This Year’s Job Market Is Shaping Up to Be Surprisingly Stable June data was underwhelming. But today’s job market is much less concerning than it was six months ago.

Halfway through the year, the labor market is flashing a virtue that proved elusive for much of 2025: stability. American hiring in 2026 hasn’t quite boomed, but it has broadly improved.

The latest numbers, released Thursday by the Labor Department, were a disappointment. June’s 57,000 new jobs fell well short of Wall Street forecasts. Sectors that reflect the current health of the economy, including retail and leisure and hospitality, lost jobs.

Yet look beyond the monthly number, and the job market has steadily, if not spectacularly, added an average of around 92,000 jobs a month so far this year. That is a giant leap from average net losses of 8,000 a month over the second half of 2025. (…)

Then there is the unemployment rate: In June, despite the lower-than-expected hiring, it actually dropped to 4.2% from 4.3%. That is good news on its surface, but the cause is less than ideal.

Thursday’s data indicates that the unemployment rate was pulled lower by the unusually large decline in the number of Americans who are either employed or looking for work. With fewer people in the job market, the unemployment rate can fall even when the number of jobs created is falling too. (…)

The size of the workforce has been stagnating for months. Retiring baby boomers are one likely reason. Immigrants dropping out of the workforce during immigration crackdowns is likely another. There were about 169 million people in the civilian labor force in June, down nearly 2.2 million from a peak in November. (…)

In each of the past few months, more industries have been adding jobs than losing them. A year ago, the opposite was true. (…)

When July data arrives next month, economists will be watching closely for confirmation that the June labor-force drop wasn’t the start of a trend. (…)

The WSJ needed to sugarcoat June’s job report after its June 8 headline The Great American Job-Creation Machine Comes Back to Life.

My June 8 post was titled Soccer Punch! pointing out that Leisure and Hospitality contributed 70k to the 172k new jobs in May (41%), well above its normal monthly average of 14,000 as hotels, restaurants and attractions were staffing up ahead of the World Cup. FIFA expected the event to generate the equivalent of about 185,000 full-time jobs in the US alone.

In fact, “data compiled ahead of the tournament by hospitality agencies like Oysterlink highlighted that World Cup host metros saw an average 30.3% spike in hospitality hiring in May.“

Note that May’s total was revised down from +172k to +129k but Leisure and Hospitality numbers remained intact at +70k, 54% of the total increase.

Goldman Sachs analysts expected June’s figures to show the competition boosting employment by another 40k jobs.

However, June’s Leisure and Hospitality jobs slumped 61k, the industry’s largest monthly employment decline since the pandemic. Projections simply did not materialize.

Last week [June 21-27] represented the most concentrated portion of the tournament to date, with 27 matches played across the United States from Sunday through Saturday, yet host-market occupancy and demand continued to trend below last year’s levels.

Last week’s U.S. hotel performance reinforced trends that have been evident since the start of the tournament. Since the start of the games, occupancy remained essentially flat. Performance outside of match nights has been markedly different (…) occupancy has declined 3.1 percentage points. Every host market except San Francisco has declined in occupancy on shoulder days. (…)

Despite hosting a combined seven World Cup matches, Atlanta, Seattle and Kansas City experienced a combined 13.5% decline in hotel demand and an 11.2-percentage-point drop in occupancy. (…)

Both [Toronto and Vancouver] markets saw decreasing occupancy. (CoStar News Hotels)

The sucker punch!

Early demand projections were artificially inflated because FIFA blocked out massive chunks of rooms months in advance. However, FIFA quietly exercised opt-out clauses and canceled up to 70% of those room blocks in cities like Boston, Dallas, Los Angeles, Philadelphia, and Seattle just weeks before kickoff. Hotels that had rejected regular summer bookings to save space for FIFA were suddenly left with empty lobbies.

Visas and travel friction played a significant role. Strict US visa policies made it substantially easier for international fans from South America, Europe, and Africa to book trips to Mexico and Canada instead. (Front Office Sports)

Mobility data from the opening matches near SoFi Stadium in Los Angeles showed localized restaurant traffic surging 26% directly ahead of kickoff. Restaurants situated outside the immediate stadium radius or official FIFA Fan Festival zones have reported sharp drops in baseline consumer traffic. Normal local diners are actively staying home to avoid gridlock, and high-spending business travelers have abandoned the cities entirely, creating quiet dining rooms away from the soccer footprint.

• “The Hotel Association of New York City slashed its tournament revenue forecast from $200 million to $100 million and initial results pointed to an outcome even below that revised target.”
• In L.A., “June saw a pull-back in broader California hospitality metrics as employers leveled out their workforce”.
• “In mid-June, Visit Seattle reduced its projected tournament economic impact by 9% (down to $846 million) as 75% of local hotel members reported occupancy rates below expectations, causing hospitality hiring to freeze.”

Also,

The WSJ’s “broadly improved” labor market does not verify with so much of the gain coming from health care and social assistance and 7 of the 14 major sectors losing workers. The diffusion index declined 1.6pt to 54.4 MoM and by 2.4pt to 56.2 on a three-month basis.

The earlier impressions of a turnaround is also deflated by revisions to April and May jobs data: –31k in April and –43k in May.

In May, the 3-m average gain was 188k; in June it’s down to 114k and Goldman now estimates that the underlying pace of job growth based on the payroll and household surveys now stands at 74k.

Employment growth has decelerated from 2.2% YoY in 2024 to 1.2% in 2025 and to 0.2% in H1’26.

Average hourly earnings were up 0.3% MoM for all workers, but only 0.2% for production workers. For the year, AHE for all workers were up 3.5%, below inflation of 4.1%.

The household survey numbers were weaker, declining 507k with both part-time and full-time employment down.

The two BLS surveys typically move roughly together. Let’s hope it’s the red line that will reach back to the blue line.

The household employment series adjusted to match the payroll concept is down 0.7% YoY in June (payrolls +0.3%). It has been negative for 6 consecutive months, from +1.7% in December 2025.

The quarterly bar chart shows that labor income growth has stabilized at 4.0% in the past 12 months. Hopefully, the physical employment part (hours + jobs) will not shrink post World Cup but contribution from wages is now below 3.0%, well below inflation, and declining.

A Sudden Glut of Oil Threatens to Weaken Iran’s Hand in Talks

Oil prices have fallen to prewar levels. Tanker traffic through the Strait of Hormuz is recovering fast. Gulf producers are already restarting idled wells.

But one thing will take much, much longer—refilling the world’s oil coffers.

Speed matters. The amount of oil in storage around the world is playing a central role in the U.S.-Iran power dynamics. The faster countries restock their buffers of crude, the weaker Iran’s ability to threaten the world economy by holding the Strait of Hormuz hostage.

Vice President JD Vance explicitly connected oil storage and negotiating leverage last week. He said in an interview with media personality Michael Knowles that the U.S. signed a memorandum of understanding with Iran to allow the world to “refill some stocks and then to see where the hand is,” referring to Tehran’s position at the table. (…)

Some are predicting oil prices—currently around $70 a barrel—will fall even more in the months ahead, providing further relief to drivers and airlines. Analysts at Macquarie and Citigroup both forecast this past week that prices could sink to $60 in coming months.

Part of the reason for the projected price decline is that it will take time before strategic-reserve managers start buying again.

OECD nations are projected to begin refilling strategic reserves in the fourth quarter of this year, with the U.S. starting its own replenishment in 2027, first at only 100,000 barrels a day and then ramping up to about 170,000 barrels a day in the second half of the year, said JPMorgan’s Kaneva. (…)

Tanker traffic out of the Strait of Hormuz has entered a new normal of around 30 to 60 a day. That is less than before the war, but enough to relieve pressure in global markets.

