Numbers Behaving Badly

Inevitable Inconsistency

Last week, I talked about “confirmation bias,” our tendency to seek out data that supports our existing beliefs while ignoring or discounting data we dislike. We’re all prone to this, which is why the best friends are those who tell you things you don’t want to hear.

We see this now in the talk about the Federal Reserve’s next move. It is endless and everywhere. Nonetheless, their decision hinges on data that is increasingly subject to dispute.

This is partly because we have more numbers now. Inconsistency is almost inevitable when we measure so many things in so many different ways. Often the differences have good explanations, but by that point it doesn’t matter. Many people will see what they want to see.

On paper, at least, the Fed’s job isn’t complicated. The goal is to keep the economy balanced between price stability (i.e., low inflation) and full employment. Both these would be subjective even if we had perfect data. How much inflation is acceptable? I’ve argued their 2% target is too high. I have been in the room with several Nobel laureates who maintained the target should be 4%. Further, what do we consider “full” employment? Equations can’t answer these questions.

Ed Yardeni recently walked through the latest numbers in his private letter. He started with this and then shared several charts.

“The Fed is required by law to keep inflation low and stable while also maintaining full employment. Achieving that dual mandate isn't always easy. Currently, the labor market is at full employment [Ed’s definition thereof –JM], but there are a few signs of weakening. Inflation was on track to fall to the Fed's 2.0% inflation target but has been stuck around 3.0% recently. Some Fed officials believe that Trump's tariffs are only temporarily boosting inflation and that the Fed should ease as soon as possible to avert any further weakening of the labor market. Other Fed officials are opposed to easing until they can be more certain that inflation is declining to 2.0%.”

Ed falls on the hawkish side because he sees inflation is still elevated. He pointed to an interesting indicator with this chart. The red line, which sums up the ISM survey for manufacturing and non-manufacturing prices paid, tends to lead PPI final demand inflation (the blue line) by about six months. The chart shifts the red line six months forward, showing where PPI inflation “should” go in the future. If accurate, it means inflation will keep rising at least until year-end.

As for employment, Yardeni notes (as I did recently) that initial unemployment benefit claims have been steady at a low level. Employers are holding on to workers but are reluctant to hire more. That’s causing a higher unemployment rate in younger groups.

Look at the dark blue line in Ed’s chart—workers and jobseekers in the 16–19 age group. Their unemployment rate is 15%, about where it was heading into the 2008 recession. The next youngest (20–24) isn’t far behind. Both are rising in a way that, in the past, would have said the economy is already well into recession.

Obviously, this is awful for young jobseekers. The other side, though, is the much larger number of older workers (the 25–54 “prime age” group as well as the 55+ cohort) who are still employed at historically strong rates.

The current employment problem is highly concentrated in the youngest workers. We need policies to help them, but it’s hard to see how lower interest rates would make much difference. Yardeni sees this as more reason to be hawkish.

Considering all this, it’s a strong argument against looser Fed policy. The combination of persistent (if not rising) inflation and steady employment doesn’t suggest any need for rate cuts. But are these more numbers behaving badly? Let’s look at some others.

Pent-Up Demand

The numbers say that Americans over age 25 who want to work are mostly employed. That doesn’t mean they are happily employed. The post-COVID era of job hopping and “quiet quitting” seems to be over. Just as employers are reluctant to hire, workers are reluctant to leave. This helps explain why wage growth is slowing.

The chart shows wage growth for both current workers and new job postings running around 3%. If the official inflation numbers are roughly correct, this means real wage growth is close to zero. And often below zero, since some major living costs (housing, healthcare) are rising faster than 3% for many households.

What do people do when their cost of living rises faster than their wages and they can’t easily find better jobs? Credit can sustain lifestyles for a while, but eventually they cut back on spending. Multiply that by millions of workers and the result is recession, or at least a consumer-led slowdown.

