Some folks are bullish on Treasurys.
Meanwhile, some folks are bearish on the short-end.
Specifically, the juxtaposition between positioning in TU and TY has raised a few eyebrows of late, especially considering the massive short in the former was built ahead of a Fed meeting that no one expected anything from.
The lackluster incoming inflation data only added to the confusion.
“The mindset of the market is that inflation continues to disappoint the Fed — most people think the Fed is on hold for the rest of the year, to be honest,” Charles Comiskey, head of Treasuries trading in New York at Bank of Nova Scotia, told Bloomberg earlier this week, adding that “being short the two-year note here just doesn’t make sense.”
As it turns out, it does “make sense” from an RV perspective, something we discussed in “Solving The Mystery Of The Big 2Y Short: ‘It’s Too Damn Rich!’”
Well, fast forward to Friday and we now know that spec positioning in Treasury futs hadn’t really changed that much through Tuesday:
Still “bigly” short TU and “bigly” long TY.
More than a few folks have suggested that the positioning in the 10Y could well be a contrarian indicator, especially given the CTA dynamic (more on that latter point here).
With all of that in mind, and with the knowledge that the man who triggered a mini-tantrum in rates just one month ago is going to be addressing Jackson Hole, consider the following out from Citi on Friday evening…
Vol: Sellers Should Walk To The Nearest Exit
The steepening of swaption skew and the still long spec positioning in TY futures continues to be the main risk to our bullish duration call. We advise scaling down short gamma positions going into August due to event risks, seasonality, and the risk of a selloff.
Swaption skew now indicates a bearish risk to rates…
One of the chief risks to our bullish call on Treasuries heading into July was the recent steepening of swaption skew. Since then, the bid for payers has continued which, in general, is a bearish indicator for duration (Figure 15). The 3m change in 3m10y swaption skew has some predictive power over the next one month move in 10y rates (Figure 2). We believe the predictive power of skew is related to positioning, given core longs may hedge with OTM payers which steepens the skew.
While we continue to be bullish on Treasuries, we must highlight the fact that the current change in 3m10y skew is now decidedly in the highest 25th percentile of the historical distribution. Historically, this implies a sell-off in rates over the next one month by ~5bp on median, with a sell-off occurring 60% of the time, and is a strong bearish indicator for the next one month (Figure 16). Hence, skew is implying core longs may be concerned here and are hedging in the options market.
…as does spec positioning in TY futures
In fact, those core long positions may be overweight in TY futures as seen by the still high spec positioning when normalized by open interest (Figure 17). Spec positioning in TY futures, normalized by open interest, also adds some additional predictive power to our swaption skew metric. Currently both indicators are in the top 25th percentile of their historic range going back to 2007. This implies a strong risk of a bearish move in 10y yields over the next one month with a median sell-off of ~9bp when both signals are in this percentile bucket (Figure 18).
Short gamma positions should watch for increased risks
Given increased risks of a market selloff, we advise scaling back short gamma positions going into August. Implied vols are likely to move higher on the back of higher rates. In addition, seasonality also argues for upside risks to vol. In fact, August has been the worst performing month for systematic vol sellers since 2010 if evaluated by returns on a daily delta-hedged systematic short 1m10y program (Figure 20). Over a longer sample, starting in 1997 and excluding the financial crisis, August is the second worst performing month for systematic sellers. Shortdated implieds also tend to richen in August which, on average, is the strongest month for implieds (Figure 19). Over the past five years, 1m10y vol has seen a median richening of 3.5 normals during August. This is also true over a longer 10y sample, excluding 2008 and 2009, with a median richening of 2.2 normals for 1m10y. We believe this seasonal pattern may reflect poor liquidity conditions in August coupled with event risks, such as Jackson Hole. In fact, we may see something similar this August with Draghi speaking at Jackson Hole.