+1 Added to my documents.
Please be aware your selection is temporary depending on your cookies policy.
Remove this selection here

High Yield default rate cycle and perspectives

The essential

Apparently in a quite paradoxical divergence between macro and financial trends, the current cycle of HY default rates in the aftermath of the Great Financial Crisis is the most benign of the last three decades. Therefore, to what extent “this time looks different” and may remain so?

This focus seeks to address these and other questions from both a top down and a bottom up perspective. The combination of macro and fi nancial conditions is therefore considered, together with refunding risk and other risks more recently arising, like higher debt issuance from low rated names and rising lite covenants deals. Overall, we share the comforting view of major rating agencies on a persisting low default rate scenario in Europe and US.

What a peculiar default cycle…

As the first graph shows, with respect to both previous two experiences of early ‘90s and early 2000’s current default cycle proved to be remarkable rapid in the rise to peaks as well as in the subsequent fall to lows. Even surpassing previous absolute peaks in the worst year of the cycle, in fact, default rates were above the long term average just in the 9 months before reaching the peak and in 10 months following it, or in other words for less than 2 years overall. The graph 1 shows how different the previous cycles were: early ‘90s and early 2000s saw defaults remaining above long term average for respectively 37 and 57 months, namely for periods almost twice and three times longer. All this has to do with the much lower cumulated defaults suffered by speculative grade companies in the aftermath of the Great Financial Crisis, apparently in a quite paradoxical divergence between macro and financial trends. Moreover, the new peak reached by HY defaults in 2009 may also mislead in the comparison with previous cycles. Actually, the higher peak was not the result of worse default performance by each rating category but most of all the effect of the higher weight reached by lowest rated CCC debt in overall speculative grade market just before the crisis.

… Therefore, are the famous (and dangerous) four words “this time is different” right?

The graph 2 shows annual default rates of US B rated names over the last three decades. Both 10-yr averages and worst yearly defaults followed downward trends. Interestingly, ten year default averages fell from 8.5% in ‘84-‘93 to 4.2% in ‘94-‘03 and finally to just 2.0% in the last ten year ending 2013. Worst yearly defaults also fell from 15% (1990) to 9.5% (2001) and finally reached just 7.4% in 2009. What is most striking about last decade performance, however is that default rates surpassed the 5.2% long term average only once, while they remained between 0% and 1% in eight years. A very similar pattern is shown also by defaults of Ba rated issuers, and to a lesser extent by defaults of CCC names, too. We all know that the very good performance of the four years following the peak of the crisis, namely the period between 2010 and 2013, has certainly to do with an unusual phase of abundant liquidity and search for yield, ultimately supporting the demand for speculative grade bonds and therefore keeping default rates from rising. However, the question is about the future trends: is this time different, to what extent this divergence with the past may last?

At the end of the day, defaults are very much a macro growth story…

The third graph shows the link between macro real growth and speculative grade default rates: peak years and default trends very much depended on changes in GDP trend. Long term downward trend in US GDP growth is captured by the linear regression reported in the graph.

In order to better analyze to what extent growth may explain companies’ defaults we therefore calculated deltas between GDP changes and their LT trend on one side and differences between annual default rates and their long term average. The chart 4 shows the resulting link. In a nutshell, not the level but the change of GDP growth with respect to its long term trend seems to be very much the real drivers. The graph shows that most of the years when growth is below its long term trend default rates tend to be above their long term average, therefore higher than 5.6%. At the same time, as in the present phase, when GDP is moving above its LT trend, companies’ defaults tend to be in a low regime and to stay below their LT average levels.

Interestingly, despite GDP growth having slowed down significantly over the last three decades, the default cycles proved to be less and less painful for US speculative grade companies, confirming that it’s not the level of macro growth that affects defaults.

