Inflation: As ‘Contained’ As An Arrow From A Bow
Is 15 months in a row of rising core inflation ‘contained?’
Year-on-year core CPI has now risen for 15 consecutive months. At some point, it will seem reasonable to let it have a month off, but until now it hasn’t needed it. Fifteen months in a row. That’s impressive. It’s so impressive, in fact, that it hasn’t happened since 1973-1974, when prices were catching up from the failed experiment of price controls imposed by President Nixon in 1971-73. Core inflation has never, in the history of the data (which exists since 1957), accelerated for 16 consecutive months. So, next month we have a chance for a record!
Headline inflation was softer-than-expected by 0.1%, even as the NSA CPI index itself came in higher-than-expected. As I pointed out yesterday, that was a semi-predictable consequence of the change to new seasonal adjustment factors. Core inflation was 0.218% month-on-month, however, which actually generated a rise in the rounded year-on-year index to 2.3% (2.277% to three decimal places). The table below shows the evolution of the year-on-year changes for the eight major subgroups from 6 months ago to 3 months ago to last month, to now.
| Weights | y/y change | prev y/y change | 3m y/y chg | 6m y/y chg | |
| All items |
100.0% |
2.925% |
2.962% |
3.525% |
3.629% |
| Food and beverages |
15.0% |
4.212% |
4.452% |
4.470% |
4.001% |
| Housing |
40.2% |
1.876% |
1.874% |
1.869% |
1.453% |
| Apparel |
3.5% |
4.664% |
4.573% |
4.194% |
3.056% |
| Transportation |
16.5% |
4.961% |
5.197% |
9.185% |
11.980% |
| Medical care |
6.9% |
3.605% |
3.491% |
3.116% |
3.199% |
| Recreation |
5.9% |
1.372% |
1.027% |
0.253% |
-0.173% |
| Education and communication |
6.7% |
1.838% |
1.670% |
1.371% |
0.982% |
| Other goods and services |
5.3% |
1.740% |
1.701% |
1.660% |
0.847% |
Compared to last month, Apparel, Medical Care, Recreation, and Education/Communication accelerated, groups which total 23% of the consumption basket. Transportation and Food & Beverages both decelerated, and they total 31.5% of the basket. Now, notice that Transportation and Food & Beverages are the two groups that are most affected by direct commodity costs – energy and food, respectively. So…don’t get too excited by the deceleration there, although new and used motor vehicles and other components of Transportation also decelerated and that doesn’t have much to do with energy prices. In Food & Beverages, “Food at home” is decelerating (about 57% of the Food & beverages category) while “Food away from home” and “Alcoholic beverages” (the balance of the category) are accelerating.
Yes, you can get eyestrain looking too closely at these figures, but doing so does help.
For example, one theme I think the Fed is counting on is that the “Housing” component of CPI is expected to decelerate due to the still-high inventory of unsold homes and the fact that foreclosure sales can now proceed. It has been a conundrum why rents have been rising while home prices stagnate (actually, not much of a conundrum: there is an underlying inflation dynamic that in the case of the housing-asset market is being overwhelmed by a decline in multiples. But this is a conundrum to the Fed, and to be fair I also expected Housing inflation to be lower than it has been recently). And in this month’s data, you can see that the year-on-year increase in Housing CPI flattened out. But, as the table below shows, the Shelter component wasn’t what flattened out. Housing only went sideways because the “Fuels and Utilities” component declined – again, a commodity effect.
| Weights | y/y change | prev y/y change | 3m y/y chg | 6m y/y chg | |
| Housing |
40.2% |
1.876% |
1.874% |
1.869% |
1.453% |
| Shelter |
30.92% |
1.983% |
1.905% |
1.792% |
1.399% |
| Fuels and utilities |
5.27% |
1.941% |
2.432% |
3.483% |
3.201% |
| Household furnishings and operations |
4.03% |
1.035% |
1.000% |
0.561% |
-0.224% |
I still expect Housing inflation to level out and probably to decline, but so far those expectations have been dashed. It will be uncomfortable for the Fed if it remains this way; a significant part of their expectations for a visually-contained core inflation number is (mathematically) due to the expectation that housing inflation isn’t going to keep rising. As you can see in the chart below (Source: Enduring Investments http://www.enduringinvestments.com), the rest of core inflation outside of Shelter is continuing to rise. Inflation is not ‘contained’, except maybe for housing. Maybe.
