Once upon a time – well – a few short years ago, the market was burned by something called correlation risk (among other things). Things that were deemed not to be correlated were.
Tranches of mortgage loans were impaired simultaneously. Players assumed defaults could not climb above a certain level. They suddenly found that defaults could indeed climb that high and higher – and quickly.
It became a big problem for those who were levered and over-exposed to the asset class.
The rest is history.
In the aftermath we implemented legislation and monetary policies that were intended to expedite an economic recovery and prevent another crisis. One result of these policies has been tight credit spreads.
The market currently rationalizes the credit spread environment by assuming that bond market participants are cautious and realistic – that tight credit spreads represent a reality of 1) low default risk and 2) high recoveries.
Which brings me to the current commodity and currency market moves.
This is what is scaring me: If commodities/currencies stay here, bond market participants are wrong. Very wrong.
Think about what happens if these commodity/currency moves are not done AND prolonged/permanent.
Basically, any levered player with commodity exposure has a credit impairment and/or a default.
The ramifications of this are dramatic. There is a crap-load of paper in this sector - and it all becomes impaired. The current tight spread environment (both high yield and investment grade) is rationalized based on the assumption that default rates will 1) be low and 2) have high recoveries.
A prolonged commodity rut will turn that assumption on it heels, and I think the market is just starting to digest that reality. It’s not hard to envision a scenario where cross-default risk balloons - coal, mining, E&P, Frackers, MLPs, etc, etc — it’s basically Crisis 2.0.
What killed us last cycle was cross-default risk. We all assumed things weren’t correlated and they were.
This is clearly not a tomorrow trade, but something worth thinking about.