What we are not doing in High Yield
We have often started our portfolio discussions with clients this year with what we don’t like in fixed income. Unfortunately the list for 2018 is relatively long and quite large when measured on a market cap basis. Our regular readers will know that we mostly refer to long dated government bonds and long dated highly rated credit. But within the sectors that we still like there is also a sub list of do’s and don’ts. In the next couple of paragraphs we will address our list for High Yield markets, which are usually one of the main providers of negative headlines as we get past the peak and approach later stages of the cycle.
The first and probably the most obvious place to start is CCC issues. Defaults in CCCs are materially higher than those in single Bs. According to Moody’s Annual Default Study the average yearly default rate in CCCs from 1990 to 2017 is 13.1%, while for single Bs the number is 3.7%. 1990 was the worst year in the period for both, with defaults reaching 43.9% and 13.7% for CCCs and single Bs respectively. As the cycle ages borrowers with higher leverage, lower free cash flow generation or ambitious/transformational expansion plans/mergers to integrate are most at risk. There is no short one line summary that describes all CCC issuers out there, but it is fair to assume that some of these characteristics will be in the list of the majority of CCC issuance. We also note that defaults are caused by a variety of different reasons. It could be the case that the issuer does not have enough cash to make an interest payment. However, in our most recent experience with default rates spikes in 2009-2010, the majority of defaults were due to distressed exchanges where creditors take a haircut and a new bond. This might be caused by issuers not being able to refinance their debts as they mature due to financial markets being closed. The point here is that in periods of market stress, which typically come at the end of the cycle, a “healthy” CCC issuer might find itself locked out of capital markets and forced to restructure its debt in some way. Allocations to CCC issuers should gradually decrease as the end of the cycle approaches and fall to nil way before the cycle ends.
Secondly we have senior unsecured bonds. Recovery values in unsecured bonds are typically lower than those of secured bonds for a given issuer (and ratings are lower). Probabilities of default might be the same but the loss given a default is higher in the case of unsecured bonds. Consequently even if the issuer does not default its unsecured bonds will be a lot more volatile. As the cycle matures exposures to senior unsecured parts of the structure should decrease in general. We do note, however, that compared to CCC credits there will be a bit more of credit specific issues when assessing whether we like a senior unsecured issue or not. As we mentioned earlier defaults in single Bs are materially lower than those of CCCs, for example, if there is no secured debt in the structure then we would be more comfortable holding senior unsecured debt of that issuer than otherwise.
Last but not least we have extension risk; a good part of bonds in High Yield have call schedules. The most common structure at issuance is a 7 year bond with a non-call period of three years and yearly calls thereafter. Issuers generally call bonds when they are able to refinance them at a cheaper level. At issuance the call is at the money, but if the price of the bond goes up (as has been the case at least until January 2018) then the call will be in the money at some point. When this happens the bond is said to be trading to a call, with today’s information the most likely scenario is that the bond gets called in the future. If you have a relatively newly issued bond whose call is two or three years away, the issuer is in the single B bucket and you forecast that the cycle might end over the next couple of years, even if the cycle does not end with a big spike in defaults, the bond in our example will most likely not be trading to a call as spreads might widen. Essentially we had a short dated bond, and applying our assumption of it being called in a couple of years’ time, that is now a much longer bond. Credit spread duration changes quite dramatically here and the bond goes from being a relatively low volatility/low yield bond, to a longer more volatile bond. If you were relying on this sort of bond to reduce your volatility and duration, you might find that actually this does not happen. Bonds issued not long ago that are trading above par (but not by much) are the ones that are most at risk.
The list is of course not an exhaustive one and there is no substitute for due diligence and detailed credit work. But it’s also as important to have a more macro view even at the sector level, as the asset allocation is a major driver of absolute and relative performance. This is not only true at the portfolio level but also at the sector level.
