Over the past two weeks, we’ve begun to document (see here and here) what may end up being a dramatic unwind in the CLO market.
Supply all but vanished in the wake of the crisis but staged a comeback starting in 2012 and by 2014, issuance was running at a $125 billion per year clip. As we noted last week, that’s roughly the equivalent of how much auto loan-backed paper came to market last year.
The problem is that stress on US O&G occasioned by the Saudi’s quest to bankrupt America’s oil patch is contributing to what certainly looks like a HY meltdown and that, in turn, has very real consequences for CLO collateral pools. In fact, 1.4% of assets held by US CLOs have either been downgraded or placed on credit watch negative this year, according to S&P.
New issue spreads are rising, issuance is collapsing, and both S&P and Moody’s recently downgraded several CLO 2.0 tranches for the first time ever.
Unless you think the acute stress in the HY market is set to abate - and trust us, it’s not - then you can bet that things are about to get very ugly, very quickly at least for junior CLO tranches.
For those curious to know how this debacle plays out, history offers a useful guide. As Morgan Stanley notes, first come the downgrades, then come the over collateralization triggers, and it’s all downhill from there.
“From late 2007, CLO equity investors suffered from a sharp decline in loan prices followed by a rapid increase of asset downgrades. The number of US CLOs failing junior OC triggers climbed and led to an increasing number of deals missing payments to equity tranches,” the bank wrote, in a note out Friday. “The proportion of deals cutting off payments to equity tranches peaked in 2Q 2009, right after the bottoming of loan prices and the peak of loan downgrades in 1Q 2009."
This dynamic (i.e. declining loan prices, downgrades, and missed payments in the equity tranches) is likely to repeat itself this time around. Have a look at the following, which shows that just as the preponderance of CCC+ loans peaked, things quickly went to pieces and didn't recover for years.
Here's Morgan on where we stand and where we're headed if history is any guide:
A decline in loan prices leading to a pickup in asset downgrades has already happened in the current US credit cycle, just in a more stretched fashion. US leveraged loans started a long selloff journey from 3Q 2014, when average BB/B loans bids were at 99.41 and average CCC loan index bids were at 97.23. By 4Q 2015, average bids of BB/B loans were at 95.30 and the average bids of CCC loans plunged to 80.28. About one year after loans started to sell off, in 4Q 2015, the downgrade/upgrade ratio of US loans started to increase to significantly above 1 and is currently standing at 2.29.
It is likely that US CLOs' OC cushion will erode and/or breach the junior OC test as downgrades in loans accumulate. In our recent report, A CLOser Look at Asset Downgrades, January 29, 2016 , we introduced a framework to estimate the impact of loan issuers’ one-year rating migration on CLOs. Assuming that excess CCC assets in a CLO portfolio are carried at 60% of par value, and that defaulted assets are carried at 50% of par value, our framework projects that on average the junior OC cushion of US CLO 2.0 deals with a reinvestment period ending in 2017 or later will decline by 1.61% over the next 12 months, and 6.8% of these deals are projected to fail the junior-most OC test in 12 months.
As Morgan goes on to note, CLO tranche spreads generally lag movements in the underlying bonds:
"In the current US credit cycle, we have already observed considerable weakness in CLOs after loans started to sell off in later 2014," Morgan observes. "The spreads of US CLO 2.0s BB and single-B (which are rare in CLO 1.0s) widened significantly in October 2014 as loan prices first started to fall, with the BB/B loan index bid starting to drop from par in March 2014. Following suit, US CLO new issue spreads widened. From September to October 2014, they moved from an average of 650bp to 700bp for BBs and from 775bp to 830bp for single-Bs. While in the first few months of 2015 there was a short rally in loans and CLOs, CLO mezzanine tranches capitulated quickly to the second, deeper dip in loan prices in 2H2015."
So how bad are things going to get you ask? Well, probably really bad - to put it colloquially. "Both loans and CLOs have had severe peak-to-trough price losses and strong rebounds. Putting these into context, the peak-to-trough price decline of the BB/B US loan index was 38 points from par, and the peak-to- trough price declines of US CLO single-A's, BBB's, BB's were 71.4 points (by 88%), 71 points (by 92%), and 64.9 points (by 96%) respectively," Morgan warns, summing things up.
But for those who are inclined to take a glass-half-full approach (and that's obviously us), the bright side is that CLO spreads will recover faster than the underlying loans: "...the glass-half- full view is that in percentage terms, the price recovery in CLO junior mezzanine tranches out of the credit cycle was asymmetrically higher than that of loan prices."
So there you go BTFD-ers. If you can manage to pick a bottom in bad loan performance you can get an asymmetric risk-reward opportunity in CLO junior mezz. What could go wrong?
In any event, we'll close with an amusing compare and contrast exercise.
Morgan Stanley, February 9: "We reiterate our view that the levels of distress in the US market may create 'option-like' payoffs in CLO equity in the secondary market, especially in deals by managers who are better 'credit pickers'".
Morgan Stanley, February 26: "If the current US credit cycle lasts longer than the 2008-09 cycle, we think more vintages of US CLOs are likely to miss payments to equity tranches, and it may take longer for managers to cure the failed coverage tests."