Ship tracker Vortexa estimated around 140 million barrels of crude oil left in June—an average of about 4.7 million barrels a day, up from just two million a day in May. The crude exodus accelerated in early July to about 40% of prewar levels, according to Vortexa.

On Sunday, the Organization of the Petroleum Exporting Countries and its allies agreed to raise oil output by 188,000 barrels a day in August, the fifth straight monthly increase. With shipping traffic through the strait recovering and Gulf producers restoring production, the cartel’s announcements of output increases are less symbolic than a few months ago.

The United Arab Emirates, which left OPEC in May after years of chafing under production quotas, has been one of the quickest Gulf producers to dial exports back up.

It is using a bypass pipeline from Abu Dhabi to Fujairah outside the strait, plus it is taking advantage of ships streaming out of the Gulf.

Kuwait has recovered production volumes faster than expected. Its export loadings rose to around 1.6 million barrels a day last week, compared with prewar levels of around 2.4 million barrels a day, Johnston said. Saudi Arabia, meanwhile, has kept sending oil via a bypass route to the Red Sea in addition to its tankers now exiting the Gulf. (…)

Replenishing the SPR back to prewar levels will take 15 to 18 months at a rate of 200,000 barrels a day, said Hamad Hussain, commodities economist at London-based research firm Capital Economics. And that is an optimistic rate of buying. (…)

“Washington did not rebuild the SPR after the previous drawdown cycle, and with focus on keeping prices low, it has little incentive to bid aggressively for barrels to refill it now,” said Rahul Choudhary, an oil and gas research analyst at Rystad Energy. (…)

China doesn’t appear to be in a rush to refill. Vortexa data shows that China imported just six million barrels of crude a day via sea in June, roughly four million barrels a day fewer than it averaged in 2025.

Not everyone thinks the current calm will last. The oil price is reacting to the idea that the “hostilities have largely finished for good,” said Neil Crosby of market-intelligence company Sparta Commodities. “I and many doubt that this outcome is real and lasting,” he said. But for now, it is hard to bet the other way until conflict flares up again.

Foreign investors are piling into US equities at a record pace

(…) The YTD pace has already exceeded the full-year total recorded in the middle 50% of years since 2002. Current demand for US equities is unprecedented.

Small Stocks Are Having Their Biggest Run in Decades The Russell 2000 has climbed some 22% in the first six months of the year

(…) It is an often-overlooked bright spot in a market that has for years now been dominated by the trillion-dollar giants of the artificial-intelligence trade. Some investors are taking it as a sign that the recent run-up in stock prices, confined mostly to big chip companies such as Intel or Micron Technology, is at long last spreading to other corners of the market.

“Investors are stepping back and saying ‘Where might the next leg of alpha come from?’” said Joshua Schachter, chief investment officer at Easterly Snow, referring to the trading term for outperforming market benchmarks. “The market was missing a lot of interesting ideas with great valuations.” (…)

The prospect that the Federal Reserve would lower borrowing costs in 2026 helped companies with smaller market-capitalizations notch a strong start to the year, before the war with Iran upended those bets. Now, small-caps are benefiting from a resilient economy that appears set to improve in the coming months. (…)

That shifting outlook is particularly beneficial for small-cap companies, which tend to earn much of their money domestically and—having less access to capital markets than multinationals—often borrow using rates that adjust with those set by the Fed.

Add to that the fact that U.S. companies are earning gobs of money, and the windfall isn’t limited to the likes of Microsoft or Nvidia. Among the biggest gainers in the S&P Small Cap 600 index last quarter: the buy now, pay later platform Sezzle and the restaurant chain Cracker Barrel. (…)

Francis Gannon, a managing director at Royce Investment Partners who primarily invests in small-caps, noted that analysts project 54% earnings growth for companies in the Russell 2000 in 2026. That is more than double the rate expected for the Russell 1000, which tracks large-cap stocks.

“The earnings story for small-caps is pretty strong, and it’s just beginning,” Gannon said. “We have a bit of a run coming.” (…)

Wait, wait! (Goldman Sachs)

• This outperformance represents just a modest reversal following nearly 300 percentage points of Russell 2000 underperformance vs. the S&P 500 during the past 15 years.
• AI infrastructure stocks have contributed roughly 40% of the YTD return for the Russell 2000. Small caps have also benefited relative to the S&P 500 by avoiding the drag from the Magnificent 7, which returned 0% in H1 2026. However, last week’s index reconstitution cut the weight of AI infrastructure stocks in the Russell 2000 from 15% to 7%, including the removal of some of the largest contributors to the index’s YTD return.
• Biotechnology accounts for 11% of Russell 2000 index weight, a 9 percentage point tilt relative to the S&P 500, and has driven roughly 10% of the small-cap index YTD return. One reason for this strength is the ongoing wave of M&A. Announced US healthcare M&A totaled $236 billion in the first half of 2026, a 90% increase relative to H1 2025 and the strongest start to a year since 2021.
• Analysts model 48% EPS growth for the Russell 2000 in 2026. However, analysts have cut Russell 2000 2026 EPS estimates by 9% YTD. Rising valuations have accounted for about half the small-cap return so far in 2026. Analysts are notorious optimists on small caps earnings. Revisions have been negative in 19 of the last 21 years averaging –21%.
• The combination of elevated valuations and near-trend US economic growth points to low single-digit Russell 2000 returns in the next 12 months. The biggest upside risk for small-caps is stronger economic or AI capex growth than the market currently prices. On the downside, in addition to disappointing growth, a hawkish Fed would be a particular challenge for small-caps. Nearly 30% of Russell 2000 stocks are unprofitable and 29% of Russell 2000 debt is floating rate vs 7% for S&P 500 companies.

 

 

 

Ed Yardeni’s Blue Angels valuation chart is in “Buy High” territory, even before considering that nearly 1 in 3 companies are losing money and that forward earnings typically don’t materialize.

If you want to go small, the S&P 600 Index sells at 16x forward EPS, all profitable companies but also subject to analysts’ optimism.

 

AI CORNER

The AI Trade Is Losing One of Its Key Signals

The Silicon Data LLM Token Expenditure Index, which tracks what users pay for AI tokens, is down almost 20% from a high in May after nearly doubling since its inception in December. The gauge is the cleanest read anyone has on the $700 billion-plus capex boom that has done the sector’s heavy lifting.

For stock investors, that could be flashing a warning that AI companies are losing pricing power with increasingly cost-sensitive customers, and that expectations for an eventual AI bonanza could prove misplaced.

“There are increasing reports that users of AI solutions, priced in tokens, are having to restrain unlimited use due to high costs,” said veteran investor Louis Navellier. “The chatter that OpenAI is pushing back its IPO to next year is seen as a sign that, currently, profitability remains a problem.”

Just to clarify, a softer index doesn’t mean AI is getting cheaper. The gauge blends prices and usage, meaning a dip can imply very different scenarios: either list prices are falling, or demand is shifting toward cheaper models. It could also point to a genuine softening in what buyers are prepared to shoulder.

Each of these possibilities carries different implications. Silicon Data, which built the index, has warned people to stop reading it as a price tag. The firm calls it a proxy for marginal willingness to pay.

Let’s start with a benign read: While token prices have collapsed more than 90% since 2023, total spend has roughly doubled since last year. Cheaper tokens have expanded the market. This means that an index pause is simply digestion, while demand is real and capex is money well spent. The bull case for Nvidia Corp., memory makers and data-center names rests here.

Bears warn that sustained weakness in the index could end the trade that saw nearly the entire AI cohort rally hard this cycle.

