Inflationary Questions
Last week I compared our jobs data, which is sometimes questionable, to World War II weather forecasts. Those were also questionable but necessary anyway. The generals needed them “for planning purposes.”
This analogy has a flaw, though. A forecast for future weather isn’t the same as data on past weather. The data—wind speed, barometric pressure, and so on—is the input that goes into the forecast. Assuming the instruments work, the data is what it is. It may or may not lead to an accurate forecast.
With economic data, the challenge is that our instruments are more error prone. The monthly jobs reports rely on employers responding to surveys correctly and quickly. The fact this doesn’t always happen causes revisions that call the whole process into question. That’s not because anyone is hiding anything. The process is understood and transparent. Just by their nature, surveys are error prone and even more so of late.
Notice the revisions below cycle through periods of mostly higher, then mostly lower. This is because the BLS’s methodology is backwards looking and trend following, which causes large revisions when the trend changes. The data responds more slowly at turns in the economic and employment cycle. By the way, notice that most of the revisions are undercounts. That is because the economy was growing most of the time since 2008.
Source: Unleash Prosperity
All that said, I think a big part of the problem is not with the data itself, but with how we all use it.
Many, perhaps most people don’t look at economic data with open minds. They’ve already decided which economic narrative they believe; they want data to confirm it. If the data doesn’t cooperate, they think the problem is in the data, not in their narrative.
I’ve talked about this before. In behavioral economics it’s called “confirmation bias,” an incredibly powerful psychological force that can unconsciously twist our thinking unless we make a concerted effort to resist. And believe me, it takes effort. I am the chief sinner.
Confirmation bias is even stronger in today’s politically polarized atmosphere. It affects all sides. We assume data will be good when our side is in power and bad when it isn’t. If the data doesn’t reflect our view, we often discredit the data rather than asking whether we should change our view.
We should all want accurate data above all else. Then we can have productive debates about what to do next. We also have to accept that our data is never perfect; the economy is just too big and complex to summarize in a few charts.
Our data methodology has room for improvement. We can always do better. Unfortunately, the economy doesn’t stop while we rethink everything. Revising those methods is kind of like changing the oil while your car is moving. If you do it, you’d best do it carefully.
That being said, I am part of a chat group that has been vigorously debating how we should actually develop the BLS employment data. I have generally been on the side that the BLS is doing the best they can and the flaw is in the surveys. Another group said we should use AI and known data. Color me skeptical.
Then one person simply went to Chat GPT 5.0 and asked some very specific questions about what we can do to improve the BLS unemployment data. To my amazement, Chat GPT came back with a rather lengthy explanation of how we can use current databases of actual employment, government databases, etc. and an actual plan to create a new employment analysis system over the next five years. Very public and transparent.
It turns out there are a lot of databases on employment out there, and the AI suggested ways to combine them with unemployment data from the government to create a more real-time unemployment database. Given the problems with the current methodology, it’s certainly worth looking at. And by the way, it would be much less expensive.
I say all that as a preface to this week’s big data points, the CPI and PPI inflation reports. I’ve written many times about the many flaws in these reports. But we can accept those limitations and use them to inform our decisions.
Inflated Supercore
Markets celebrated the July CPI inflation data, but not because it improved. It didn’t. Stocks rose because inflation, at first glance, appeared to have stabilized enough to let the Fed address rising unemployment with a rate cut next month.
Unfortunately, looking deeper than the surface says inflation remains a serious concern. While headline inflation held steady at 2.7%, it had a lot of help from falling gasoline prices. Core inflation (ex-food and energy) rose 3.1% from a year ago.
Worse, the so-called “Supercore” index, which also subtracts housing costs, was up 3.6%.
Source: RSM
This chart highlights the monthly change by category.
Source: RSM
The categories that rose more than average tend to be either labor-intensive or imported. Shelter costs don’t stand out in percentage terms but remain a problem because of their large weighting in the index.
Here’s the quandary in a nutshell: Service inflation, including but not limited to housing, isn’t improving. Now the goods categories that had been showing low or even negative inflation are beginning to tick higher.
Add those two together and it’s hard to make the case inflation is on the mend. Here’s David (Rosie) Rosenberg:
“On the low-flation side, we did get benign readings out of hotels/motels (-1.3% MoM, and riding a five-month losing streak), appliances (-0.9%), new autos (flat), video/audio equipment (flat), apparel (+0.1%), and toys (less than +0.2%).
“But there were too many items with outsized increases that make me a touch uncomfortable: air fares (+4.0%, ending a five-month string of declines and the biggest print since May 2022), home improvement (+1.6%), delivery services (+0.9%), furniture (+0.9%), medical care services (+0.8%—driven principally by a record increase in dental costs), recreation services (+0.4%), used cars (+0.5%—this was a surprise), sporting goods (+0.4%), and personal care goods and services (+0.4%).
