Tea Leaves

Tea Leaves #10: MAVERICK MOMENT

Phillip Stegner — Thu, 20 Aug 2026 01:27:04 GMT

The chip manufacturer Etched is reportedly in talks to raise money at a $20bn valuation. An unsurprising development in the world of AI.

What is surprising? These reports emerged while Etched was actively finalizing yet another round, at a valuation of only $10bn.

The sequence reminds me of Top Gun – of Maverick and Goose sauntering to their jet, high fiving over their “need for speed.”

While “speed” has numerous economic applications, I’m most intrigued by its derivative cousin, “acceleration” – for our purposes, what I’ll define as the shrinking interval between economic events.

Specifically, I’m interested in what happens when acceleration picks up, forcing new action before the last has fully played out –

– what we might call a “Maverick Moment,” riffing on the obvious intersection of Tom Cruise and Hyman Minsky. A moment when capital commits before cognition kicks in.

We’re generally taught to avoid this behavior. Just two weekends ago, the Journal ran a personal finance piece on sound money management with the subhead, “The key is to put a little friction between us and our purchases.”

Could the same extend to capital markets? Consider the following.

Speed: On a Friday, SpaceX goes public at a $1.8tn valuation. Acceleration: The following Tuesday, SpaceX uses its newly public equity to purchase Cursor in a $60bn all-stock deal. Outcome: Tuesday’s M&A currency has less than two business days of ground training before wheels up.

Currency that closed yesterday nearly 30% off its post-IPO peak.

Speed – in our case, large dollar transactions – isn’t inherently bad. If Etched sees an opportunity to raise capital, it should have at it. My brief stint as an investment banker notwithstanding, I’m grossly unqualified to judge.

But when acceleration forces the next decision before the feedback loop closes on the last, we may sacrifice sound judgment at the altar of rapid response.

We may find ourselves in a Maverick Moment.

And we know how that ended for Goose.

Tea Leaves #9: VEBLENIAN MARKETING

Phillip Stegner — Thu, 20 Aug 2026 01:26:05 GMT

Thorstein Veblen is hardly a household name. To the extent you’ve heard of the 19th-century economist, I’m guessing it’s in the context of “conspicuous consumption,” his best-known contribution to the theoretical lexicon.

More on that another time. Today, I’m interested in Veblen’s thoughts on marketing.

As John Cassidy notes in Capitalism and Its Critics, Veblen’s opinions on the topic were none too flattering. In Veblen’s telling, marketing adds nothing “to the serviceability of the output,” offering no “utility to the last buyer.” Veblen decries advertising as “wasteful.”

Much of it, “parasitic.”

Had a resurrected Veblen read this weekend’s Wall Street Journal, I assume the Exchange section would’ve infuriated him. Veblen would’ve loathed an article on a Wendy’s turnaround effort built, in part, on “stepped-up” marketing; in Heard on the Street, he would’ve fumed over Reformation’s perceived need to advertise for growth.

To be clear: I’m no Veblenian purist. There is clear value in telling consumers about new products and services; in reinforcing a brand’s value proposition. I work with brilliant marketers who’ve taught me to think in these terms.

But Veblen does provide an interesting lens through which to view this weekend’s reporting. Yes, near-term growth can often be purchased – companies can convert browsers to buyers via a one-time promotion, a clever ad campaign, increased acquisition spend.

A less cranky Veblen, however, would smirk as he probed further: Do those buyers ever come back?

In the same weekend Journal, Under Armour’s lackluster quarter – a quarter in which it leaned heavily on promotions to offset weak traffic – may provide a clue. The CEO intends to “cut back its reliance on promotions,” noting the company “must earn [consumer demand] through more compelling reasons to buy.”

Sometimes, sustainable growth must be earned, not simply purchased.

Perhaps Veblen had a point.

Tea Leaves #8: PATRONIZING PASTRIES

Phillip Stegner — Thu, 20 Aug 2026 01:24:58 GMT

Sopaipillas. Beignets.

Once relegated to the last page of the menu, you’ll now find these delicacies in SEC filings. They’re moonlighting as the names of trusts financing Meta datacenters – nods to the states in which they’re being built.

