Month/Quarter-End Trading Causing Lots of Volatility

This morning's most notable scheduled event and biggest potential market mover was the release of May's PCE price index, the Fed's favorite inflation metric. Indeed the biggest volume spike and most directional movement of the day followed that data faithfully, helping yields move to the lowest levels of the past 3 days.  But things changed a short while later with a reasonably big sell-off to the highest yields of the week, all without any overt justification in terms of data or new news.  Combine it with the fact that Treasury performance is vastly different across the yield curve and this is a classic symptom of month/quarter-end trading.

To visualize the yield curve movement mentioned above, consider a chart of10yr and 2yr yields with equal y axes. Note the 10yr spiking much quicker than 2s during this morning's sell-off.

Some smart people are considering the possibility that bonds are reacting to the presidential debate and that the improved odds of a Trump victory somehow precipitated this selling.  In our view, that's hard to justify considering the random mid-morning timing despite an absence of similar trading earlier in the day.  It's not as if traders changed their minds about the debate implications during this time.  On the other hand, it lines up quite well with past examples of month/quarter end trading on Fridays.

What does this mean for the future?  Nothing.  It's just an explanation of this morning's otherwise perplexing volatility.