Slowing

CESIUSD Index {GO}

The USD is selling off as the market starts to sniff out slowing growth for the rest of this year and into 2027. U.S. economic surprises appear to have peaked, and this may well not be random. The two biggest drivers of U.S. growth are slowing. Fiscal is turning from boost to drag, while AI Capex is still growing, but not as fast. Here’s the fiscal story.

And AI Capex. This next chart is from DoubleLine, but you can find similar data at BofA and other places, too. DoubleLine has 85% and 28% for 2026 and 2027 while BofA has 91% and 35%. Same ballpark.

Depending on who you ask, AI Capex and U.S. fiscal have had varying impacts on U.S. growth over the past four quarters, but regardless of who you ask, the impact is large. Two of the biggest engines of U.S. growth are going to be less powerful and we are already seeing a string of weak data points for Q3 (negative NFP, soft inflation, weak Retail Sales). I showed the relationship between yields and economic surprises in am/FX on Friday and while it’s not perfect, it’s pretty decent.

An important psychological aspect to any slowdown or soft patch in the U.S. is that the market is really not used to seeing economic weakness in America anymore, and so the reaction could be more extreme than you might otherwise expect simply because there is a pretty big surprise factor to it (if it happens). We have become accustomed to year after year of solid growth and a permanently antifragile U.S. economy, so if it somehow ends up showing signs of fragility, that could have a bit of a wow factor.

Atlanta Fed’s GDPNow reading, which opened above 5%, is now at 4.3% and so there is still plenty of room for expectations to fall. There is also plenty of room for Fed pricing to drop as the US 2-year yield still trades around 50bps above the current implied overnight rate and banks keep adding more hikes to their forecasts even though a very realistic outcome is that the Fed actually hikes zero times.

As discussed over the past week or so, I am looking for trades to get ahead of narratives that I believe the market will find compelling when it returns from holidays and is ready to deploy capital after Labor Day. With the recent run of soft U.S. data, a Fed that probably turned max hawkish at the worst possible time, and two big reasons to think U.S. growth could limp into 2027, the first trade on my radar is the U.S. soft patch / slowdown trade.

As I get these ideas, I will aim to initiate between now and Labor Day, depending on the urgency I feel. This does not feel super urgent as bonds cannot hold a bid right now and the market is still super sleepy.

I wanted to note here, too, that tech analyst and friend of the show Jamie Saettele pointed out to me that when TLT makes a new 1-year low on 1.25X or more normal volume, then reverses the week after that, again on 1.25X or more normal volume… That’s been an excellent reversal (buy) indicator. This is a non-random indicator, to me, because the combination of new price extreme on high volume and failure to continue is a classic reversal. Slingshot reversal + high volume reversal = good indicator. The buy signal outperforms the upward drift of the series, especially after three months.

There have been ten buy signals (one still open, triggered August 7) and here are the returns (total return, including carry). The juiciest holding period is three months because as you move further out, the return becomes more dominated by carry and less by directional price movement.

Note the sample here is small, and the October 2023 signals overlap because they are two weeks apart. Still, the market psychology setup is logical so I will consider the 89% win rate meaningful despite the small n.

I am not quite ready to pull the trigger on TLT, and might consider a shorter-duration instrument when I am ready to play for economic weakness because the further out you go on the curve, the less certain you can be that the economy is the main driver of yields. Here is the history, in chart form.

---

Calendar

The highlight of the week is FOMC Minutes; that’s not super encouraging for volatility.

This Jiji article suggests the possibility of a Japanese FinMin shakeup. I will discuss in tomorrow’s am/FX. Positioning report follows.

It is ~100% certain that someone reading this has their birthday today. So to you(s) I say: Happy birthday.

Hi. Welcome to this week’s report. (To read about how I use and trade this report, see here.) The market is back to dabbling short USD as the economic surprises in the U.S. turn south, and gold has found its footing. JPY, GBP, AUD, and NZD are all getting a bit of love, while EUR and CHF are not. Gigantic 1.1500s in EURUSD this week, while the RSI heatmap shows that MSFT (the only outlier last week) fell back to earth after a bad week last week and EWC (the Canadian equity ETF) is the most overbought or oversold number on the board. The backtest shows EWC RSI needs to get above 79 to send a bearish signal.

For a greater than 50% chance that at least one other person in a roomful of n people has the same birthday as you, n would need to be at least 253 (so 254 people total, including you). This number is significantly higher than ⁠365/2⁠ = 182.5: the reason is that it is likely that there are some birthday matches among the other people in the room.

Green line is p(anyone has same birthday) while blue line is p(someone has same birthday as you). x-axis is number of people in the room.

good luck ⇅ be nimble