A 34-Year-Old Airline Just Died - Are You Next On the List

Consumer sentiment just hit the lowest level in 50 years of University of Michigan data. The S&P 500 just had its best month since the pandemic.

Same week. Same country. Different planets.

Americans are more pessimistic than during the Great Recession, the 2022 inflation spiral, and the COVID crash. Meanwhile, Big Tech committed $725 billion to AI infrastructure. With a straight face. While beating estimates.

This isn't cognitive dissonance. It's a structural split.

The economy bifurcated into two parallel realities. One has a stock portfolio. The other has a Costco gas card. Knowing which one you're in determines whether you make money or lose it this year.

Let's review.

The headline is genuinely historic. Michigan Consumer Sentiment fell to a final April reading of 49.8. Down from 53.3 in March. Below the prior all-time low of 50 set during June 2022's inflation peak.

Preliminary data briefly hit 47.6. That's worse than 2008. Worse than the 1980s stagflation shock. We had to invent a new floor.

Year-ahead inflation expectations jumped to 4.7–4.8%. Translation: Americans think prices keep climbing, things get worse, and the Iran war is the villain on the marquee.

The Iran link is direct. The Strait of Hormuz closure pushed gas up roughly 40% since the conflict began. Goldman Sachs pegs the annualized hit to household incomes at $140 billion. The lowest-income quintile spends four times as much on gas as the top quintile, because they spend a much larger share of their paycheck at the pump.

And yet.

March core retail sales rose +4.6% year-over-year. Fastest pace since last August. AmEx says premium cardholders are "rolling with it." Travel bookings are strong. PepsiCo sees Frito-Lay volume recovering. Ulta says skincare is recession-proof.

Spoiler: skincare has been recession-proof since the Great Depression. People will eat ramen, but they will not look ashy doing it.

So what gives?

It's not a paradox. It's a K-shape inside the K-shape. Upper-income households are spending fine. Lower-income households are being flattened by $4-plus gas prices. Surveys capture mood. Retail data captures wallets that still have something left in them.

When sentiment readings hit 1980 levels, the wallet usually follows. Eventually.

This was the week markets answered the only question that mattered. Would the hyperscalers flinch on AI capex?

They did not. They doubled down. Then they bought another data center to celebrate.

Going into earnings, combined 2026 AI capex commitments from Microsoft, Amazon, Alphabet, and Meta sat at $630–670 billion. After Wednesday, the number jumped to roughly $725 billion. Every single one raised guidance:

All four beat estimates. The market's split reaction (Alphabet up, Meta and Microsoft down) wasn't about the numbers. It was about optics.

Meta raised spending more aggressively than the others, while Reality Labs continues its slow-motion arson. Cumulative losses since 2020 just crossed $83 billion. That's roughly Lithuania's GDP, set on fire to make legless avatars.

The market is keeping score on Zuckerberg's metaverse. The score is still bad.

NVIDIA closed Thursday down 4.6%. The S&P printed a record. That's not a coincidence. That's a memo.

Both Google and Amazon used their earnings calls to announce something quiet but seismic. They're moving custom AI chips outside their own clouds.

Sundar Pichai confirmed for the first time that Google plans to deliver TPU chips directly to customer data centers this year. Andy Jassy hinted Trainium racks could ship beyond AWS "over the next couple of years."

Translation: Nvidia's two biggest customers are about to become its two biggest competitors. By appointment.

To be clear: this isn't a coup. Google's TPUs and Amazon's Trainium currently run narrower workloads. Inference is their initial turf, not the heavy frontier-model training Nvidia dominates. Custom AI chip sales are projected to grow 45% in 2026, compared with 16% for standard GPUs. NVIDIA still holds 92% of the specialized AI chip market.

What changed is the direction of travel. One analyst called the shift "extraordinary" and "irreversible." When the people writing your largest checks start printing their own silicon, that's not a moat. That's a countdown.

The line to watch is the split between commodity and premium chips. As long as H100s and Blackwell deliver better performance for frontier training meaningfully, the moat holds. The moment "good enough" inference silicon becomes acceptable to cost-conscious enterprises, the rotation accelerates.

Sound familiar? It should. This is how every monopoly ends. Slowly, then all at once.