I am fairly confident, though, that if Housing inflation does not decelerate as expected, then the Fed will find some other reason to ignore the very clear acceleration in inflation. The economists at the FRB are for the most part true believers in the notion that the output gap constrains any possible acceleration in inflation, despite ample evidence that output gaps don’t matter (or, anyway, matter far less than monetary variables). For another view of this proposition, see the Chart below, taken from this article by economist John Cochrane.
Fed economists also feel strongly that “well-anchored inflation expectations” means that they can ignore 15-month trends in core inflation, despite the fact that by Chairman Bernanke’s own admission we aren’t really very good at measuring inflation expectations (to be kind).
They have time. The Fed has recently begun to treat 2% (on core PCE, not core CPI) as more of a floor than a target, so it will be some months, even if core inflation doesn’t pause for a month or two pretty soon, before the Committee starts getting at all warm under the collar about inflation. Even then, they are extremely unlikely to take steps to reduce liquidity while Unemployment remains high. The Fed is in a political bind, and the only easy path for them is to “see no evil” on inflation while hoping that Unemployment drops swiftly enough for them to act before prices really get out of hand. We will see.
Leave a Reply to Michael AshtonCancel reply
“Core inflation has never, in the history of the data (which exists since 1957), accelerated for 16 consecutive months. So, next month we have a chance for a record!”
hi, it is certainly an interesting fact but the implication that inflation is breaking out like an arrow from a bow, is highly misleading, to say the least.
this measure is NOT current inflation, it is a comparison of the index now, vs 1 year ago. it would have been relevant to GRAB the bbg charts for series CPI XYOY and CPUPXCHG which show a/ that this 15 month run started at 0.6% and b/ was only that low due to a very sharp spike down in the monthly data, to below -0.1% (m/m) at the start of 2010. core CPI is currently about 0.2% m/m – an entirely typical level for it for quite some years. as it has not been above 0.25% m/m since 2007, so it is hard to see how the y/y figure can rise much above its current level anytime in the foreseeable future. obviously, for the y/y number to get above its current 2.3%, you need at least some of the monthly numbers in that 12 month period, to annualise to greater than 2.3%, right? that is simply not the case, and i say that as a numerical fact, not as expert analysis.
i expect this rise in y/y core CPI will flatten off noticeably in the next few months.
As I said, I expect that we will get some flattening for a few months in the core CPI, because a large part of the index is still recovering from a bubble. But core CPI is certainly not going to be lower a year from now, ex-housing and probably including housing, than it is now. We can certainly differ in our opinions. It would be shocking if y/y inflation did not hit 3% in the next year (at least, ex-housing), with some chance of being not-insignificantly higher than that depending on a number of variables beyond the scope of this reply.
However, the measure I’m accentuating is not the difference in the price index now, versus a year ago. It is the difference in the rate of change in the index – that is, it is acceleration (which is why i call it acceleration) rather than change. The year/year increase is becoming larger each month. And the ‘spike’ down in October 2010 was, as you can tell if you look at the data, not at all a spike but rather a fall due entirely to the crunch in housing. I know that because in early 2010 I was predicting the low would be in September or October 2010, after years of saying that inflation was decelerating.
Have you seen this http://www.clevelandfed.org/research/Data/inflation_expectations/
I am curious what you think of it?
I read their research piece when it first came out. While I admire the Fed for trying to figure out how to measure inflation expectations, this is clearly not it. You can look at the data and, by talking to 10 people you know, prove to your satisfaction that it isn’t right in that it’s not really capturing true inflation expectations.
I recently published a paper laying the groundwork for how to model inflation expectations; it was published in January’s edition of Business Economics (the journal of the NABE). I think it’s much more realistic and thoughtful than what the Cleveland Fed has done here.