If that’s the case, the Fed arguably needs to get ahead of it ASAP. But it must do so in a way that doesn’t aggravate inflation. How do they balance this?

Bary Habib says they may not need to because the inflation numbers are behaving badly. He thinks both CPI and PCE are overstated because of the way they incorporate shelter prices. Here’s Barry.

“Zillow reported that blended rents in July fell from 2.9% to 2.6% year over year. We have been seeing a deceleration, but the figures in CPI and PCE are much higher.

“The biggest reason is the BLS is using guesswork and imputed figures. Owners’ Equivalent Rent is at 4.1%, and when adjusted vs. the Zillow numbers, and applying the weightings in with CPI, Core CPI is being overstated by 0.5%. That means that Core CPI would actually be 2.6%, instead of 3.1%, which is much better.

“Applying the same formula for Core PCE, and using the weightings, Core PCE is overstated by 0.2% and would be 2.6% instead of 2.8%.

“Both of the year-over-year Core inflation readings at 2.6% also include some impact of the tariffs, which is hard to measure.”

I agree with Barry that a 2.6% read on core CPI is much better than the 3.1% reported for July. But even at that adjusted level, inflation remains well above the Fed’s target. The numbers are still behaving badly.

Barry went on to say lower rates might (counterintuitively) help reduce inflation by enabling growth in housing supply. More:

“Housing Permits, the most forward-looking supply indicator, fell almost 3% last month and continue to trend lower to the lowest reading of the year. All of the decline was in multi-family, with single-family up a modest 0.5%.

“Housing Starts increased 5% to 1.43M, but most of the gain was in multi-family. The single-family starts level is the second-lowest of the year and points to limited supply coming to the market from the new construction front in the future.

“Completions rose 6%, with all of the gain coming from single-family. This should help with a bit more single-family supply in the near term.

“There is a lot of pent up demand and people on the sidelines waiting for rates to come down. If the Fed cuts rates and mortgage rates fall, there will be a lot more demand.

“The NAHB Builder Sentiment report has stabilized around 32 the last few months, which is a weak figure and in contraction. Traffic increased two points to 22, but from anemic levels.

“Builders’ biggest challenges are elevated mortgage rates, weak buyer traffic, and ongoing supply-side challenges. They said that affordability is the biggest challenge, and builders are pushing for the Fed to cut rates to help financing costs and indirectly help mortgage rates and in turn demand.”

Barry knows housing better than anyone in my Rolodex, so I won’t argue with him. I do wonder how low rates would need to go to achieve this effect and how long it would take. The Fed cut rates last year and the ten-year bond yield went up.

AI Is the Swing Vote

The one place we see unambiguously good news lately is the stock market. The indexes have occasional down days but remain near record highs. Investors are betting heavily that artificial intelligence technologies, and construction of the needed infrastructure, will outweigh other economic headwinds.

Here’s a capex chart almost too hard to believe. Look how much capital spending plans grew at Alphabet, Amazon, Microsoft, and Meta since ChatGPT’s 2022 launch.

The trillions in spending these four companies are planning (estimated to be almost $7 trillion) in 2026–2030 is staggering. All of it will go into someone’s pocket, constituting a giant economic stimulus comparable to anything the federal government could do.

That’s impressive but again, these numbers may be behaving badly. Will spending really reach those levels? Are all the things the companies want to buy really available? And when will it produce actual profits?

This is important because the entire economy really needs the AI boom to work. Brad Delong, who shared the chart, put it this way.

“It is anticipation of AI‑driven productivity—and the capital expenditures to chase it—that keeps the ship sailing forward. But will, in the end, these software architectures be useful enough for their broad adoption? Will enough of the value generated be diverted from consumer to producer surplus?

“… Whether this ends in a bust or a true transformation, and how big a bust, is for now hidden behind an impenetrable veil of time and ignorance. For now, we see that without this, employment and production would almost surely be in a Bad Place, and that there is not yet a marked-out path to mammoth profitability consonant with the mammoth investment scale.”

