value rose from USD 94 bn to its current level of USD 212 bn over the same period. As a consequence, both the number and debt face value of issues grew from 10%-11% to the current 15%-16% of the asset class. Remarkable investments linked to shale technology, together with other factors, were behind this fast growing funding activity. However, though quite big in terms of overall dimensions, the energy sector is composed of sub-sectors with very different sensitivity to trends in oil and energy prices. Therefore, the combination of vulnerability to the eventual fall of energy prices and the average rating or credit quality of each sub-sector has to be considered to fully estimate the commodity market’s impact on the default rate outlook. The purpose of the following table is to summarise this combination of high/low quality and high/low sensitivity to changes in energy prices. Exploration & production, together with Oil Field Equipment & Services look much more vulnerable to a fall in energy prices than the other two sub-sectors, namely Gas distribution and Oil Refining & Marketing. At the same time, we divided each sub-sector into three credit quality categories, based on their letter rating (BBs, Bs and CCCs or lower).
Results point to a mitigation of concerns about default rate risks driven by a fall in oil prices. The major default risk, in fact, lies in the low-rated/high-sensitivity bucket, which accounts for around 11% of the overall sector. Low-rated names with a low vulnerability to oil swings represent just 1.6%, while less vulnerable/mid-rated names represent a remarkable proportion, with 27%. On a reassuring note, 25% of energy debt looks to be of high quality, with low sensitivity to oil price swings, and another 8% is mid-rated with low exposure to commodity volatility. Total weights, then, show that two thirds of debt is more at risk, but at the same time 53% of debt is still highly rated.
The second table reports spread levels for each bucket, which combines ratings and simplified sensitivity to energy prices, as of end of July: the table seems to confirm the picture and the quite different risk-reward profiles for each of the identified subgroups. Sensitivity to energy prices looks to be the main discriminating factor for spread levels, while quality also plays quite a role within the two groups of sub-sectors.
What default rates are implied in current spreads of energy companies?
Implied default rates greatly differ among the above buckets: we start from a very low 1.2% for the best-quality/low-sensitivity bucket (top left), then move into the 3.5%/4.0% range for the two light green buckets, then move up again to 11% for mid-rated/highly-sensitive names to finally top out in the 38% range for the most risky bucket on the bottom right. These are one-year implied default rates based on recovery assumptions of 30% and after subtracting an average 260 b.p. excess return historically offered by HY bonds. The entire sector’s average one-year implied default rate is at around 10%.
Therefore, now let’s apply a few “stressed” scenarios to default rates in one year from now
As we previously outlined, current default rates for the energy and metals & mining sectors stand at 2.3% and 10%, respectively. The energy sector default rate is still quite low and in line with the overall HY market average. This is due to the fact that default volumes started to rise only in very recent months and from below average levels. As the default rate is a 12-month rolling ratio, it takes time to fully incorporate the trend. Let’s assume now that distress ratios are good predictors of future defaults: currently, around 30% of the energy sector is in distress, meaning a 3% default rate in one year’s time. In a particularly stressed scenario we may also assume that, within the banking loans funding channel, lending standards and conditions applied to energy companies have recently become tighter, though still within banks’ generally easy stance vs. the majority of US non-financial companies. We therefore double the effect of distress in the bond market, assuming a comparable impact is produced by tighter bank lending conditions. We therefore assume a potential rise to a 6% default rate in one year’s time. Assuming a further deterioration of this scenario with a rapid rise of the distress ratio to a 50% level, this would mean a default rate of between 5% and 10% in the next 12-15 months. As the energy sector represents 14% of the overall speculative grade market, this could mean around 1.4% in the worst outcome. The metals and mining sector already stands at a 10% default rate, but it is a relatively small sector. Doubling its default rate would add another 0.6% to the overall average rate. Therefore we would get around 2% from these two sectors to be added to other sectors’ volumes, which are currently in the 1.5% range. This would lead to a result of 3.5%, very close to our regression outcome, which is actually slightly higher at 3.8%. Contrary to metals & mining, the energy sector has a relevant weight of companies with a relatively low sensitivity to energy prices. At the same time, as the recent years have shown, defaults are still very much a CCC or lower rated story.
The strong link between oil price levels and two thirds of the US HY energy sector will inexorably drive default rates higher in the next one to two years. In recent months, we have already seen an increased volume of credit events concentrated in the energy and commodity sectors and more are likely to come over the next quarters. Current spreads imply quite a remarkable probability of defaults over the next year, pointing to a 10% default rate vs. the current 2.3%. Defaults are likely to come more from low rated-highly exposed issuers, which already imply quite a very high level of distress. Therefore, the contribution of energy and metals & mining names will be quite relevant and it could even account for more than half of the overall default levels. Our regression based on leverage, bank lending standards and the distress ratio is pointing to a 3.8% rate in one year’s time, consistent with an absolute contribution of 2% from the two above sectors plus a 1.5% rate for all other sectors. A rise to the 4% range would mean almost doubling the current rate of 2.3%, but the default rate would still remain substantially below the historical average and, as has been the case over the last five years, it should remain concentrated in the lowest rated names. Furthermore, the fall in oil and commodity prices is supportive for most of the non-energy sectors and the maturity schedule of the next two to three years looks manageable. At the same time, the Fed’s rate normalisation cycle represents a potential risk to the asset class and recent Chinese forex policy developments and their link with commodities also pose further upside risk to the scenario. In light of these risks and trends, we reiterate our preference for high-quality and low-duration segments of the US HY market.
Commodity-linked sectors currently make for 60% of US HY defaults
Not all energy companies are equally vulnerable to oil price swings
Most of the rise in the distress ratio was driven by energy and, more recently, also by the metals & mining sector
25% of energy debt looks to be of high quality, with low sensitivity to oil price swings
Current default rates for the energy and metals & mining sectors stand at 2.3% and 10%, respectively
The strong link between oil price levels and two thirds of the US HY energy sector will inexorably drive default rates higher in the next one to two years
The Fed’s rate normalisation cycle represents a potential risk to the HY asset class and recent Chinese forex policy developments and their link with commodities also pose further upside risk to the scenario