A lot depends on the CPI and core CPI print today

Outlook: A lot depends on the CPI and core CPI print today. The Bloomberg survey gave us CPI probably at 8.7% y/y from 9.1% in June, led down by gasoline. But core is seen picking up to 6.1% from 5.9%, if under 6.5% from March.

A Reader sent us a tweet from a fund manager who notes the Cleveland Fed estimates have been 0.23% below the actual since Oct ’21, on average, and almost 40 bp below in May and June. “If the same holds true, CPI could be 9.0% to 9.2%” and not the estimated 8.7%. Remember, the Cleveland Fed has a series of estimates involving the trimmed mean (excluding extreme outliers), one of which is a lagged version. The current Cleveland Fed Nowcast is 8.82% and 6.1% for the core. 

More than one analyst is looking at the Cleveland Fed’s Nowcast but hardly anyone is mentioning that it goes into August, too—at tiny dip to 8.76%. This reminds us, or should, that when it comes to making the famous data-based decision, a single month of inflation data doesn’t cut the mustard. Even if the Cleveland Fed is right about both July and August, and trimming the mean means it will miss by some amount by definition, it’s not enough data for the Fed. We will need a rolling 3-month average (or some other statistical variation) that is demonstrably, inarguably, falling.

Besides, inflation doesn’t just slide away overnight. It takes months. Goldman is forecasting core CPI sticking to 0.4-0.5% m/m for the next few months and managing to get down only to 0.3-0.4% by December 2022. Goldman expects core at 6.1% in December, moderating to 2.7% by Dec the following year. To emphasize the improbability of a rapid correction in inflation, BoA shows that to get under 5% would take a steeper drop than is realistic.

Yikes! Nobody in his right mind should be trading anything until the numbers are out ands the dust settled. We all know the response will be bigger, the bigger the divergence between expected and actual, but if somebody says he knows in which direction, he’s lying. It’s possible a lower number brings out the still prevalent peak inflation talk and expectations of lesser Fed action after the September hike.

It’s also possible a much higher number restores the Fed to its sort-of plan and even ramps it up a bit. You’d think a high-ish number would promote higher yields and higher yields favor the dollar, but that’s not what we have been seeing, despite consensus on 75 bp in Sept. Traders are still over-invested in the idea the Fed will stop raising rates next year and starting cutting them.

Oxford Economics expects Fed funds at 3.9% “early next year,” with QT continuing and the equivalent of additional hikes. The current rate is 2.25-2.50%, so that means a total of 140 to go, assuming “early next year” means Q1. If we get 75 in Sept and 50 each in the next two, that’s only 125, so another 15 in there somewhere. We don’t pretend to understand 15 bp when the Fed traditionally doesn’t do 15 bp, but never mind. 

Again, even assuming another 50 bp in both the UK and eurozone (with another 50 by the BoE before year-end), the US will be far, far ahead in the return, without necessarily having so much more inflation that it gets eaten up in the real return sweepstakes. The dollar “should” benefit. In fact, it should be benefiting now and especially if inflation is on the high side this morning and not subsiding, whether headline or core. But that’s not what we are getting, and we have yet to see a convincing explanation except the wishful thinking that has the Fed cutting next year. 

We can try the stock market. One newsletter writes “Over the last twelve months, CPI has come in hotter than expected two-thirds of the time. Notably, CPI has not come in below expectations over the last twelve months but did match expectations one-third of the time.

“In terms of market expectations, over the last twelve months, the S&P 500 has averaged an opening gap of -49 basis points following a CPI print that was hotter than expected. That’s about twice the average gap lower of 25 bps following all higher-than-expected prints over the last ten years. The market tends to gap higher following an inline print, averaging a gain of 5.3 bps over the last twelve months and 9.8 bps over the last ten years.”

So, if the number is 8.7% or lower, the S&P should go up, implying risk on, and the dollar falls as riskier stuff looks nicer, like the peso. If it’s a higher number, we can get a fat drop in the S&P, which might mean a gain for the dollar. This is at least consistent with the yield story. Again, this is the most hostile trading environment it’s possible to get, on a par with payrolls. Stay out.

This is an excerpt from “The Rockefeller Morning Briefing,” which is far larger (about 10 pages). The Briefing has been published every day for over 25 years and represents experienced analysis and insight. The report offers deep background and is not intended to guide FX trading. Rockefeller produces other reports (in spot and futures) for trading purposes.

To get a two-week trial of the full reports plus traders advice for only $3.95. Click here!