A lot of thoughtful commentary and analysis has been put out there this past week on swap spreads going negative this past week. Good examples can be found here and here. Well I’m going to throw in my two cents on the subject as well.
In trying to analyze the issue, I pulled data from the Federal Reserve to look at spreads at various points along the yield curve and compared the average swap rate each day to its equivalent constant maturity Treasury. The following two graphs show this in two ways. First, I looked at swap spreads for the 2, 3, 5, 10, and 30yr swaps to Treasuries through time. While all the talk this past week was about 10yr swap rates going negative to Treasuries, they have been negative for at least 6 months at the 30yr:
Since I also wanted to view swap spreads as a yield curve in their own right, I wanted to see what the term structure of swap spreads looked like. This was even more telling in some ways. Not only is it negative sloping, note the degradation and shift in this term structure in the past month:
The only things I can think of that would have had such a profound effect on this curve was the Fed’s “extended period” phrase being left in after the last FOMC meeting and we had a bad Treasury auction.
But as I look at these charts and data, I can’t help but think both of the dominant theories about this phenomenon are at work. In the near term in the 2s,3s, and 5s we have swap spreads that seem to be moving together both in direction and in the size of the spreads. The fact we’re seeing all of them move lower in unison speaks of extreme market complacency in the near term – confidence that adjustable rates will remain super low and fixed keeps grinding lower in the process. Again, the image of kids carry traders picking up pennies in front of oncoming buses comes to mind.
But at the 10 & 30 yr, things seem to be different. There, I think David Merkel of the Aleph Blog may be on to something:
Which is what I also tend to believe. I think this is more about a Treasury glut than it is a credit risk issue. Because governments can tax and they can print more currency. Debt issuance is not the same as issuing currency because debt is borrowed. Currency isn’t. Defaults in a sovereign context really have to do with the tendency of a government to do a strategic default on its debt. Think of strategic defaults on mortgages except bigger. Other events can serve as a proxy for default like IMF/World Bank assistance, or currency devaluation.
So given those are the default definitions we have to work with, can we really say the US is going to do one of those things? It’s hard to imagine them doing so, but there is no such thing as a zero probability event. Over a long enough timeframe, anything can occur.
But as great as it is to imagine all of this, the big question is what will the rate curve look like going forward? I don’t know, but Dr. Donald van Deventer from Kamakura Corp. illustrated their 10yr forward forecast of the Treasury yield curve:
The thing I note? The compression of the curve. At the left (i.e. present day) you can see a nice upward sloping curve. But as you move right, you see the curve is being forecasted to get tighter and tighter, with most of the increase coming in the short maturities/front-end. Call this another “tell” that the market is being too complacent. If they were more focused on the idea that the curve’s shape can change, they wouldn’t be looking to gain extra basis points off the present shape, but would consider what else could happen.
Oh well. Just don’t come crying to me if it all goes wrong…