A hawkish read of Warsh puts two Fed hikes in the price before year-end

The projections are the reconciliation event. A Jackson Hole reading is a price put on a tone, and prices put on tones get marked down; the numbers the committee has to publish alongside its September decision ^[1] are a different object, because someone has to sign them, and if that written path stops short of what August read into the speech ^[2], the hawkish position gets unwound at the very meeting that was supposed to ratify it. That is one way this goes. Another is that Japan's data does the work instead, since a central bank whose business investment keeps shrinking ^[4] cannot match a tightening Fed on schedule, and the gap then widens by arithmetic rather than by intent. A third is a soft print somewhere in the middle, which is the dull outcome and therefore the likeliest.

What is more interesting, or rather the version of this a book can actually express, is how little of any of it shows up in aggregate. Compound Chandler's two annual index moves and the recovery comes to about 4.7% ^[1], which is what you have after four consecutive down years running from 2021 through 2024 ^[2]. Underneath that near-flat number, the Australian dollar has appreciated almost 7.3% against the US dollar through late August on the back of three first-half hikes ^[7], roughly 2.4 points of currency per hike delivered ^[4], while sterling's six-month high of $1.3675 came out of an August gain of about 0.4%, its first back-to-back monthly rise of the year ^[6], which implies a starting point near $1.3620 and a total advance of about half a cent ^[3]. Call that a level rather than a trend.

This is probably wrong, but I would rather own the dispersion than the direction: the currencies whose central banks have already delivered are the ones with the least left to price, and the currency whose central bank cannot deliver is where a widening differential quietly accumulates without anyone announcing it. The counter-thesis belongs in the same breath, which is that a speech-driven repricing is the least durable kind there is, and that Chandler is reporting what the market took away from Warsh rather than what the committee has committed to ^[2]. I would drop the view on a Japanese consumption print that turns positive, or on a set of Fed projections carrying one move rather than two. Meanwhile, a market spending September on a single speech is a market not spending it on the Xi-Trump meeting or the German state elections sitting on the same page of that calendar ^[1], and those are the items that can move an inflation forecast without any central banker opening his mouth. Chandler has been reading calendars like this for forty years ^[8]. The line in his summary I would price off is the capex one.