“The rental metrics remained stubborn, at +0.3% MoM, as has been the case for the better part of the past year, though the broader-term trends are slowing down. But the overriding issue here is that there were more items in the CPI showing a pulse than playing dead. If the Fed decides to cut rates, it is not because of disinflation but because it has shifted its attention to the weakening labor market. All in all, however, I am less than impressed with this particular report.”
This week’s Producer Price Index report was on the high side, indicating that reducing inflation on the retail side may be an issue over the next few months. Expectations were for a rise of 0.2% but what we got was a rise of 0.9%. These reports do swing around a lot, but this is not what we want to see. This graph is from Mike (Mish) Shedlock:
Source: MishTalk
So, assuming recent patterns continue, the key question is whether goods price inflation will continue accelerating. To answer that question, we must discuss the elephant in the room.
Structural Breaks
Let’s take a long-term look at inflation. This chart shows headline CPI for calendar years 1985–2024—the last four decades.
Source: FRED
In the 1990s and early 2000s, inflation stayed generally between 2% and 3%. The range shifted after the 2008 recession, falling mostly between 0% and 2%. That was the era when Fed officials worried inflation wasn’t high enough. They got their wish in 2022. But in the meantime, we all came to think sub-2% inflation was normal. Federal Reserve and mainstream economists talked about “structural shifts” that would keep inflation low, allowing for lower interest rates.
|
A prime reason inflation stayed so low for so long was our growing reliance on low-priced Chinese goods. This had benefits for Americans but also costs, which were unfortunately concentrated in certain regions and occupations. The resulting backlash is the reason President Trump is now using tariffs to discourage imports.
Let’s note something which should be obvious but deserves repeating. Rising goods inflation isn’t an unintended side effect of tariff policies. It is one of the goals. Higher prices for imported goods are supposed to incentivize more domestic production, create more manufacturing jobs, and make the US less reliant on imports.
That’s the theory, at least. Whether it will have those effects is another subject. But the higher-prices part is here, and it’s becoming evident in the inflation data.
Sidebar: The new Federal Reserve governor nominee, Stephen Miran, Trump’s current head of the Council of Economic Advisers, actually advocates for a weaker dollar that would create opportunities to manufacture more products in the US as foreign products begin to cost more. I would argue quite aggressively this is a bad idea, but the point is that there are advocates in this administration for these policies.
Rather than throw more numbers at you, I’ll share a chart that makes it clear. Each bar below shows annualized inflation by month since August 2024. They are divided by color to show the contribution by category. Segments below the zero line were negative that month, adding a disinflationary influence. What jumps out at you?
Source: EconReporter
The biggest inflation factors are obviously shelter (purple) and core services (red). Food prices (blue) have been a smaller but steady contributor. Energy prices (yellow) helped pull inflation down in nine of these 12 months. So did core goods (green) until December.
Starting in April, core goods began running above zero instead of below it (see dotted circle), with an increasingly positive contribution each month.
Correlation isn’t always causation, but in this case it seems pretty obvious. In April the president announced his intent to aggressively discourage imports by imposing high tariffs. The rates are settling down to lower levels than Trump initially ordered, but still much higher than before.
To be clear, most of this inflation was already in the mix before Trump took office. Shelter and core services are still the main sources. But the recent increase in inflation, and particularly goods inflation, looks tariff-driven.
For example, here are the one-month (not annualized) price increases the CPI shows for some import-heavy goods.
- Infant & toddler clothes +3.3%
- Women’s dresses +2.7%
- Tools, hardware & outdoor equipment +1.6%
- Dishes & flatware +2.0%
- Bedroom furniture +1.5%
- Coffee +2.3%
(I am very curious about how the public will react if coffee keeps rising at that rate. Most people don’t consider it optional. And, other than small amounts grown in Hawaii and Puerto Rico, the US has no way to produce coffee domestically. Or bananas, or many other goods.)
Note also, price increases aren’t limited to the goods that are being tariffed. The effect spills over as demand increases for US-made substitutes.
Other research supports this, too. Alberto Cavallo of the Billion Prices Project, an alternative inflation measure I’ve mentioned before, has a new paper showing how the US marked a broad-based inflation trend change this year. He found 75% of the non-shelter CPI categories recently had what he calls “positive structural breaks.”
Source: Alberto Cavallo
I could go on, but I think you get the point. A counterpoint is that the tariff effects will be a one-time increase that policymakers should “look through.” That’s a reasonable position but presumes tariff rates and related policies will reach some point of stability. I will “look through” more easily when we get the necessary and needed stability.
Ineffective Tools
The Fed is caught between inflation and unemployment. Which will it decide to fight? Wall Street thinks the latest reports show unemployment is a bigger problem, and the Fed will therefore cut rates at its next meeting, with more cuts likely to follow.
(Note the Fed faced a similar problem in the 1970s. Then-Chair Arthur Burns kept saying that price increases on this or that item or service were one time and we really needed to “look through” the data points. He was clearly wrong. I’m sure every FOMC voter knows the risks of repeating that mistake.)
