G
Gainbrief
Tag

#business

2 posts in this community.

AAAaron···5 min read

The Health-Care Bill Is Becoming a Paycheck Story

The sharpest cost surprise in the U.S. economy right now may not be at the gas pump or in the Treasury market. It may be sitting inside employer health plans. That is why this week’s new Milliman Medical Index landed as more than a healthcare statistic. It looked like a wage story, a corporate margin story, and a consumer spending story at the same time. Milliman said the annual cost of healthcare for a typical American family of four covered by an employer plan reached $37,824 in 2026. For the average person, the cost reached $8,460. The average-person figure rose 7.9% from 2025, which Milliman said was the biggest increase in more than a decade if you set aside the pandemic distortions. That is a big number on its own. It matters even more because most workers do not experience it as one obvious bill. They experience it as slower wage growth, narrower benefit choices, higher deductibles, tighter provider networks, and more anxiety about actually using the coverage they technically have. That is the part of the health-cost story I think markets still underprice. Investors spend a lot of time debating consumer resilience. Can households keep spending? Can employers keep hiring? Can margins hold if rates stay higher for longer? Employer health inflation cuts through all three questions at once because it acts like a quiet tax on compensation. If a family plan is now pushing toward $38,000 in total annual cost, the issue is no longer just whether the employee’s payroll deduction went up. The issue is what the employer is not doing with that money instead. Less room for raises. Less room for bonus pools. Less room for richer benefit design. Less room for aggressive hiring in labor-intensive businesses. That is why I read the Milliman release as finance news, not niche health-policy news. The drivers matter. Milliman said outpatient facility care now makes up roughly 31% of employer-sponsored healthcare spending, making it the biggest single category. Pharmacy was the fastest-growing component, rising 14.8% year over year for the average person. Milliman also said outpatient care and pharmacy together accounted for 69% of the annual increase. That should sound familiar to anyone watching public markets. Outpatient settings have become a bigger revenue engine across the healthcare system. GLP-1 drugs keep pulling more dollars into pharmacy spend. Provider consolidation keeps weakening the buyer’s leverage. The result is that employers, even very large self-insured ones, are often negotiating from a structurally weak position. KFF’s employer survey already showed where this was heading. In 2025, the average annual premium for family coverage hit $26,993, with workers contributing $6,850 toward that total. The new Milliman number is not directly comparable because it includes the broader cost of care, not just premiums. But together they tell the same story: the visible premium pain was already high, and the invisible cost base behind it kept getting worse. That is the hidden squeeze. For workers, healthcare inflation does not always feel like inflation because so much of it is filtered through employers. The money disappears before it shows up in take-home pay. That makes it politically quieter than rent or groceries, but not economically smaller. In some cases it may be more important, because it directly competes with wages. For companies, especially outside big tech, this is a harder problem than it looks. Employers can push more cost onto workers, but only up to a point before retention gets worse. They can narrow networks, but that tends to create employee blowback. They can push high-deductible designs, but that often just changes when people seek care, not the underlying system prices. They can carve out pharmacy strategies, but drug trend is still drug trend. This is where the New England Journal of Medicine perspective from earlier in May was useful. Its argument was that self-insured employers are a sleeping giant in health-care affordability. That framing feels right to me. Big employers have scale, claims data, and purchasing power on paper. In practice, they are still fragmented buyers dealing with a healthcare system that is better at monetizing complexity than reducing it. That gap matters for investors because it pushes this story beyond insurance stocks. If employer health costs keep compounding this way, the effects show up in more places than UnitedHealth, CVS, Humana, or hospital operators. They show up in wage negotiations, labor cost assumptions, benefits consulting, pharmacy benefit management, and consumer discretionary demand. They also show up in the earnings quality of companies that look healthy until you notice how much of their compensation growth is being eaten by healthcare. I think that is why this topic deserves more attention right now. The U.S. market has spent the last year obsessing over AI capex, tariffs, rates, and the consumer. Fair enough. But employer healthcare costs are one of the clearest examples of an economic pressure that is both large and easy to miss. It is easy to miss because the burden is dispersed. No single worker sees the full employer contribution. No single quarterly report calls it out cleanly unless medical trend really blows up. No single inflation print captures the labor-market consequences. But the drag is real. The near-term question is whether employers become more aggressive buyers or continue acting like price takers. Milliman pointed to GLP-1s, changes in how pharmacy costs flow to plan sponsors, and AI-assisted billing as important forces to watch. None of those look like one-quarter issues. They look structural. My read is that 2026 may be the year more executives stop treating health benefits as a background HR expense and start treating them as a capital-allocation problem. If healthcare claims inflation keeps outrunning comfort, then benefits strategy becomes part of margin strategy. And if benefits strategy becomes part of margin strategy, this stops being a healthcare-side story and becomes a mainstream business story. For U.S. readers, the practical takeaway is simple. When you hear that the labor market is holding up, or that wage growth looks decent, or that the consumer has more room than expected, it is worth asking what part of that picture is being quietly offset by employer health costs. This week’s Milliman data suggests the answer is: more than most people think.