It’s token spending that justifies the next capex order, and the bill is already looking stretched. Allianz Research said there’s nearly a 46% growth gap between AI investment and sales. That’s worse than the 32% divergence measured during the 2001 telecom bust. (…)

There’s also a more recent, demand-side reason why the bearish read may have legs. Washington has a newfound willingness to exert control over a pivotal industry. The US government only this week removed foreign access restrictions on Anthropic PBC’s Fable 5 model, days after regulators requested OpenAI to stagger the roll-out of an upcoming release.

Meanwhile, the European Union’s AI Act targets frontier models for mandatory evaluations and stringent transparency requirements. None of these caps prices directly, but does create a deployment-and-compliance burden on top platforms that lesser-but-still-useful systems don’t carry. This consideration may just offer companies’ financial chiefs a rational reason to route workloads to cheaper models.

To be sure, this isn’t a chip-glut call. Top-end graphics processing units and high bandwidth memory are sold out through 2026, with no real relief arriving until 2028. The hardware tell is more subtle in that it points to a demand-mix shifting away from top-end training GPUs, toward inference-optimized parts. This changes the mix of winners, but doesn’t hand you a short. (…)

The conclusion is that the token chart cuts both ways, and one should hold both reads at once. If the late-June flattening holds and the dip was just mix-shift digestion, cheaper tokens will keep on expanding the markets, meaning capex spending stays justified, leaving the bull case intact.

If instead this is the point where customers’ willingness to pay peaks just as regulatory headwinds nudge demand down-market, then the most expensive part of the trade is also the first to crack. That’s because it’s a pricing-power story, not a silicon story, that is funding the march toward $1 trillion of capex in 2027.

Compute demand is not slowing…

Google has placed limits on Meta Platforms Inc.’s use of its Gemini artificial intelligence models because it could not provide as much computing capacity as the social media company wanted, according to the Financial Times.

In the latest sign of AI infrastructure constraints, Alphabet Inc.’s search giant has enforced restrictions on several clients, with Meta particularly affected, the FT reported on Sunday. The move has had a knock-on effect on Meta’s internal projects, and meant the company has told staff to make more efficient use of AI tokens, the newspaper reported, citing three unidentified people familiar with the matter. (Bloomberg)

If Google had the capacity they would sell it to Meta but they don’t.

… but demand may be shifting:

• How China overtook the US in AI token usage—and why it matters

Chinese models are gaining global ground.

Data from OpenRouter shows that from March 30 to April 5, global token usage reached 27 trillion, with Chinese models accounting for 12.96 trillion, up 31.48% week-on-week. US models, meanwhile, generated 3.03 trillion. This marks the fifth consecutive week that Chinese models have surpassed their US counterparts.

Given that OpenRouter aggregates usage from over five million developers across more than 400 models worldwide, this isn’t a local phenomenon. Chinese models are competing, perhaps even winning, on the global stage.

The driving force behind comes down to that most basic of economic factors: price. Chinese models are highly cost-competitive. DeepSeek’s V3.2, for example, charges just $0.42 per million output tokens, compared with $75 for Anthropic’s Claude Opus. That’s a cost of more than 170 times less.

Market sensitivity to pricing applies equally to AI model usage. OpenRouter COO Chris Clark noted that Chinese models are widely used because they are “disproportionately represented in agent workflows run by US companies.” In short, this isn’t just about technology leadership. It’s about cost-performance—and that’s where Chinese models are carving out an edge.

China’s surge to 140 trillion daily tokens—and its overtaking of the United States in usage—signals a shift in how AI leadership is defined: not solely by model sophistication, but by the ability to scale, deploy, and embed AI into real-world workflows at speed.

This lead has been driven by cost efficiency, developer-led adoption, and breakthrough application scenarios such as AI-generated video, alongside a quieter but more strategic wave of enterprise deployment.

Yet the significance of this moment lies beyond the numbers. Token consumption is an imperfect proxy—one that captures intensity of use, but not necessarily quality or impact. As tokens increasingly become a metric of progress, the real competitive advantage will belong not to those who consume the most, but to those who translate usage into durable capabilities, productivity gains, and defensible value.

This recent “AI Intelligence Index” shows the evolution of various models since 2022:

This summarizes the above:

This FT chart shows the trends:

Here’s a cost comparison:

Summarized:

A good case in point: Microsoft!

Microsoft is eyeing DeepSeek as a hyper-cheap, optional alternative to the expensive OpenAI and Anthropic models currently powering its enterprise agent tool, Copilot Cowork. To cut its own soaring internal AI token usage costs, Microsoft issued a firm June 30, 2026 cutoff deadline for its Experiences + Devices (E+D) division—the engineering teams behind Windows, Office, Teams, and Surface—to stop using Anthropic’s Claude Code. (Bloomberg)

Another sucker punch:

Best-in-class SemiAnalysis debunks stories about widespread cancellations of datacenter expansion:

The claim that half of 2026 US datacenter capacity will be delayed or canceled has been circulating widely across financial and social media. (…)

We update the dataset by reviewing every site dozens of times a year. However, over the last 6 months, our YE2026 NA Hyperscaler Self-build forecast only moved by ~1%, and NA colocation <5%. What is causing the discrepancy?

In our view, the culprit is obvious: the data sources behind these claims of “50% of 2026 datacenters are delayed” are essentially uninformed vibe-coded datacenter forecasts that take announcements at face value, without any bit of critical judgement. (…)

Thankfully, that’s not how we built our model, which is trusted for billion-dollar investment decisions by all the world’s largest tech companies in the world, as well as energy and industrials giants, and all the largest investors on Wall Street. (…)

Let’s be clear, delays and cancellations are occurring. (…) The question worth answering is why our 2026 outlook holds despite the headlines.

The short version is that the cancellations and delays cluster in a layer of the pipeline we already treat as structurally oversupplied. The veritable 2026 projects, meaning the ones with site control, equipment on order, interconnection agreements signed, and/or vertical construction underway, continue to progress on schedule. Hyperscalers, whether its self-build or colocation tenancies, are routing around the constraints in ways that announcement-stage tracking does not pick up. (…)

The Equipment Supplier Narrative Is Wrong Too. The fear priced into names like Vertiv is that the wave of datacenter delays and cancellations will start eating into record orders and backlog. However, our Datacenter Model capacity forecast shows quarter by quarter delivery, and it’s clear to us that delivered MWs are accelerating.

Equipment lead times remain long. (…) the way capacity actually gets allocated is through prepayments. Buyers wire money up front to hold a slot in the queue, which boxes out players with less capital. (…)

Now run the cancellation fear against that backdrop. The projects getting canceled sit in the early stage layer that never placed equipment orders in the first place; a speculative announcement dying in a county commission hearing removes zero orders from anyone’s books.

The projects that do sit in OEM backlogs have prepaid for queue position, locked in long-lead SKUs during scoping, and in many cases already broken ground. And in the rare case a real order does fall away, a queue running three to four years deep means the slot is reallocated to the next buyer in line rather than vanishing. (…)

When the noise is filtered out, our 2026 US datacenter outlook is largely intact. 24GW is expected to come online this year, which are already under construction and tracking. The cancellations and delays dominating financial media headlines are concentrated in the early-stage announcement layer that was always going to be culled. (…)

In a recent update:

Today, the US grid is serving most datacenter load in the US, but we’re reaching a tipping point. As the insatiable demand for power of AI Labs and hyperscalers keeps accelerating, the grid simply can’t add capacity fast enough. That leaves Behind-The-Meter as the only way for the largest players to secure the power they need. (…)

Let’s start with key numbers: first, we continue to see a record datacenter buildout in the US, going from +21GW in 2026 to +84GW by 2030.

