A few weeks ago, I mentioned that the interest rate swaps market was pricing in lower long-term rates than the U.S. Treasury cash bond market. Some media outlets opined that fixed income market illiquidity might be skewing the relationship between swaps and Treasuries by pushing swap spreads too low. I believe the opposite might be true. The swaps market is even more institutionally driven than the U.S. Treasury market and has less central bank participation, as well as very little retail involvement. Since the financial crisis, the 10-year US Mid-Swaps Spread usually traded between 10 and 20 basis points above the yield of the 10-year UST note (it traded much higher than that prior to the financial crisis). Swaps tend to trade above the corresponding UST yield due to counterparty risks when engaging in a swap agreement. However, at the time of this writing, 10-year U.S. Mid Swaps were trading at more than 16 basis points under the yield of the 10-year UST note. This is the widest negative spread ever.

Alright, 10-year Mid Swaps are trading below the yield of the 10-year UST note. How can I be confident that it is the 10-year UST note which is mispriced? If I am correct, the breakdown in the relationship should have occurred around the time we saw a spate of central bank selling of U.S. Treasuries. Not only that, but the breakdown should have occurred by UST yields moving higher while Mid Swap spreads should have remained little-changed. Let’s look at what happened and when:

The data indicate that the swaps UST relationship was fairly normal prior to late August, when the spread began to narrow. That was about the time we first heard of central bank selling of Treasuries, trying to defend their currencies as EM currencies were battered. The trend continued through September as central bank selling continued (confirmed by TIC Flows data). By October we saw retail Outflows and then some shorting by some institutional speculators around the time of the October FOMC meeting and again after the strong Nonfarm Payrolls report, released on November 6^th. It is my belief that, because they are less impacted by central banks and retail investors, the 10-year Mid-Swaps Spread might be giving us a better indication of where the fixed income market believes long interest rates should be than does the 10-year UST yield.