Here is an interesting data point that may have been missed in today’s noise: the core CPI surprised jumped by 0.349%, barely missing out on a 0.4% print. Had the BLS not revised its seasonal adjustments (which dragged the Jan CPI by 4bps lower vs the old seasonal factors), the monthly CPI jump would have been 0.4%.
A second notable data point: According to SocGen, in January, 43 of the 62 core CPI subcomponents were positive on a monthly basis, the most since May 2016, and above the 2017 monthly average of 35. However, while the breadth of gains was impressive, the strength was concentrated in a few components which the French bank does not expect to put persistent upward pressure on the core rate moving forward. Furthermore, what stands out is that about 14% of the core CPI was responsible for nearly 50% of the rise in the core.
And a third point of note: Of all CPI components, apparel was the biggest contributor by far to January’s inflation spike (0.064pp), posting an eye-popping 1.7% SA rise in January. That was the fourth-largest increase since 1947 and the biggest since a 1.8% jump in February 1990. On an NSA basis, it registered its biggest January increase since 1951.
Yes but… according to SocGen, Q1 strength has been a feature of this category over the last few years (likely reflecting changing discounting patterns. However, that Q1 strength has generally proved short-lived. Indeed, looking at first prints of apparel, Q1 over the 2015-2017 period averaged a rise of 0.4%, but the balance of the year it averaged a drop of 0.2%.
It wasn’t just apparel, however: in January, the BLS reported that several components posted the largest gains in their respective histories. As SocGen analyst Omair Sharif writes, “we cannot recall having seen that many “L-EVER” mentions in any CPI report before. Perhaps they were early-year price increases that are unlikely to be repeated, but some, in our view, could be related to changes in methodology and seasonal patters.“
For example, the “window and floor coverings and other linens” category, which is a measly 0.3% of the core CPI, jumped by 4.1%, adding 0.135pp to the core by itself! Within appliances, laundry equipment climbed by 3.4%, perhaps reflecting the new tariffs on washing machines, and the reported price increases instituted by firms hit by the charges. Women’s apparel climbed by 3.4%, which helped to boost the overall apparel category. The information technology commodities gauge rose by 1.0%, with the telephone hardware component climbing by 2.5%. Elsewhere, the medical care index increased by 0.41% SA, the strongest reading since August 2016. The bulk of that rise (about 85%) was centered in the hospital services index, which climbed by 1.3% and added 0.037pp to the core increase
In terms of some of the other components, rent climbed by 0.34% while OER advanced by 0.28%. There was less giveback in those shelter components in January than we had expected. The new vehicles index fell by 0.1% while airfares slid by 0.6%, both of which were in line with our expectations. Hotel rates tumbled by a surprising 2.4%, the third-largest decline in four years. Recreation was softer than expected, posting an unchanged reading, while communication was also flat.
* * *
But the bottom line about this particular inflation report is that it was too good – or rather strong – to be true. Quote SocGen:
Out of 62 subcomponents we track in the core, 14 posted monthly increases that were more than two standard deviations above their norm.
That was the highest number since June 2008 and was well above the 20-year average of 9 components. In other words, there were a number of outliers that boosted the monthly core rate by much more than their usual contribution. In fact, January tends to be an extremely volatile period with exaggerated strength. Indeed, in the last 20 years, January holds four of the top ten spots for the month with the highest number of outliers.
Which brings us to the bank’s conclusion and why it is skeptical of the CPI’s strength:
First, a large number of core components were, in our view, outliers to the upside and exaggerated the core advance. Second, although many core items increased in January, much of the strength was concentrated in a handful of items that seem unlikely to be persistent sources of strength moving forward. In short, we are still comfortable with our view that the core CPI will advance in 1H and climb to around 2.0%-2.1% by mid-year before moderating in 2H
Judging by the schizophrenic market, while the dollar agrees thoroughly with this skeptical assessment, 10Y bonds couldn’t disagree more. Stocks, of course, go with whatever says to BTFD.