Our research suggests that BTM will power well over half of new US datacenters in 2028+, and the Total Addressable Market (TAM) for DC BTM equipment to cross 50GW/year by 2029. New Grid Capacity isn’t growing fast enough, and also needs to serve non-datacenter load growth. (…)

As such, we expect power generation to be a major bottleneck to grid-connected datacenter load growth. (…)

These generation and transmission constraints, combined with inadequate market incentives, makes Behind-The-Meter often the most attractive solution for GW-scale newbuilds. And we’re already observing that many top-tier developers are planning 5GW+ behind-the-meter facilities in Texas, where permitting onsite gas is easier. (…)

This is, of course, a material tailwind for BTM solutions and providers. However, key beneficiaries aren’t the usual suspects. (…)

Our analysis of 40k generators reveals one of the core challenges for the US grid: the lack of firm, dispatchable capacity added to the market over the next two years. Our nameplate capacity forecast shows that the US industry will add less than 10GW of gas per year in 2026 and 2027, with additions only picking up in 2028 and beyond. (…)

The 2026–27 shortfall isn’t the product of one bottleneck but a stack of them. (…)

We see solar and BESS [Battery Energy Storage System] each adding over 20GW nameplate per year in coming years. However, from a grid perspective, their value is significantly lower (…) on an ELCC basis (Effective Load Carrying Capability), their contribution is minimal. (…)

Our point is not that renewables don’t matter, they are, and will remain, an important part of the AI buildout, just not its most important one. It is that nameplate badly overstates what they add to firm capacity: on an ELCC basis, each incremental GW of solar, wind, and storage is accredited at a steep and widening discount as the risk it addresses gets saturated. That gap between nameplate and accredited capacity is exactly what determines how much new load a market can actually host — which is the subject of the next section: grid headroom.

Grid headroom is the capacity a market has left to absorb new load once it has covered its own peak demand and required reserve margin. (…)

Headroom goes “red” when a market’s reserve margin falls below its required target: at that point there is no spare accredited capacity to host an incremental large load like a datacenter without eroding reliability. Across a growing set of subregions, our analysis shows that threshold being crossed by 2027. (…)

With grid supply structurally constrained — too little firm capacity, too little effective capacity from renewables and storage, and vanishing headroom — the decision now sits with the buyer. And we think BTM is now the most attractive option. (…)

The key advantage of BTM vs Grid is speed and certainty on the timeline of power. On speed, onsite generation can be energized in a fraction of the grid-interconnection timeline — requested BTM in-service dates cluster around 2027–28, against grid timelines that routinely slip toward 2030. On certainty, the schedule sits in the buyer’s hands rather than the utilities.

(…) timelines provided by utilities are notoriously unreliable and they often push back or revise down the load that they promised datacenter operators… with little to no penalties at all.

This doesn’t work for AI Labs, for which access to large scale compute is the lifeblood of their business. They need power both to generate revenue (inference) and to fuel future revenue growth (training).

In addition power as a percent of total TCO is mostly insignificant, meaning that any amount of power secured by an AI Lab is actually worth billions.

For example, our Tokenomics Model susbcribers know very well the margins Anthropic is currently making on API, and the implied tens of billions of dollars of annual revenue per GW. Given the cost of GW-scale DCs and revenue potential, it simply doesn’t work if it risks multi-year delay, or if the load-serving entity faces no symmetric dollar penalties for being delayed.

In the grid vs BTM debate, another key factor is redundancy and uptime, historically a major advantage provided by the grid. (…)

However, AI labs and some hyperscalers have relaxed those requirements as there is now a lower uptime tolerance applied to both inference and training, not just training. (…)

This removes historical cost barriers to BTM adoption. (…) Providing four or five nines of redundancy at a BTM site is a recipe for unbearable costs. But now that customers are willing to accept lower redundancy, the economics of grid vs BTM are much more balanced. (…)

Now, none of this means the grid is being written off and that the uptime and reliability requirements of a datacenter are no longer of importance. Rather, it reflects our view of how market participants are adapting to meet the rampant power needs of AI DCs – and it’s our view that tenants are becoming more lenient with these requirements as they have nowhere to go. (…)

Given the context of BTM vs grid, it is crucial to understand the positioning of different OEMs. A month ago, our Energy Model report Grid Bad, BTM Good flagged GEV and Siemens Energy and MHI as key losers of the trend. The explanation is their portfolio positioning: as our Energy Model shows, they’re both highly exposed to the grid-connected buildout. Our grid nameplate gas forecast, shown at the beginning of the article, is largely carried by GEV, MHI and Siemens (exact data in our Energy Model). While their BTM exposure is also high, it’s not as high as that of others like Bloom.

As explained above, with BTM becoming more favorable for buyers and expected timelines being 2028, we don’t expect utility turbine orders for 2030+ capacity to go up. We see 2026 as a potential peak for turbine orders for the big 3 OEMs. Most buyers will be focused on 2028, and that’s going to flow to Bloom, Innio, Wartsila, Bergen and the likes.

(…) the surge in “contracted load” has driven massive orders, but we now think there is growing skepticism on utilities’ ability to serve this on time. Combined with the financing challenges discussed earlier, we see a good recipe for peak turbine orders in 2026.

This is also overall negative for IPPs like CEG/Vistra/TLN, exposed to grid constraints and rising power prices. As demand for grid eases (on a relative basis) and BTM surges, IPPs are negatively exposed. However, we see some interesting ERCOT plays on the IPP side. (…)

We view BYOG [Bring-Your-Own-Gas] grid-structure solutions as a growing space to watch (…).

In our view, NRG is well positioned to benefit from ERCOT’s new BYOG/WLPUN framework: it has gas turbines available to pair with co-located load, making it a natural supplier of the on-site generation these structures are built around.

On its Q4 FY25 call, management pointed to a contracted large-load opportunity that implies roughly $2.5bn of incremental EBITDA — built on blocks in excess of 1 GW under 10- to 20-year contracts with investment-grade counterparties — with first power potentially online by late 2029, which implies that some deals would need to close in 2026 to hold that timeline.

Against the backdrop of the recent 20-year Microsoft–Chevron agreement in West Texas (the ~2.67 GW Project Kilby), we see no reason NRG could not land a comparable long-dated, hyperscaler-anchored gas deal of its own. Management nonetheless continues to frame front-of-the-meter generation as its primary near-term focus which may be exemplified with the recent ERCOT PUN/BYOG rulings. (…)

• The future of AI may need less electricity than we think Tomorrow’s data centers may look very different from today’s as innovators seek to offset AI electricity demand

(…) The newest AI campuses require hundreds of megawatts of power, and gigawatt-scale facilities are already moving from concept to construction. Utilities across the United States report growing interconnection queues, while technology companies are competing for scarce transmission capacity and generation resources.

There are many examples of technologies that initially appeared destined to overwhelm available resources before innovation changed the equation. (…)

A growing number of innovators are questioning assumptions that have long shaped the design of data centers. This report highlights three examples. They are not the only ideas under development, nor will they necessarily succeed.

Assumption 1: Electricity must flow through copper.

(…) the economics of data centers may change the calculation, making exotic materials more competitive.

A class of materials called high-temperature superconductors (HTS) eliminates virtually all electrical resistance when cooled with liquid nitrogen, allowing substantially more electricity to move through much smaller cables. VEIR, a venture-backed company headquartered in Woburn, Massachusetts says its superconducting cables can carry roughly 10 times the power density of conventional conductors while dramatically reducing electrical losses. (…)

At the scale of today’s hyperscale AI campuses, operators may be willing to devote substantial energy to cryogenic cooling if it enables dramatically larger gains in computing performance and overall energy efficiency.

One company pursuing that vision is Snowcap Compute, a Silicon Valley start-up developing superconducting processors for AI and other high-performance computing applications. Superconducting circuits have essentially no electrical resistance under cryogenic conditions, allowing them to operate with far less energy than traditional silicon-based processors. (…)

Assumption 2: Data must move electrically.

(…) as AI systems scale to unprecedented size, approaches that were once too expensive may now make business sense.

