It's All About Duration

I’ve been seriously concerned about the state of the credit markets for quite some time, as chronicled in my posts.  A dramatic reduction in interest rates in all classes of credit (Treasuries, leveraged loans, high yield, mezzanine, bank loans…) has been coupled with a significant deterioration in credit quality (manifesting as higher leverage and looser terms).  This is being driven by a number of factors including a wave of capital that has entered into the asset class.  Investors recognize corporate credit performed very well during the crisis and, given the Fed’s policies, are forced to move out in the risk curve to find suitable yield.  Much of this is well comprehended by students of the markets, but others may not understand that recent events don’t necessarily have historical precedent and could result in significant future disruption.

Let’s talk about duration.  It is the weighted average number of years it takes for a bondholder to get back the purchase price of that bond, including both scheduled interest and principal payments.  Duration is an indicator of a debt’s sensitivity to interest rates.  The longer it takes to get your money back, the more volatile price changes of the bond will be to interest rates.  The lower your current coupon, the slower you get your money back, the higher your duration, and the higher your sensitivity to changes in interest rates.  With interest rates and coupons at all-time historical lows, duration has become quite elevated, resulting in the market’s correlation to interest rates being higher than ever.  It’s helps to conceptualize this visibly.

Many market participants brush off the duration issue by claiming historically when interest rates rose risk spreads compressed, so a significant move in bonds prices shouldn’t be expected.  Maybe… but is this conventional wisdom really true and applicable? The risk spread measures the incremental yield generated for owning debt that is riskier than a Treasury bond with the same maturity.  Without even looking at the data, my first rebuttal is that those prices have moved up significantly in the last few years as a result of interest rate declines.  Intuitively the reverse should be true when they rise.  More concretely, interest rates and spreads have had positive correlation in the past.  For example, from 1950 to 1981 it was +.54.  We can also look at the last two cycles of Fed increases in 1994 and 2004.  In the latter, spreads tightened as the Fed hiked, but in the former spreads rose for almost two years after the Fed’s tightening began.* Finally, the correlation has been positive since April of this year. The chart below shows High Yield bond spreads and nominal yields being almost equal for the first time in history… sure to cause some surprises when rates eventually rise.

These are some of the fun things we think about at Catalus.  In addition to our core structured debt investments, we’ve been spending a lot more time on preferred equity and equity co-investments.  We’re always happy to consider interesting deals so send ‘em our way!

** Morgan Stanley

*Morgan Stanley