… And a story of fi nancial conditions…

Together with macro growth, financial conditions play a significant role in leading default rates. The reason has to do with the high dependency of low rated companies on external funding and their vulnerability to changing sudden changes in liquidity conditions and investor’s risk aversion. As the bond market has become a major funding channel for US speculative grade companies vis a vis bank loans, also the role of financial conditions in the bond market has become powerful, as a driver of defaults. Our forecasting model (see: Cross asset investment strategy special focus published January 2012: “Credit: Corporate default rate forecasting”) is based on both financial conditions’ trends for bank loans and bonds. Bank Lending Standards (BLS) survey conducted and published by the Fed captures the availability of bank loans, while the Distress Ratio (Dr) or the percentage of HY bonds trading at or above 1,000 b.p. spreads captures the openness and availability of bond investors to fund speculative grade companies. Both the percentage of banks tightening lending standards to companies (BLS) and the DR lead default rates by around four quarters on average: intuitively it makes sense, as the cut of credit lines or the sudden closure of the bond market takes some time to produce a “drought” of liquidity and to subsequently contribute to an extreme credit event. The two reported graphs (5 & 6) show the changing role played by bank and bond market financial conditions in the last two default cycles: the tightening in bank conditions played a major role in early 2000s default cycle, while the last cycle saw a very strong relationship with trends in funding available in the bond market. Interestingly, the graph 5 shows that US banks started to move to a less easy (though not yet a tight) stance already during the credit bubble years (2005-2007), at a time when bond market was still fully “giving credit” to speculative grade companies. Also in the very last quarters, it seems that default rates are following more closely the DR rather than BLS conditions, confirming a stronger role played by capital markets than by banks in determining the default cycle. Both factors, however, are still pointing to a low or even lower default regime for at least the next four quarters.

… Which U.S. companies wisely capitalized on…

The major positive factor for future trends in companies’ default arises from the fact that refunding risk over the next few years has dramatically fallen. As the graph 7 shows, as of December 2013 less than 10% of overall outstanding debt was due to mature between 2014 and 2016, a much lower proportion then just 2 or 4 years ago, which saw respectively 15% and 20% of overall debt maturing in the following three years. Thanks to huge refunding and prefunding undertaken since 2010, the picture has strongly improved:  hat the graph shows, in fact is that by the end of 2009, 60% of overall debt was due to mature in the following fi ve years (namely between 2010 and 2014), quite a worrisome scenario. As of December 2013, the same proportion of debt maturing in the following fi ve years (namely between 2014 and 2018) is almost half, namely 35%. This occurred notwithstanding the strong rise in HY bond debt market size.

… And European companies, too

A recent report published by Moody’s shows a reassuring trend for European companies, too (see: “High Yield Interest-European Edition”, Moody’s 16-July-2014). Among the three major areas of improvement cited by Moody’s, the strong reduction in next years’ refinancing needs is dramatic: “The growth of the EMEA high-yield market over the past three years has been driven primarily by the refinancing of bank/CLO-funded LBOs in the capital markets….This structural transformation has slowed down as it nears completion.” The other two areas of improvement cited by Moody’s for European companies have to do with liquidity and rating trends. The liquidity profile is at its best level since Moody’s run its periodical study, improving from already solid levels. The good news is about periphery issuers, which improved significantly from previous studies. The third area of improvement is on rating trends. The percentage of companies with negative outlooks or under review for downgrade has fallen to 17% from a high of 23% at the end of Q1 2013.

Is a leverage bubble building in HY bond market?

Two sources of concern about HY default perspectives are more frequently cited: 1) the recent rise in the proportion of lowest rated debt in overall speculative grade new issuance and 2) the progressive easing of required covenants accepted by investors in primary markets, especially within leveraged loans. Many observers are concerned about an excessive risk taking attitude shown by investors in recent months: in other words, the persisting search for yield may have led investors to fund low rated companies which would not have access to market financing in normal times. Therefore, this could create the conditions for a new bubble forming in credit markets.

Let’s then address the first point: on this side, data referred to developed HY bond market do not seem to subscribe to this concern. Contrary to previous cycles experiences, in fact, the proportion of new B- or lower rated debt remained in the 33%-37% range of the overall new issuance over the last five years, and actually the proportion slightly fell to the low level of the range in 2013 (33%) and in H1-2014 (32%). A rise in the proportion of low rated new debt, actually, was more the case in the leveraged loans market, the one cited by chair Janet Yellen in her July testimony to US parliament. B rated leveraged loans in fact rose from 40% levels in 2010 to a 54% in the last few years. The same is true for the second issue of concerns(covenant-lite issuance) which seems to affect much more the leveraged loan sector than the HY bond market itself.