As AI clusters grow larger, moving data is becoming almost as important as processing it. Every time GPUs exchange information, electrical signals travel through copper connections that consume power, generate heat and eventually require amplification. At the scale of modern AI superclusters, that communication overhead has become one of the fastest-growing sources of electricity demand inside the data center.

The challenge is becoming increasingly important because interconnect technology is improving far more slowly than AI computing power. Lightmatter, one of the leading companies developing photonic networking technology for next-generation AI systems, notes that while frontier AI models have expanded roughly 240-fold in only three years, and AI clusters have grown tenfold, interconnect bandwidth has improved only about twofold — creating an increasingly severe bottleneck.

Photonics transmits information as pulses of light through optical waveguides (e.g., fibers) instead of electrical signals through copper. Because light can travel farther with much lower loss, using photonic interconnects could reduce both power consumption and heat generation, while allowing larger AI clusters to function as a single computing system.

Lower networking power also reduces cooling requirements. Every watt that is not dissipated by high-speed electrical interconnects is a watt that does not need to be removed by cooling systems, allowing photonics to reduce both direct electricity consumption and the secondary energy required to keep AI hardware within operating temperatures. (…)

Lightmatter has developed photonic interconnects and optical “engines” designed to connect AI chips with substantially lower power consumption than conventional electrical links. Rather than replacing GPUs, the company’s technology focuses on reducing the energy required to move data among them — an increasingly important challenge as AI systems scale into tens of thousands of accelerators.

Other companies are pursuing similar goals using different architectures. Ayar Labs, a Silicon Valley start-up, has developed optical input/output (I/O) “chiplets” that replace short copper connections with fiber-optic links. The company states that its technology improves data bandwidth per watt substantially. Ayar has attracted strategic investments from Nvidia, Advanced Micro Devices and Intel, reflecting growing industry confidence that optical interconnects will become a core component of future AI systems. Another company, Celestial AI, is developing what it calls a “photonic fabric” that uses light to connect processors and memory more efficiently, while reducing the energy required to move data across increasingly large AI clusters. (…)

Assumption 3: AI must run in giant data centers.

(…) As everyday devices employ increasingly powerful AI processors, some companies are pursuing ways to get hundreds of millions of existing devices to shoulder more of the work.

(…) The AI work most users actually see — answering questions, generating text, recognizing images or translating languages — is called “inference.” Every inference task performed locally means less electricity consumed inside a remote AI campus.

Edge AI is a concept that takes advantage of the “latent compute” of billions of processors that already exist but remain underutilized much of the time. Rather than transmitting every request to a hyperscale facility, local devices — such as increasingly powerful smartphones, PCs, vehicles and industrial equipment — perform AI tasks and communicate with cloud systems only intermittently. (…)

Companies pursuing this strategy include Qualcomm, whose AI processors enable on-device inference in mobile and industrial applications; Nvidia, whose edge computing platforms target factories, robotics and autonomous systems; and Apple, which increasingly performs AI workloads directly on iPhones and Macs. Each of these approaches reduces network traffic by shifting computation away from centralized facilities. (…)

None of these technologies has yet demonstrated commercial success. Superconductors remain expensive. Photonics must compete against decades of investment in conventional interconnects. Edge AI raises difficult questions about software deployment, security and model management.

Meanwhile, someone has a plan…

Trump administration moves to gut energy efficiency rules for home appliances

As punishing heat sends the U.S. power grid to the brink and Americans reel from rising electricity costs, the Trump administration is moving to weaken efficiency standards for home appliances proven to cut power demand and lower utility bills.

The Energy Department on Thursday said the proposal would “Permanently End Green New Scam Appliance Mandates.” It takes aim at energy conservation standards for a host of appliances, including air conditioners, refrigerators and washing machines.

For decades, Congress has required those standards to be continually updated, creating pressure on manufacturers to make each generation of new appliances more efficient.

The Trump administration plan would prohibit further updates in many cases, as part of an effort to “preserve consumer choice and lower costs.” (…)

“In America, you should be able to choose a dryer that dries clothes on the first try rather than one that takes multiple cycles — unfortunately, past administrations thought otherwise,” Energy Secretary Chris Wright said in a statement.

“This proposed rule will preserve the American people’s ability to choose home appliances and equipment that actually work — at prices they can afford.” (…)

Trade groups argued that manufacturers had already made considerable investments in more efficient products and that rolling back the rules would put them at a competitive disadvantage.

The next round of federal efficiency updates, which would take effect between 2029 and 2035, have the potential to reduce peak summer electricity demand by 34 gigawatts in 2040 — equivalent to 34 large nuclear reactors or the power used by tens of millions of homes, according to an analysis by the Appliance Standards Awareness Project, a coalition of consumer groups, utilities, state regulators and environmental advocates that support efficiency.

That could ease strain on the power grid, as electricity companies struggle to keep the lights on while also meeting the explosive energy demands of the AI industry.

The Trump administration has already declared multiple power emergencies because of spikes in power demand on the grid this year. The declarations include orders for utilities to spend hundreds of millions of dollars to keep operating coal plants that had been slated for retirement. (…)

“This doesn’t make any sense,” said Andrew deLaski, executive director of the Appliance Standards Awareness Project. “They want to handcuff a policy that works well to save energy in our homes and businesses and frees up electricity capacity to meet the growing needs of our economy. It also saves people money.”

DeLaski said he was confused by Wright’s claims around dysfunctional clothes dryers because the existing rules do not allow products on the marketplace if they do not function properly. Appliances typically have settings that allow consumers to run them in more energy-intensive ways if they prefer, he said. (…)

• US clean power prices set to soar as AI demand coincides with subsidy cuts

US companies are facing a sharp increase in green energy costs as the Trump administration halts renewable tax credits at the same time as cash-rich data centres are snapping up the available supply.

According to a survey of solar and wind developers across the US by LevelTen Energy, a clean energy marketplace, the cost of clean energy power purchase agreements (PPAs) is set to increase by 40 to 120 per cent once Inflation Reduction Act subsidies for solar and wind projects come to a close.

For instance in Texas, PPA prices could jump from $55 per megawatt-hour to $121. (…)

The tax break had helped turbocharge the growth of US renewables projects. According to Wood Mackenzie, US solar capacity almost doubled between 2022 and 2025, jumping from 141 gigawatts to 279 gigawatts. But that growth is now predicted to slow markedly, with the cost of clean energy PPAs expected to soar as developers gain the upper hand in negotiations. (…)

The price squeeze comes against a backdrop of rapidly rising energy demand in the US, as the tech giants engage in an AI arms race and as household goods and transport are increasingly electrified. (…)

The Energy Information Administration forecasts that US electricity demand will grow between 25 and 50 per cent by 2050. Large energy users like factories, retailers and data centres buy clean energy PPAs to help meet their sustainability targets, by matching their grid-powered or on-site electricity use — which could be fossil-fuel generated — with renewable power generated elsewhere.

By guaranteeing a buyer for the energy, PPAs give developers the certainty they need to finance and build new projects.

The Trump-led backlash against renewables and the boom in US power demand has made companies less concerned about environmental factors than in previous years. (…)

“Demand [for clean PPAs] is going to continue,” said Don Leavens, chief economist for the National Electrical Manufacturers Association. “Global corporations have to respect what’s happening globally, not just what’s here in the US.”

Clean energy purchases also help buyers hedge, by giving them a fixed price for electricity so they are less exposed when wholesale prices rise due to fuel costs or demand spikes.

Supply chain issues, labour costs, interest rates and long waiting times to connect to the grid are also pushing up the cost of renewable energy agreements.