Stock prices respond bullishly to falling rates for a purely mechanical accounting reason. Lower interest rates raise the net present value of future cash flows, which is the basis for stock valuations. The Fed can magically make equities more attractive even if nothing else changes. Of course that’s what bullish Wall Streeters want. Will they get it? The futures market says so. I’m not so sure.
For one thing, FOMC voters consider themselves data driven. Before their next scheduled vote on Sept. 17, they’ll get the second estimate for Q2 GDP on Aug. 28, a PCE inflation report (Aug. 29), the August jobs report (Sept 5), and a CPI report (Sept 11), plus a host of lower-profile releases. Any of those could change the outlook.
The fact that all this data and its compilers are suddenly under a new level of scrutiny might complicate the situation a bit, but I don’t think it will affect their votes. FOMC members will, like generals do with weather forecasts, consider the reports “for planning purposes” and apply whatever level of doubt they think appropriate.
New governor Stephen Miran’s presence at the meeting, assuming the Senate confirms him by then, will add another complication. He may dissent, either because they don’t cut rates or don’t cut them enough, but I don’t think he will change the outcome. It will be significant if we see three of the seven governors dissent. Two did last time and Miran may join them. I’m not sure that has ever happened before. Regional Fed presidents can dissent, but multiple governors dissenting is rare.
Marketwise, the biggest tantrum risk is if the committee doesn’t cut because a majority thinks inflation is still too high. As of now, it’s hard to argue inflation is near their goal. The Atlanta Fed maintains an Underlying Inflation Dashboard that shows various inflation measures relative to the Fed’s 2% target.
Source: Atlanta Fed
The red backgrounds are a color code of how far each benchmark is from the target, with each adjusted to a level that would match 2% core PCE. Here’s the color key.
Source: Atlanta Fed
All red means all nine inflation measures are more than 50 basis points above target. They can always change their target, of course, or just ignore it. This chart was all red a year ago, too, and it didn’t stop them from cutting rates. But we know some now regard that move as a mistake. It’s hard to imagine them doing it again.
There’s reason to think inflation could rise over the near term. The minor-but-visible tariff impact we see so far is happening even though companies are still absorbing most of the impact. A Goldman Sachs analysis found that as of June, US consumers were only bearing 22% of the price impact.
Source: Scott Lincicome
This suggests the tariff impact on core goods could rise from here if companies decide to pass along more of the tariff costs as price increases. (Of course, if they don’t pass them on that means lower profits, which is also a problem.)
Why haven’t businesses done this already? I suspect most are still in wait-and-see mode. Nobody wants to aggravate customers only to see the tariffs get negotiated away or stopped by court decisions. Better to absorb the costs as long as you can. And with a lot of inventory built up before higher rates took effect, they’ve been able to manage this. But they can’t do it indefinitely if they want to maintain profit margins.
I also suspect the recent lull in new hires is a result of lower profit margins. Remember, monthly swings of 100,000 jobs are still a very small percentage of the number of people employed. The employment number we all obsess about is actually quite small in the grand scheme of things and happens “on the margin.” And if a small percentage of businesses are not hiring because their profits are squeezed, it shows up “on the margin” in the employment numbers to which the Fed pays attention.
Remember, too, weak energy prices have been suppressing headline CPI and PCE. That could change at any time for all kinds of reasons. Remove that disinflationary factor and the situation could change faster than we would like.
The Fed could easily find itself in the worst of both worlds: rising inflation and rising unemployment at the same time. Then what? I really don’t know. What kind of job-producing activity that isn’t already happening will take place if rates drop a percentage point or two? Is that really enough to put employers in hiring mode? Similarly, what kind of demand will disappear if the Fed holds rates at the current level or a little higher?
We may be approaching a time when the Fed’s tools just aren’t suited for the problems we face. I criticize central bankers a lot. I’ve even fantasized about a world without them. In the next year or two, we may see what that world looks like. I am not sure that I will like it.
West Coast Fishing, Austin, and Houston
Shane and I will be leaving in 10 days to go to British Columbia and the West Coast Fishing Club to meet with 33 readers hoping to catch trophy salmon, halibut, and cod. Last year was one of the best experiences I’ve had in a long time just in terms of conversation and relationships. I am really looking forward to this.
In the middle of September, I will be going to Austin. There is a weekly Zoom group that I am on that has about 40 participants and many of us are flying into Austin for a birthday party, and as it turns out a lot of other cool people will be there. And I have lots of friends in Austin.
From there I will fly to Houston to be with David Bahnsen and his team at their office there. If you are an accredited investor and interested in joining us for dinner, drop me a note at business@2000wave.com. We will talk about portfolio management and then of course have an ask David and me anything session.
It is good to be back in Puerto Rico. The door-to-door round-trip from Puerto Rico to California is a 14-hour day and just tiring.
You have a great week and let’s enjoy the rest of the summer! And don’t forget to follow me on X!
|
Your hoping the Fed can navigate this conundrum successfully analyst,
Scroll down for comments.