On Wednesday, Nomura’s Charlie McElligott marveled at the sheer scope of the recent duration purge which is part and parcel of November’s pro-cyclical rotation that found Treasury yields surging more than 20bps last week amid rampant (and possibly misplaced) trade optimism.
Hedge funds, he wrote, have been pressing the rates selloff hard. Last week’s COT/TFF data showed specs net sold $32mm/01 the week prior to the breakout from the high-end of the UST yield range.
That, Charlie observed, was “the single largest absolute duration change in either direction since the UXY contract was introduced in 2016”.
Given that, the logical next question is what history says about these kinds of epic duration purges.
Cue a backtest.
“We didn’t just look at an ‘extreme sale’ in Duration in isolation (identified as 1w DV01 sale as 0th %Ile since 1995), but made it contingent upon the prior 12m period seeing such an extreme accumulation of Duration (tested at > 78th %ile since 1995)”, McElligott writes.
In other words, Nomura looked at scenarios where big one-week duration unwinds were preceded by massive duration accumulation over the previous year.
As it turns out, the trade out of those scenarios typically entails bonds stabilizing or outperforming for a spell, along with bond proxies (e.g., defensives), while equities more generally (and especially cyclicals) underperform.
That’s generally in keeping with what we’ve seen, and McElligott expects the analog to hold.
“The market is trading almost exactly as it has in the past into similar ‘Duration Puke after large Duration Build’, which too then aligns with my view that Bonds will locally rally thereafter (NOW) while expecting Equities to ‘pull back’ in the next 2w period”, he writes. That lines up with the notion that “extreme greeks” (as it were) and stretched positioning in equities has the potential to collide with any macro disappointment (e.g., a snag in the trade talks) to catalyze a pullback in stocks off the recent euphoria.
But, nobody should take that as an overtly bearish call. The same backtest shows stocks rising 1- and 2-months on, although the returns for things like energy and crude do not suggest that the pro-cyclical rotation is likely to gather a ton of momentum into the new year (again, going strictly off the backtest mentioned above).
“[After] this next two-week ‘Equities pullback window’, I expect to see a resumption of [the] risk rally which ‘fits’ SPX historical year-end seasonality, albeit with the ‘Duration Sensitive’ tone”, Charlie concludes.