According to data from the American Society of Civil Engineers, the cost of transformers has risen 60 to 80 per cent since the start of 2020, while solar labour costs increased 15 per cent in 2025. (…)

Some industrial buyers say that as prices shoot up, data centres, which are desperate for as much power as possible and willing to pay sky-high rates for it, are pushing them out of the market for clean energy contracts. (…)

US MANUFACTURING PMIs

S&P Global: US manufacturing growth expands solidly despite renewed fall in employment and waning optimism

The headline seasonally adjusted S&P Global US Manufacturing Purchasing Managers’ Index™ (PMI®) registered 53.9 in June, down from 55.1 in May. The latest reading marked the eleventh consecutive month above the crucial 50.0 threshold and signaled a solid improvement in operating conditions.

The PMI was supported by further marked and above trend gains in both output and new orders. That said, growth moderated slightly across both variables, with new orders rising at the softest pace since March. According to panel members, new product launches and pre-orders placed in anticipation of further price rises supported demand.

Where firms reported higher sales, the data pointed to a continued reliance on the domestic market, as new export orders faltered again. Exports declined for the twelfth consecutive month, albeit modestly overall. Tariffs reportedly remained the main factor behind the fall in foreign sales, while some companies also noted subdued international demand, in part linked to the war in the Middle East.

Despite sustained demand growth, US manufacturers refrained from additional hiring in the latest survey period. In fact, employment fell for the second time in three months and at the steepest rate since May 2020. Lower headcounts, combined with rising order inflows, placed further strain on capacity, as backlogs of work rose modestly and at a faster pace than in May.

Inventories of finished goods increased only fractionally in June. Survey evidence suggested that firms, where possible, used existing stocks to complete orders, limiting the extent of replenishment.

US goods producers recorded another marked deterioration in vendor performance during June, with reports of shipping delays and port congestion disrupting supply chains. In response, companies sought to protect input availability by building stocks of purchases, which rose at the strongest rate since May 2025. Purchasing activity also increased at an accelerated pace, with growth the strongest in over four years.

Tariffs and higher raw material prices remained key drivers of input cost inflation in June. The latest data signaled another historically elevated rise in costs, though the rate eased from May’s recent peak. Selling prices were also raised solidly, but factory gate inflation followed the trend in cost burdens and softened to a three-month low.

Finally, firms remained positive about output prospects for the coming year at the end of the second quarter, but the overall degree of optimism fell to an eight-month low. Positive sentiment was linked to hopes that easing inflationary pressures and geopolitical tensions would help stimulate sales and production. However, concerns over the health of the domestic economy continued to weigh on confidence.

ISM:

The Manufacturing PMI^® registered 53.3 percent in June, 0.7 percentage point lower than in May. The overall economy continued in expansion for the 20th month in a row. (A Manufacturing PMI^® above 47.5 percent, over a period of time, generally indicates an expansion of the overall economy.)

The New Orders Index expanded for the sixth consecutive month after four straight readings in contraction, registering 56 percent, down 0.8 percentage point compared to May’s figure of 56.8 percent.

The New Export Orders Index returned to contraction territory with a reading of 48.5 percent, 2.1 percentage points lower than the 50.6 percent registered in May.

The Prices Index remained in expansion (or ‘increasing’ territory), registering 73 percent, a 9.1-percentage point decrease from May’s reading of 82.1 percent.

The Employment Index registered 49.7 percent, up 1.1 percentage points from May’s figure of 48.6 percent,” says Spence.

In June, 34 percent of the comments were positive and 66 percent negative, with a 1-to-1.9 ratio of positive to negative sentiment. Among negative comments, the Iran war was mentioned in 31 percent and tariffs in 17 percent; 50 percent of the panelists mentioned pricing volatility as an issue for their companies.

• “The conflict in Iran has impacted pricing in every category of raw materials. Especially, items that have a heavy concentration of oil in the components like our adhesives.” [Chemical Products]
• “Continued pressure from conflict in Middle East is resulting in a more conservative approach to capital expenditures.” [Computer & Electronic Products]
• “While overall economic growth remains resilient, it is slowing as consumer spending weakens under pressure from higher costs for energy and essential goods, reducing demand visibility and increasing cost sensitivity for buyers. Meanwhile, supply chains have stabilized compared to prior years but remain structurally complex, with trade policy volatility, geopolitical tensions and regulatory changes now ongoing cost drivers rather than temporary disruptions. Our organization continues balancing cost control with resilience, shifting sourcing strategies, tightening inventories and prioritizing supplier diversification and risk management.” [Computer & Electronic Products]
• “Retail electronics sales seem to have stabilized to some extent. The pause in tariff changes has been welcomed the last two months, but it’s only a matter of time before more confusion is introduced.” [Electrical Equipment, Appliances & Components]
• “Input costs remain elevated across key categories, driven largely by Middle East conflict impacts and ongoing tariff uncertainty. Supplier lead times have stretched, which is influencing our inventory strategy and sourcing decisions. We are managing exposure through diversified supplier bases and contract structures that balance cost certainty with operational flexibility.” [Food, Beverage & Tobacco Products]
• “Conditions are optimistic but not yet booming for our company, even though many others, it seems, are experiencing growth. Machinery in support of defense and semiconductor manufacturing is very strong, a bright spot for our team. Industrial and medical clients are slow to purchase, focusing more on refurbished and upgraded units versus new ones.” [Machinery]
• “Core business remains solid in the face of ongoing geopolitical uncertainty. AI industry continues to have huge capacity consumption for critical electronics. Monitoring impact of U.S. defense industry needs on supplier capacity.” [Miscellaneous Manufacturing]
• “No major changes from last month. With the potential ending of the Iran war, management is expecting us to go back to February pricing structures and plans since the increase in oil prices was driven by the war and not regular market influences.” [Petroleum & Coal Products]
• “Requests from suppliers in Europe and India for ‘energy surcharges’ have stopped this past month. We’re seeing continued capacity growth in the Asia-Pacific region (excluding China), including Vietnam, Thailand and South Korea. Most suppliers are building for the longer term as geopolitical protection from all sides.” [Transportation Equipment]
• “The new Section 232 tariffs continue to destroy our profitability and demand as we have to raise prices to deal with this gigantic tax. Add the ‘incentives’ for our company to pivot to purchasing non-U.S. sourced material, and one realizes the total ineptitude of this tariff policy.” [Transportation Equipment]

• Retailers Rush In Holiday Imports Early to Avoid Looming Costs New tariffs and fuel surcharges spark competition for space on containerships

Ocean shipping rates are surging as U.S. retailers rush in clothes, electronics and holiday items to get ahead of rising costs caused by tariffs and the Iran war.

The average cost to ship a 40-foot container from China to the U.S. West Coast hit $5,933 on Friday, a threefold increase from the end of February and the highest rate since September 2024, according to transportation intelligence company Xeneta. Peter Sand, Xeneta’s chief analyst, said rates could rise by another 30% before they peak.

Logistics industry specialists say importers are bringing forward orders for the busy end-of-year shopping seasons, including Halloween and Christmas, to avoid new tariffs taking effect at the end of July. Some companies have also been scrambling to beat a new round of bunker fuel surcharges that ocean carriers are expected to implement in the coming weeks to account for rising oil prices resulting from the war. (…)

Importers say competition for space on ships is fierce. “We’ve had more delays getting container space in the last three weeks than we’ve had in a long time,” said Bronwen Sainsbury, president of Stack Resources, a Seattle-based wholesaler of home decor. (…)

Freight specialists say some importers are rushing in goods now to take advantage of a brief lowering of tariffs after the Supreme Court struck down most of the levies in February. The administration introduced lower, temporary tariffs in recent months, and is expected to institute higher tariffs, using a different legal mechanism, at the end of July. (…)

(last data point is April)

• Global manufacturing upturn continues in June, but shows signs of easing

(…) There were some signs that the upturn had passed its peak. Growth of both output and new orders eased, employment decreased and business optimism fell for the first time in three months. Incoming new business rose for the sixth month in a row, but at the slowest pace since March, with a contraction in international trade volumes acting as a drag on overall intakes of new work.

