2015-11-21

Treasury / Swap spreads are negative. And what?

In recent weeks there were several news articles related to the negative treasury/Swap spread. This is in particular the case of a Bloomberg article and a Zero Hedge blog.

Some of the “information” in those notes are

Swap rates are what companies, investors and traders pay to exchange fixed interest payments for floating ones. That rate falling below Treasury yields […]  is illogical in the eyes of most market observers, because it theoretically signals that traders view the credit of banks as superior to that of the U.S. government.

It’s hard to overstate how illogical it is when swap spreads are inverted. That’s because it suggests that governments are less creditworthy than the very financial institutions they bailed out during the credit crisis just seven years ago.

Let’s be clear, to me, those claims are purely wrong, nothing is “illogical” in negative swap spread. But why do I rant about that in the “multi-curve framework” blog? Because it is exactly the same that the spread between swaps of same tenor illustrated in Figure 1.1 of the book. Remember, those spreads going up and down while they refer to the same swap tenor. If those swap rates were representing the banks creditworthiness over the term of the swap, there would be no spreads between swaps referencing different indices. The multi-curve framework foundation is a contradiction with the idea that swaps represent the creditworthiness of banks over the tenor of the swap. Those press articles indicate that the basis of the multi-curve framework are not understood by “most market observers”. That leaves a lot of potential sales for the book ;-)

What does a swap represent? As indicated in the first extract above - yes part of those quotes are correct - the swap rate is a fixed rate paid in exchange of floating ones. The floating rates are overnight or term Ibor rate indices. How are those indices computed? As the average of interbank rate over their term - actual rates for overnight and invented rates for Ibor. They represent some fixed amount that the market thinks is equivalent to periodic reinvestment with banks with substitution. The banks that loose too much in creditworthiness or default during one period are not included in the next period. The rate is somehow the creditworthiness of good banks with elimination of weakening ones on a regular basis. It is the fear of bank default without the actual defaults.

On one side there is the risk of default of the state over 30 years and on the other side the risk of default of a basket of banks over 3 months - the standard in USD; see Appendix B.8 in the book - rolled over 30 years with the weak members of the basket eliminated every 3 months.

The question of the swap spread is: can the difference be negative? You may have a quantitative economic model that tells you that the difference should always be positive, but there is nothing “illogical” per se is negative spreads. How do you compare 30 years risk with 3 months repeated risk of something else that is changing every 3 months? There is no obvious comparison. And if your model tells you that the spread cannot “theoretically” be negative, it just mean that your model is wrong, not that there is a problem with the economy.

And if you ask me my opinion about the U.S. government creditworthiness - I know you have not really asked and you probably don’t care about my credit risk opinion, but I give it anyway -, with a very high and increasing deficit, periodic fights about debt ceiling from a government that already morally defaulted on its debt when it abolished the USD/gold convertibility, 30 years is a very long time!