Try to imagine what the economy would look like right now if AI expenditures weren’t stimulating growth and creating a wealth effect on high-end consumer spending. How will these companies generate the multiple trillion dollars of revenue needed to show a reasonable return on investment of $6 trillion by the end of the decade?

Remember, too, these data centers are more than AI. We also have monstrous demand for cloud services. But the capital investment is the same. I think Brad is right: We’d be in a Bad Place.

(WSTFD) What Should the Fed Do?

I asked a number of my friends who think and care deeply about Fed policy about what they think the Fed should do at the September meeting. Normally, there is some sort of consensus in the perspectives. Not this time. And not just among my friends. I read a lot of source material and letters. This is a very disputed topic that does not go the way I thought it would. (Note: This was asked prior to Jerome Powell's speech on Friday. See below.)

I mentioned Ed Yardeni and Barry Habib above. Ed thinks they should do nothing. Barry wants a cut and the promise of more.

David Kotok, who has made his long career deeply studying the Fed and counts many former governors and regional Fed presidents as close friends, sent along this thoughtful note:

“John, I hope you’re doing well.

“My opinion is consistent with what we learned from the minutes. The inflation risk is greater than the recession risk, so I favor no change based on what we know today (Wednesday night). The pressure on the Fed is unlike anything I recall from my half century-plus career. The latest personalized attack on Cook is extraordinary and may succeed because of the double mortgage application issue.

“In my opinion, an easing Fed that moves too much and too soon can take a tariff-triggered price level adjustment and monetize it into a wage-price spiral. Waiting too long means a slowdown becomes a recession because we are taking the labor force lower in headcount and therefore raising wage pressure. Watch behavior of steel workers and autoworker unions now that they have protectionist tariffs.

“Thank you for asking. I miss fishing together.

—David”

David Bahnsen simply replied:

“Yes, they should cut (and I don’t believe their present level has anything to do with inflation risk).”

This squares with my belief that Fed policy is already becoming less meaningful (except at the extremes).

David Rosenberg (Rosie) noted:

“I’m in the camp that believes that the economy is slowing not just on the demand side but the supply side too. So from an inflation standpoint, it’s a wash. I think the Fed should stand pat next month, not just because of that but also the latest inflation statistics were more disturbing on the services front than the goods part of the equation. And we all know or should know that the tariff impact lies ahead of us on that score, the only question is magnitude and persistence.

“I’d rather the Fed await more decisive signs of slack building up before moving. Not to mention the super-inflated multiples in the stock market, razor-thin credit spreads and this year’s USD bear market. Nothing in the financial markets is screaming for a rate cut around the corner. Only the President and his long list of [friends], two of them currently sitting on the FOMC and another set to join, seem to think it’s a good idea.”

The normally super-inflation hawk Dr. Lacy Hunt thinks the Fed should cut a lot more than 25 bps. Like 1% soon. His rationale?

“I believe that many risk markets are in the condition referred to as ‘Mania’ by the late and great MIT professor Charles Kindleberger. This occurs when valuations are extremely high and in many past cases when valuations were unprecedentedly high. My friend Burt Malkiel wrote in last Friday’s NYT that stock market valuations are among the highest in the past 230 years. I know from my own research that the S&P price-to-book ratio is at the highest since 1947 and the price-to-sales ratio is at the highest since 1995. I see a major discrepancy between the S&P earnings and the corporate earnings in the national income and product accounts.

“To me this is a risk since a recent analysis conducted by the WSJ found that more than 70% of the earnings reported for Q2 have not followed GAAP. I can list numerous other serious risks that are being completely ignored. No one including me can know timing, but I believe Kindleberger’s research shows unequivocally that ‘panics’ and ‘crashes’ follow manias.”

Peter Boockvar had the best answer (IMHO). Summarizing:

“I’m actually indifferent if all they cut is 25–50 bps over the next year. The real mistake is if it is any more than that. They need to make a hawkish cut.”