Some firms noted that the recent boost from clients strategic stockpiling (to guard against supply disruptions) had started to wane. (…)

 

CONSUMER WATCH

US Car Payments Hit a Record $777 a Month as Down Payments Drop

The automotive affordability crisis is deepening as average monthly car payments reached a record $777 in the second quarter, with nearly a quarter of US car buyers taking out loans of seven years or more.

The average amount financed on a new car also reached an all-time high of $44,156 in the period, according to data released Wednesday by automotive shopping site Edmunds.com. At the same time, stretched buyers have less money to put down on a new car, as the average down payment fell 10% from a year ago to $5,815, Edmunds said.

It’s the latest sign that there’s no letup in accelerating automotive inflation as the average price of a new car in America has stubbornly remained around $50,000. To afford what was once luxury prices, more than a fifth of new car buyers signed up for monthly car payments of $1,000 or more in the second quarter, according to Edmunds. And that’s led many to owe more on their car than it’s worth and others to default on their loans. (…)

“When you see loan terms extending to record lengths, down payments shrinking and monthly payments hitting all-time highs, you’re looking at a clear recipe for long-term financial strain.” (…)

“Car shoppers are caught in a dangerous practice of focusing heavily on their monthly payment while ignoring the potential long-term damage to their wallets,” said Ivan Drury, Edmunds’ director of insights. A seven-year loan at the current 7% interest rate means “you’re signing up to hand over nearly $10,000 on average in interest alone.”

Warsh Says Inflation Outlook Has Improved but Won’t Say if Fed Should Hike Rates

Federal Reserve Chairman Kevin Warsh declined to say Wednesday whether the central bank needed to consider a rate increase later this month but said his first weeks in the job have seen risks of higher inflation recede. That was evidence, he said, that markets have already grasped his hard line on prices.

“Expectations of future inflation [over the last four weeks] have come down. Inflation risks have come down,” Warsh said at a conference in Portugal alongside foreign counterparts. Anyone expecting the Fed would tolerate inflation running above its 2% goal “would be disappointed,” he added. (…)

Warsh brushed aside complaints Wednesday that investors need a clearer sense of how the Fed would adjust as the outlook changes. He pointed to falling interest-rate volatility, lower Treasury yields and expectations that inflation would decline over the coming year or two as early evidence his strategy is succeeding.

“I hear this as if people don’t understand,” he said. “I think they actually understand quite well.”

The outlook for inflation has improved in recent weeks in part because a deal meant to end the war in Iran has pushed energy prices down. (…)

Earlier Wednesday, Kevin Hassett, a senior White House economic adviser, said in an interview on Fox Business that it would be “a macroeconomic mistake” for the Fed to raise rates. Hassett said he believed Warsh shared his view that AI-fueled economic growth wouldn’t be inflationary and wouldn’t require higher interest rates.

Warsh wasn’t asked about Hassett’s remarks. But pressed on whether the AI boom was stoking inflation, he offered an answer that pointed the same direction. Warsh emphasized the longer horizon, when AI could expand the economy’s capacity to produce goods and services and ease price pressures. “That has huge implications” for the Fed if businesses expand that capacity, he said. (…)

In the Fox Business interview, Hassett delivered an implicit warning about the coming decision. Pointing to Jerome Powell, who stayed on as a Fed governor after his term as chair ended in May, Hassett said he was concerned that “a majority of people over at the Fed” might vote for higher rates “not necessarily…because they’re patriotic, but rather because they want to get Trump.”

Hassett exempted Warsh from that critique, portraying him as saddled with an unruly committee. “It’s not going to be Kevin Warsh’s fault, but he’s got to herd some cats over there, and it’s a really difficult job,” Hassett said.

The [9 on 18] officials who have argued for higher rates have framed their case in economic terms—that a resilient economy and sticky inflation, not hostility to the president, warrant tighter policy.

Some see an economy at risk of overheating, with demand outrunning what the Fed’s current rate setting can restrain. Others think policy was never as restrictive as assumed and that, by standing still as inflation has risen this year, the Fed has effectively been easing.

John Authers:

(…) But while he’s right that the bond market is taking him seriously and dialing down inflation expectations, that’s not necessarily great news. The yield curve — the amount by which yields on longer-dated Treasuries exceed shorter-dated bonds — is reducing, in what is known on markets as a “bear flattening.”

To explain this, Mike Howell of Global Liquidity Indexes in London argues as follows:

The US yield curve is flattening, but not in a benign way. In liquidity terms, a flatter curve is consistent with tighter conditions. Less liquidity raises systemic risks and so increases the demand for longer-dated ‘safe’ government debt. The problem is that this implies a bearish flattening… It points to a market increasingly focused on ‘higher-for-longer’ policy risk rather than on a straightforward growth and inflation slowdown.

The rise in two-year yields since inflation fears were revived by the Iran war has been much sharper than in 10-year yields: (…)

Labor Market Tightness Index

The NY Fed’s HPW index is consistent with a pace of ECI annual wage growth of 2.96 percent.

When PCE inflation is 4.1%…

About AI vs labor:

Layoffs are falling. But where companies are cutting, AI is driving a greater share of the downsizing, Axios’ Ben Berkowitz and Avery Lotz write from a Challenger, Gray & Christmas report out today.

Just this year, AI has been blamed for nearly 102,000 job cuts.

• Since 2023, when AI was first tracked as a distinct reason, it’s been cited in more than 173,000 job cut announcements, per the report.

“AI is the dominant force as companies are restructuring around it, automating roles, and reallocating budgets toward new capabilities,” the firm’s Andy Challenger wrote.

Data: Challenger, Gray & Christmas. Chart: Ben Berkowitz/Axios

U.S. Plunges Trade Pact With Canada and Mexico Into Doubt Washington declines to renew deal by Wednesday deadline, setting it up for annual reviews

The U.S. declined to extend its signature trade pact with Mexico and Canada on Wednesday, setting up a decadelong review process that casts uncertainty over businesses that move goods across the world’s busiest export borders.

The U.S. decision came as no surprise. President Trump has effectively ripped up parts of the U.S.-Mexico-Canada trade agreement that he signed in his first administration, imposing tariffs on a range of goods. He has mused about terminating the agreement altogether. (…)

The pact remains in effect, but now the U.S. refusal to renew means American trade representatives will have to meet every year for a decade with Mexican and Canadian officials to continually review the deal. Negotiations can continue in the meantime. (…)

The pact underpins nearly $2 trillion in annual trade. Last year, combined U.S. exports to Mexico and Canada surpassed $670 billion, according to U.S. statistics. In comparison, U.S. exports to China last year amounted to about $106 billion. (…)

Trump Made $1 Billion on Crypto Deals While His Fans Lost a Fortune Roughly two-thirds of investors in the president’s memecoin are currently in the red

(…) Roughly two-thirds of investors in Trump’s memecoin are currently in the red, according to crypto data provider Nansen, which tracks 1.48 million crypto wallets that bought the token since its January 2025 launch. Many fans spent a few thousand on Trump coins while the biggest spenders shelled out millions for the token. Nansen’s analysis of 26,663 wallets shows that 85% of World Liberty’s $WLFI token buyers in the secondary market are underwater.

Trump, who in 2021 described bitcoin as a “scam” threatening the U.S. dollar, now leads a White House that has pledged to make America the “crypto capital of the world.” As his administration lightened regulation of the notoriously boom-and-bust sector, the Trump family’s sprawling crypto business reached into nearly every corner of the industry, drawing conflict-of-interest concerns from ethics watchdogs. Bitcoin prices also plunged. (…)

Days before his inauguration, Trump launched his memecoin, $TRUMP, which surged to a peak market capitalization of nearly $15 billion before plunging 97% to about $400 million today. In September 2024, Trump and his sons helped launch World Liberty Financial, the family’s flagship crypto venture.