0
0
AAAaron···4 min read

The AI Trade Has Reached the Utility Bill

The AI trade just bought itself a utility problem. This week NextEra Energy agreed to acquire Dominion Energy in an all-stock deal valued at about $67 billion. On the surface, that looks like a standard scale merger in a defensive sector. It is not. It is one of the clearest signs yet that the AI boom is moving out of the data center and into the power grid, where the economics are slower, more political, and much harder to ignore on a monthly bill. The headline reason is simple. America suddenly needs a lot more electricity, and not because households are buying more refrigerators. The U.S. Energy Information Administration said in January that power demand is set for its strongest four-year growth stretch since 2000, driven largely by large computing facilities. That changes the way investors should look at AI. The first phase of the trade was about chips, cloud contracts, and model releases. The next phase looks more like substations, transmission lines, gas generation, batteries, and regulated rate cases. That is why the NextEra-Dominion tie-up matters beyond utility investors. In their announcement, the companies said the combined business would serve about 10 million customer accounts, own 110 gigawatts of generation, and carry more than 130 gigawatts of large-load opportunities in its pipeline. That is not normal merger language. It is AI-era language. Utilities are starting to describe their future in terms that sound more like hyperscaler capex decks than sleepy dividend presentations. Dominion is especially important in this story because Virginia has become the physical heart of the U.S. data-center economy. S&P Global noted this week that Dominion Energy Virginia has about 51 gigawatts of data-center capacity in various stages of authorization and contracting, including 10.4 gigawatts already under electric service agreements. You do not need every one of those projects to arrive on time for the point to land. Even a partial buildout implies massive demand for generation and grid investment. The harder part is what this means for everyone else. The independent market monitor for PJM, the largest U.S. power market, said last week that its recent capacity auctions were not competitive primarily because of forecast demand from data centers. It also said total wholesale power costs in the first quarter rose 75.5% from a year earlier. Not all of that increase drops directly onto consumer bills, and weather and fuel costs still matter. But the direction is hard to miss: large new loads are colliding with a grid that was not built for this pace of demand. That is where the politics begin. The merger pitch from NextEra and Dominion leans heavily on scale. Management says a bigger platform will buy equipment more efficiently, finance projects more cheaply, and spread costs better across a larger base. The companies are even offering $2.25 billion in bill credits over two years for Dominion customers in Virginia, North Carolina, and South Carolina. That is a meaningful concession, but it also reads like an early admission that affordability will be the central regulatory fight. Investors should take that seriously. AI infrastructure still has a popular narrative problem: markets tend to treat demand as unquestionably good and supply as somebody else's engineering detail. In power, supply is never just an engineering detail. It runs through public utility commissions, local opposition, environmental permitting, fuel procurement, balance-sheet capacity, and the simple fact that power plants and transmission lines take time. A utility cannot scale the way a software platform scales. That mismatch may become one of the most underappreciated constraints in the AI buildout. If the grid becomes the bottleneck, then some of the biggest economic gains from AI get delayed, repriced, or redirected. Data-center operators may have to pay more for dedicated generation. Utilities may push for special large-load tariffs so new industrial users bear more of the cost. Regulators may insist that residential customers be protected before any merger synergies are allowed to count as a public benefit. In plain English, AI is entering the part of the economy where growth gets negotiated. For markets, that creates a wider opportunity set than the usual semiconductor leaderboard. The obvious beneficiaries include utilities with credible growth territories, transmission developers, gas and nuclear suppliers, power-equipment makers, and companies that can finance generation quickly. But there is also a real risk layer. If customers keep seeing electricity bills rise while data centers multiply, utilities may face a backlash that limits how fast they can pass through costs or win approval for new infrastructure. This is why I think the NextEra-Dominion deal deserves to be read as a warning as much as a bullish signal. It is bullish because it confirms that AI demand is strong enough to reshape the largest regulated parts of the economy. But it is also a warning because once AI demand shows up in household utility politics, the trade stops being purely about innovation and starts becoming about who pays. That is a different market. It rewards scale, patience, and regulatory skill more than flashy product cycles. It also pulls the center of gravity away from Silicon Valley and toward state capitals, grid operators, and utility commissions. The AI story is still expanding. It is just expanding into a part of America where every growth plan eventually meets a ratepayer. For U.S. readers, the takeaway is straightforward. If you want to understand the next leg of the AI economy, do not only watch Nvidia, cloud capex, or model benchmarks. Watch utilities, wholesale power markets, and the political fight over electric bills. The smartest way to read this merger is not that utilities are joining the AI trade late. It is that the AI trade has finally reached the real economy, and the real economy sends an invoice.

0
0