I’ve named entities like these in my career. As a fintech founder, I raised capital to fund loan sales, often financed through independent trusts. Naming those trusts was a “fun,” near-final step in each transaction, but I treated the task semi-seriously. I knew these names could resurface at inconvenient times.

I learned that lesson the hard way, years earlier at the Federal Reserve during the Great Financial Crisis, watching Citi grapple in real time with fallout from sweetly named sidecars. Sidecars like “Beta Finance.” “Centauri Corporation.”

My favorite? “Dorada,” or “golden.”

With the benefit of nearly 20 years’ hindsight, I trust Citi executives will forgive my eyeroll.

It’s no secret capital markets have no long-term memory. It’s a tune Tea Leaves readers have heard before. Today’s verse? What’s born in the dark comes to light at inopportune times. And while an innocuous-turned-ironic name is cheap, the assumption of millions or billions in liabilities once believed sectioned off … is not.

I can’t predict the financial path of Meta’s datacenters. But I do hope someone – in the offices of Meta, its investors, its bankers – has enough historical sense to take a free lesson from Citi.

To recall that desserts can spoil.

That they can become patronizing pastries.

Tea Leaves #7: NOBODY KNOWS WHAT IT IS

Phillip Stegner — Thu, 20 Aug 2026 01:23:51 GMT

Only the rarest of nerds has a favorite overview of the South Sea Bubble. And in this context, I am him. My favorite retelling of this 300-year-old mania can be found in Charles Mackay’s Extraordinary Popular Delusions and Madness of the Crowds, first published in 1841.

In the text, Mackay colorfully details the speculative frenzy that overtook London’s Exchange Alley in 1720, a year when promoters floated shares of all kinds.

Some, seemingly reasonable. For “the importation of Swedish iron.” For “supplying London with seal-coal.”

Some, perhaps questionable. For “a wheel of perpetual motion.” For “the transmutation of quicksilver into a malleable fine metal.”

Some, downright mad. For “carrying on an undertaking of great advantage;

“but nobody to know what it is.”

I recalled Mackay’s words when I came across recent FT coverage of AI-adjacent energy companies fundraising in the current AI wave. The article’s punchline, quoting TD Cowan’s Jeff Osborne, landed particularly forcefully.

“Often, those faring better [in the market] … are ‘less of a science experiment.’”

Nearly two centuries separate Mackay from Osborne, yet you could lift and shift the sentiment almost verbatim.

Words are powerful. They have meaning. They set expectations. And over time, entities that live up to those expectations earn a reputation, itself a valuable asset.

But in a bubble, fashionable labels take on a life of their own. They flood the lexicon, attracting mountains of capital until costly commitments are all that remain. Until what it means to be “AI-adjacent” is a muddled mess.

Until no one knows what it is.

Energy companies headed to market today would do well to remember history.

Labels fade. Expectations do not.

Tea Leaves #6: KALSHI & CAMELS

Phillip Stegner — Thu, 20 Aug 2026 01:20:56 GMT

CME has had a busy summer. Just last month, it sued the CFTC. And according to this weekend’s FT, its market cap has fallen nearly 20% in a matter of weeks.

Why? Talking heads would say “Kalshi.”

More specifically, the CFTC’s approval of Kalshi’s bitcoin “perps,” or “perpetual futures” – a product with which CME takes issue.

A product I won’t pretend to understand.

What I do understand: Financial history. And the story of Kuwait’s Souk al-Manakh, a shadow equity exchange established in 1978, feels pertinent.

As Ben Craig of the Cleveland Fed writes, the Souk was born of investor demand to trade stocks unavailable on the traditional exchange (the Boursa). Their needs unmet, these traders struck out for a new home.

Literally. They moved to a converted parking garage, a location that derived its name from the market that preceded it.

The Souk al-Manakh. Translated, “the market at the resting place of camels.”

To insulate the Boursa from the Souk, regulators drew what they believed were clear lane lines. Critically, they banned bank financing of Souk trades – an attempt to mitigate the risk of broader disruption if things went awry.

But as we know, systemic risk is stubborn.

Without traditional funding sources, Souk traders became reliant on a system of post-dated checks as a credit and liquidity substitute. That system grew to billions before confidence broke and the exchange collapsed.