That project subsequently released a dollar-pegged stablecoin shortly before Trump signed the Genius Act into law, establishing a regulatory framework for such dollar-pegged tokens.

White House officials say those actions don’t constitute conflicts of interest. “Neither the President nor his family has ever engaged—or will ever engage—in conflicts of interest,” White House spokeswoman Anna Kelly said. “All actions by President Trump and his administration are taken in the best interest of the American people.” [This is a recording] (…)

The WSJ Editorial Board:

• The Trump Family and ‘Honest Graft’ They are profiting off the Presidency in ways that demean the office.

(…) The Trump clan is cashing in on the Presidency in big and sketchy ways. (…)

We have detailed some of World Liberty’s dubious deals with foreign actors that may have been trying to buy influence with the Administration. This includes DWF Labs, a crypto firm based in the United Arab Emirates, and the Pakistan government. Abu Dhabi used World Liberty’s stablecoin to invest in the Binance crypto platform.

Binance pleaded guilty in 2023 to violating U.S. anti-money laundering laws and Iran sanctions. Its co-founder Changpeng Zhao, who pleaded guilty to violating the Bank Secrecy Act, was pardoned by Mr. Trump in October. The Journal reported in May that since the plea deals billions of dollars were routed through Binance and financed Iran’s regime.

Then there’s the Trump family’s financial alchemy involving government support for critical minerals. The Administration has provided billions of dollars in financing to an array of critical mineral developers to shore up U.S. supply chains so China can’t weaponize its dominance. The government investment has caused an industry gold rush.

The Trumps are seeing their opportunities and taking them, in the tradition of what Tammany Hall’s George Washington Plunkitt called “honest graft.” (For Princeton history majors, Tammany Hall was a 19th century Democratic political machine.) The President’s disclosure report shows capital gains in the range of $100,000 to $1 million from critical mineral developer MP Materials, whose share price surged last summer after the company announced the Pentagon was taking a 15% stake.

Donald Trump Jr.’s venture capital firm, 1789 Capital, invested in Vulcan Elements three months before the company received a $620 million government loan. Investment bank Cantor Fitzgerald, which is headed by Commerce Secretary Howard Lutnick’s sons, has been a financial adviser or underwriter for many startups that have received government support.

The Trump Administration has also offered up to $1.6 billion in financing for a project to develop a tungsten mine in Kazakhstan. There’s no question the U.S. needs to diversify its tungsten supply, which is critical to weapons systems. But a firm partly owned by Mr. Trump’s sons Eric and Donald Jr. invested in a company backing the project shortly before the Administration announced a minerals deal with Kazakhstan last November that paved the way for the project. Kazakhstan President Kassym-Jomart Tokayev posted a photo of himself at a White House deal-signing ceremony.

The Administration has proposed establishing price floors for critical minerals to ensure China can’t flood the market and financially undercut alternative suppliers, as it has in the past. This might make sense, but a government guaranteed price paired with financial support make these projects potentially very lucrative for investors like the Trump family.

***

Assuming all of this is legal, it’s still an unseemly display of using the Presidency for family profit. It’s hard to believe the Trump boys would be able to do the same deals if Dad wasn’t in the Oval Office. The main difference between Hunter Biden’s foreign dealings and the Trump projects is that the Trumps are brazenly open about theirs.

But there will be political costs—for Republicans. If Democrats take back the House or Senate this November, they will have a field day probing the Trump family deals. Charges of GOP corruption will resound through 2028. This will feed the left’s class warfare and facile narrative that billionaire “oligarchs” are getting rich off government.

Foreigners may come to think they can buy American goodwill or favors if they cut the Trumps in on the action. Americans, and especially his supporters, deserve better from this or any President.

“Foreigners may come to think”? Many foreigners are smarter than the WSJ EB and figured that out a while ago.

The mandatory recording again: “All actions by President Trump and his administration are taken in the best interest of the American people.”

Yesterday in the WSJ, Barton Swain, author of the “Unruly Republic” column, wrote

What the Founders Didn’t Trust Madison, Hamilton, et al. would wonder at our credulous view of human nature.

(…) The doctrine of original sin—and its implication that humans tend toward rank self-interest and degeneracy—animated every part of the constitutional structure the Founders erected. (…)

Spend the rest of this semiquincentennial year reading the Federalist Papers and mark how often the authors—Alexander Hamilton, John Jay and James Madison—base their contentions on the reality that fallen man can be counted on to conflate his own narrow interests with justice and the common good.

The essays refer often to man’s ingenuity and capacity for innovation and commerce—they didn’t think him so stupid as to require coercion. But Hamilton, Jay and Madison aimed chiefly to convince the public that the proposed Constitution, by balancing powers against each other, mitigated man’s propensity for mischief, violence and corruption.

Madison famously observes in No. 51 that government’s existence itself testifies to the morally defective state of its subjects. “If men were angels, no government would be necessary,” he writes. “If angels were to govern men, neither external nor internal controls on government would be necessary. In framing a government which is to be administered by men over men, the great difficulty lies in this: you must first enable the government to control the governed; and in the next place oblige it to control itself.”

Or Hamilton, in No. 15: “Why has government been instituted at all? Because the passions of men will not conform to the dictates of reason and justice without constraint.” (…)

The FT’s Editorial Board:

Supreme Court checks Trump, but empowers the presidency Despite losses on tariffs and birthright citizenship, the executive has emerged stronger

(…) The six Republican appointees continue to push the US right’s decades-long project to reshape the government by hobbling the administrative state and concentrating power in the executive. (…)

Although the court protected Cook, it allowed Trump to fire Federal Trade Commission Rebecca Slaughter without cause. That landmark decision overturns a 90-year-old precedent and kneecaps Congressional laws protecting the independence of federal watchdogs. The result will be more politicised and less predictable regulation, creating uncertainty for everyone from banks and dealmakers to pharmaceutical companies. (…)

The majority took another giant whack out of the seminal 1965 Voting Rights Act with a ruling that makes it much harder for minority voters to challenge legislative maps. Combined with earlier rulings that greenlight political gerrymandering, these rulings open the door to all manner of monkeying with the American electoral process.

Global experience with would-be autocratic leaders shows that free and fair elections and courts willing to check presidential power are critical to protecting a democratic future. (…)

The 2024 decision that granted Trump immunity from prosecution for official acts in his first term has empowered the president to take even more extreme positions now that he is back in office. T

his term’s decisions suggest that constitutional guardrails continue to hedge Trump in. But they are fragile and being undermined by the rightwing campaign to shift power to the presidency at the expense of Congress and independent regulators.

Justice Sonia Sotomayor’s dissent in the Slaughter case is chilling. Observing that the independent agency structure has been accepted since the early 1930s, she writes “the Court discards that democratic regime in favour of one that distorts the structure of Government to fit the majority’s theory of unitary, total executive control. The result is a President who emerges with far greater power than ever before.”

As the nation prepares to celebrate the 250th anniversary of throwing off the yoke of monarchy, it is a shame that the Supreme Court majority seems bent on recreating one.

FYI:

The Financial Times now reaches a global audience of roughly 20 million readers a month. Its readership is heavily international, affluent, and skewed toward senior business decision‑makers, with a notable C‑suite and finance professional concentration.

The FT emphasizes that roughly 70% of its readers are outside the UK, characterizing it as a “truly global” publication.

The FT’s opinion pages have outsized influence because they reach a concentrated global audience of policymakers, corporate leaders, and investors, and are widely perceived as authoritative, economically literate, and relatively centrist on business issues.