However, one very important point which is not usually mentioned about these sources of potential concern is the fact that most of the new debt issued by low rated speculative grade companies in recent years has gone to refinance maturing debt or loans and not to increase the leverage. One of the most peculiar features of the current cycle, in fact, is the cautious attitude of companies towards re-leveraging strategies despite the ample and extraordinary amount of liquidity available to them in the bond market. New debt issued for re-leveraging purposes usually starts to rise and to return close to previous peaks in the few years following the balance sheet repair phase: this is not the case currently, as for the 6th year in a row, 2014 still records new issuance for refinancing purposes making around 70% of primary market overall activity. Just to cite a few numbers, in the four years starting 2010 and ending 2013 US HY companies issued USD 330 bn of bonds for bonds refinancing and almost USD 400 bn in bonds for bank loans refinancing. Over the same four years another USD 515 bn bank loans were refinanced with new loans. Overall this means a huge volume of USD 1.24 Trillion refinancing against an overall issuance of new debt for re-leveraging purposes (namely, for acquisition/LBO and equity monetization) of USD 480 bn. Year to date figures are still in the very same proportion: USD 81 bn of new debt issued for bond refinancing, USD 72 bn for loans refinancing and USD 90 bn new loans covering for maturing ones, totaling USD 243 bn against USD 118 new debt and loans for re-leveraging purposes. A proportion of two dollars for refunding to one dollar for re-leveraging. Furthermore, combining issuance by rating and by purposes the picture looks more reassuring than some depict it. 2000 and in 2008, the two years which were followed by the bubble burst and the spike in default rates, respectively saw refinancing as a percentage of lower rated issuance falling to just 17 % and 8%. The same proportion, year to date, though lower from last year 54% is still at a decent and comforting 47%. This means that low rated companies tapped the market for “good reasons” and this should help default rates remain low rather than being seen as a negative factor.

The second factor of concern or the covenants issue looks more tricky than the rating issue. As pointed out as well by S&P in a recent report dedicated to the topic (see: “Covenant-Lite issuance casts a cloud over future default levels”, S&P 14-July-2014) “There has been a proliferation of covenant-lite corporate loan borrowing in the U.S. over the past 18 months, a majority of which Standard & Poor’s rates in the ‘B’ category”. This is a cause of concern because in rating agency own words: “In the event of a significant financial market liquidity crunch, default rates for covenant-lite borrowers could spike well above the levels seen during the 2008-2009 financial crisis”. Though there are only limited data on covenant-lite defaults and even less on their recovery rates, historical experience do not point to a worse performance of this sub asset class within the HY market: according to S&P, actually, in the 2007 -2013 period covenant-lite loans experienced lower default rates than traditional leveraged loans. This does not come as a surprise as this kind of transaction tends to reduce rather than to increase the default risk over the short term, because the company has more flexibility to face financials headwinds or unexpected drop in economic activity. Fewer default triggers, in simple words, are at work in this kind of loans. At the same time, it opens to higher risks for lenders looking forward as the borrower finds it easier to take aggressive financials strategies in terms of leverage or lower required liquidity. One points which seems to be reassuring is the fact that these transactions have their maturity wall in 2018 and 2019, therefore more than three years from now.

Therefore what conclusions to be drawn on future default trends?

a. HY companies have moved into a “low default rate regime”.

b. This occurred despite a downward trend in LT macro growth which was more than counterbalanced by the supportive effects induced from ultra-easy monetary policies.

c. The combination of macro growth and fi nancial conditions still looks supportive over the next few years, though the balance of more growth and less easy monetary policy is likely to take the place of todays’ opposite combination.

d. Our fi ndings show that it’s not the absolute level of GDP growth which matters for default rates, but actually it’s the delta vs LT trend. When GDP growth is above its LT trend, even if the latter is on a downward path, default rates remain below their LT average. This should be the case for the next two years in the US, according to our projections.

e. Despite the recent disappointing macro data, also European GDP perspectives point to better growth in 2015 and 2016.

f. On the micro side, so far, companies made a very “wise” use of exceptionally favorable funding conditions; most of them refrained from embarking into dangerous re-leveraging and capitalized on pushing forward debt maturities, fully taking advantage of low funding costs.

g. As a result, very low refunding needs looming in the next two to three years are likely to prolong this “peaceful” cycle of low defaults, consistent with the combination underlined in point c).

 

Download the article in English

Télécharger le document en français

 

Not the level but the delta of GDP growth with respect to its long term trend drive default rates

5-01
5-02
5-03
5-04
5-05
5-06
5-07
5-08

 

Refunding risk over the next few years has dramatically fallen

5-09
Sergio BERTONCINI, Strategy and Economic Research at Amundi Milan
Send by e-mail
High Yield default rate cycle and perspectives
Was this article helpful?YES
Thank you for your participation.
0 user(s) have answered Yes.