I had to ask what the heck is a hawkish cut? His answer:

“It means the Fed must communicate they only have 1–2 rate cuts left and anyone thinking they are going to 3% anytime soon is wrong. We are in a rate tweaking cycle, not a rate cutting cycle.”

As for me? I lived as a very young entrepreneur in the ’70s. It was a stone-cold bitch. Then-Fed Chairman Burns looked through the inflation in various items and essentially said it was transitory. I don’t have space to go into detail about a longish communication with Ed Easterling, but essentially the next five months will give tough year-over-year comparisons for inflation numbers. And that’s before the tariff effects. Should we look through those as transitory?

I think Job One for the Fed should be inflation. Not sure how in this environment 25 bps makes a difference, so a cut is possible without hurting. But cautious steps should be the clear communication, no matter what they do.

(All of the above was written before Friday morning, on a plane with multiple crying babies, so I might have been distracted.)

Now, What Will Jay Do?

Friday morning, we read Powell’s speech at Jackson Hole. It was modestly dovish. Basically, after he talked about real inflation risks, a few sentences later he indicated they would make a rate cut, “…with policy in restrictive territory, the baseline outlook and the shifting balance of risks may warrant adjusting our policy stance.” That was what the market wanted to hear.

Why employment risks over inflation risks?

“Overall, while the labor market appears to be in balance, it is a curious kind of balance that results from a marked slowing in both the supply of and demand for workers. This unusual situation suggests that downside risks to employment are rising. And if those risks materialize, they can do so quickly in the form of sharply higher layoffs and rising unemployment.”

With his adamant assertion that the Fed would be independent as always, and looking at the data, it looks to me like Powell is adopting something similar to Peter Boockvar's “hawkish cut.”

The economy is vulnerable in multiple ways. Vulnerable to an AI disappointment, vulnerable to inflation, and vulnerable to lower consumer spending. Any of those would be a problem and we could see more than one.

I’ve written before about the multiple cycle theories that all point to climactic events around 2030. I think those events are already underway and poorly behaved numbers are disguising them.

This doesn’t have to end badly; in fact, I think it will end with our society and economy in much better condition than they are now. But we have to get through the hard parts first.

To get through this, we need to cultivate two seemingly opposite skills. We have to proceed tentatively but also confidently. Tentatively because we really don’t know what lies ahead, but confident we are ready to handle it.

Dallas, British Columbia, Austin, Dallas, and Houston

If you are reading this Saturday morning, I am flying back to PR from Dallas where I had yet another session of Therapeutic Plasma Exchange (TPE) at our new Dallas (Frisco) longevity clinic (partnered with Dr. Mike Roizen). My Mauldin Economics partner Olivier Garret joined me. I will shortly be getting you more information on the procedure and why you should consider it.

Shane and I fly to Vancouver ALL day Tuesday and meet up with 30+ readers to go on a fabulous salmon, halibut, and cod fishing trip at the West Coast Fishing Club, then back Labor Day. I will be in Austin for Brad Rotter’s huge birthday party weekend of September 12–14. For those who used to follow Dennis Gartman, Brad was a huge floor trader who made the famous statement: “I'm flat and I'm nervous.” He is uber-connected, and the guest list shows it.

Monday night (September 15) I will be in Dallas to host a dinner for readers and friends who are interested in knowing more about our Dallas longevity clinic and just simply sharing bread. The next night (the 16^th) I will be in Houston at an event for The Bahnsen Group’s clients and my readers who are interested. If you would like to join either event, simply drop a note to [email protected].

With that I will hit the send button. You have a great week and enjoy family and friends.

Your both tentative and confident analyst,

John Mauldin

Prepare your bond portfolio for changing market conditions. Register today for the Fixed Income Symposium on Sept. 18, 2025, 11AM ET / 8AM PT.

© Mauldin Economics

Read more commentaries by Mauldin Economics