Very Helpful Cleveland Fed Answers My Question: CPI Isn't 10pc

The Cleveland Fed saw this post, which was inspired by seeing the 10 percent CPI figure whipping around social media. Doug set me straight on the calculation:

Thanks for your interest in the Cleveland Fed’s inflation indicators. I wanted to suggest an edit to your June 27 post; specifically, to clarify how the 10 percent inflation projection should be described. The Cleveland Fed inflation nowcasting model expects a month-on-month reading in June of 0.98% for CPI, translating to a year-over-year reading of 8.67%. Estimates here. The 10% figure is for CPI inflation during the second quarter, at a quarterly annualized percent change—that is, the average CPI level during April-May-June, divided by the average CPI level during January-February-March, to the fourth power, minus 1, times 100. So this inflation rate is different from the statistics that are most commonly reported.

This is a polite and helpful answer. "Read the website dumbass" is how I would answer myself, because it's right there on the page. I didn't read it carefully enough. I troll myself:

So, what did I learn aside from reading is fundamental? The 10 percent figure going around is bunk as presented. It's the annualized figure for Q2. Cleveland Fed's inflation model predicts a 1 percent mom increase for June and 8.7 percent yoy. 

The annualized quarterly number is a very helpful number for clarifying why I think second quarter GDP could be negative. The BEA similarly reports on an annualized basis and the Atlanta Fed projects 0.3 percent growth:

Atlanta Fed says the economy grew 0.3 percent (as of June 27) in Q2. Cleveland Fed says inflation is 10.0 percent in Q2. These numbers are not directly comparable because BEA has their own inflation calculation, but for illustration, the nominal GDP growth would be about 10.3 if both of these numbers were precisely accurate. A small adjustment in the inflation calculation could be the difference between a positive and negative GDP number, as could a divergence in inventories or imports.

Mish discusses where GDP it might be headed in GDPNow Forecast Inches Up to 0.3 Percent, But What's On the Way? He discusses how the Atlanta Fed model works too, and why good or bad data can produce opposite moves in the forecast because it's a model and adjusts based on incoming data and how that data relates to forecast. In other words, "less bad" can produce an upward move.

Whether BEA reports a technical recession a month from now will depend on inventory adjustments, imports and the GDP deflator. I think growth is going to be low enough that it could be within the margin of adjustment, such that statistical adjustments will decide whether it is positive or negative now and in the future. I believe with highest confidence that a recession will be backdated because I expect inflation (the BEA calculation) will be revised higher next year. If a positive number comes out next month, it'll be revised away next year. I'm not claiming the numbers are cooked, I do not believe numbers are politically massaged. Nor am I making the claim that CPI is undercounting real inflation. I'm speaking specifically of their model. I think they will revise their own numbers as they get a fuller picture from more complete data.

In sum, peak inflation is probably here on an intermediate time frame. It could rebound higher down the road, but right now it is rolling over. This assumes, because it is impossible to predict, that oil goes lower with QT and slowing demand, and that a shooting war doesn't erupt because Russia tries supplying Kaliningrad. For trading purposes, assume inflation has peaked and that inflation trades will keep taking it on the chin after a brief counter-trend bounce.

P.S. For all the crap many of us give the Federal Reserve and government economic calculations, the economists who work there are dedicated to reporting good data and quite helpful in answering questions. When I emailed the BEA for a GDP question way back when, I received a helpful answer within a day or so. The Cleveland Fed saw this post and responded, which is the most proactive outreach I've ever seen. This is the kind of responsiveness all government and quasi government institutions should have. The Federal Reserve also provides a ton of data at FRED. The addition of GDP and inflation forecast models from various regional banks provides excellent forecasting for the public. The individual investor who makes use of these resources will have a leg up on many, if not most, professionals.