Jerome Powell held rates at 3.5%–3.75% at what was widely expected to be his final FOMC meeting as chair. The decision on the rate call was unanimous. Everything around it was a brawl.

Four dissents colored the meeting. Three members disagreed with the dovish-leaning language in the statement. That's the highest level of internal dissent since 1992. The last time the FOMC fought this hard, "Achy Breaky Heart" was on the radio.

Rate-hike odds by December quietly ticked up from 0% to 9% on CME futures. Small number. Directionally loud. March inflation came in at 3.3%, hottest since May 2024. With oil still elevated by the Iran conflict, some economists now say cuts will be delayed until 2027.

Powell made one notable move on his way out. He's staying on as a Fed governor instead of leaving the board entirely. He cited the ongoing DOJ investigation and his concerns about political interference. He put it plainly: "I am concerned that these assaults are undermining the institution and jeopardizing the public's essential need for independent monetary policy, free from political considerations."

Trump's response was characteristically warm: "I don't care. If he stays on, he stays on… but I think he's a negative force."

Strong "I'm not mad, you're mad" energy.

Kevin Warsh, Powell's successor, confirmed as chair-designate after Senator Tillis unblocked the nomination, has already signaled he wants "messier" Fed meetings. He carries a hawkish bias. His arrival on May 15 will be the most consequential Fed leadership transition in at least a decade.

Enjoy that crisis.

Spirit Airlines ceased all operations at 3 AM Eastern on Saturday, May 2, 2026. The carrier that once flew millions of passengers a year on its distinctive yellow aircraft became the first major U.S. airline failure in 25 years.

The proximate cause: a $500 million federal bailout request the Trump administration couldn't close. A critical group of creditors rejected the final rescue proposal. Spirit's last flight, Flight 1833 from Detroit to Dallas, landed just after midnight.

Cinematic exit. Yellow plane fading into the night. Roll credits.

The structural cause is older. Spirit's ultra-low-cost model was already broken before Iran. The carrier filed for bankruptcy twice since 2024. What the Iran war did was apply the kill shot. Jet fuel costs spiked 40%. Spirit had no balance sheet left to absorb it.

When your business model is "we barely break even at normal fuel prices," an energy shock isn't a headwind. It's an extinction event.

17,000 jobs gone. 14,000 Spirit employees plus thousands of contractors. Millions of ticketholders stranded. United, Delta, JetBlue, and Southwest capped one-way fares at ~$200 for displaced passengers. Spirit customers with unused tickets are due refunds. Anyone holding Free Spirit points or vouchers may have no recourse at all.

Free Spirit points. The name aged like milk in a parked Cessna.

Why it matters: Spirit's collapse pulls real low-cost capacity off price-sensitive routes. That's a tailwind for Frontier, Allegiant, and Sun Country. A milder one for legacy carriers on leisure routes where Spirit competed. Airfare on Spirit's former routes will rise. For traders: watch capacity data and yield stats from Delta and United over the next two quarters. The bigger picture? Spirit isn't an anomaly. It's a preview. The Iran war is a slow-moving stress test for every capital-intensive, margin-thin business in America. Airlines went first.

There's a list. We're working through it.

The macro backdrop in three sentences. The economy is running on AI investment heat while the consumer smolders. The Fed is frozen between sticky inflation and an energy shock. Geopolitics, specifically the Iran conflict and the Strait of Hormuz, is the lever controlling it all.

Where the risk is:

Where the opportunity is:

April 2026 will be remembered as the month AI stopped being a thesis and became the economy. Five of the Magnificent Seven beat estimates. Combined capex hit $725 billion. The S&P posted its best month since the COVID pandemic.

And underneath all of it, the war in Iran quietly claimed its first major corporate victim, crushed real household incomes, and pushed a central bank into paralysis.

The market is pricing a peaceful resolution and an AI-powered profit cycle. That may well be right. Markets that price perfection just leave no room for the surprises history always delivers.

The S&P is near 7,200. Consumer sentiment is at an all-time low. A 34-year-old airline is gone. The Fed changes hands in 13 days.

Pay attention. The split-screen economy doesn't last forever. One side eventually wins.

Here's hoping it's the one with your portfolio.

This is not financial advice. It’s market commentary from people watching chaos unfold in real time.