The fallout extended to Kuwaiti banks, whose customers became overextended in the Souk and were subsequently unable to repay other obligations. The risks thought relegated to the shadows of a parking garage brought down the house.

“In the end,” as Craig notes, “only one bank in Kuwait’s robust commercial banking system was solvent.”

To be fair, Kalshi isn’t a camel trading post. It’s regulated by the CFTC, an entity whose inner workings aren’t known to me.

But if underpinning all of this is a regulatory belief that the broader market can be inoculated by isolation, the Souk says think again.

Tea Leaves #5: AI TRACKS TO NOWHERE

Phillip Stegner — Thu, 20 Aug 2026 01:19:38 GMT

This week’s Economist pithily summarized recent AI fundraising: “Nowadays it is easy to get desensitised to big numbers.

“Every week stonking deals are announced.”

This isn’t surprising: Extraordinary productivity boosts can trigger “stonking” deals.

But history suggests those deals don’t initially fund the most productive investments.

For proof, consider the British railway mania of the 1840s. In many ways, railways were 19th century data centers – catalysts for generational productivity growth.

As Edward Chancellor writes, at the mania’s apex, rail investments were “projected [to exceed] the country’s national income,” funding tracks “twenty times the [country’s] length.”

Tracks that led “to faraway places that could never repay the cost of investment.” Tracks to nowhere.

Like rail construction that preceded it, AI projects do not want for capital. But capital wants for direction.

Listen to executive interviews and speeches on the subject, and you’ll hear plenty about “falling behind.” Plenty about “catching up.” You’ll hear less about expected returns.

Just the urgent need to “adapt.”

A strategy grounded in fear may minimize executive regret, but it doesn’t maximize the good. It trades the discomfort of thoughtful decisionmaking for the comfort of the herd.

In The Price of Time, Chancellor illustrates this point with the story of Watchet. While the small British parish wasn’t exactly “nowhere” in the 19th century, it certainly wasn’t a metropolis that required three independent railways – precisely what was proposed in the manic wave.

Hidden in a footnote, Chancellor provides an update on Watchet: While “[none] of [the three proposals] got off the ground,” a singular passenger line was finally opened later.

And that line “continues in operation today.”

With the help of AI, businesses will build profitable Watchet railways. But continued pursuit of this technology for its own sake all but guarantees we’ll lay more unprofitable tracks first.

Tracks to nowhere.

Tea Leaves #4: RUBE GOLDBERG & REGULATION

Phillip Stegner — Thu, 20 Aug 2026 01:18:14 GMT

Yesterday, the WSJ ran a story on debit card fees – not a riveting topic. Not riveting, that is, until you read that banks are exploring the acquisition of payment networks to skirt a 2010 Dodd-Frank rule that caps collectible interchange revenue.

Presumably not what legislators intended in the aftermath of the Great Financial Crisis.

Episodes of financial stress are (rightfully) a magnet for regulation. But a complex system is just that: complex. A Rube Goldberg machine in which the slightest turn of a gear deploys a mouse trap. A truism we either tend to forget, underestimate, or flat out ignore.

A truism best exemplified in the savings and loan collapse of the 1980s.

When the Great Depression threatened the industry’s viability, Congress established the Federal Home Loan Bank System and governing board. It quickly followed suit with the Federal Savings and Loan Insurance Corporation, or FSLIC – a fund Congress housed in the bank system itself.

When competition for retail funding intensified, policymakers and regulators loosened Reg Q, allowing S&Ls to pay deposit rates above those of commercial banks.

When high deposit rates collided with low, fixed-rate mortgages – the industry’s bread and butter since inception in 1831 – S&Ls sought and received approval to invest in an array of alternative, higher yielding assets.

When insolvencies spread, regulators implemented more forgiving accounting rules. Rules that “[cynics] quickly dubbed … creative accounting principles.” Or “CRAP,” according to Kathleen Day, author of S&L Hell.

At each step, regulators acted. At each step, S&Ls reacted. Action, reaction. And each fix today yielded trouble tomorrow.

How did it end?

The FHLB Board devolved into a conflicted mess. The line between regulation and advocacy disappeared. S&Ls used their newfound freedom to wipe out their capital. FSLIC went bust.

A Rube Goldberg machine run amok.

Back to yesterday's Journal, the source of our story on debit card fees. Tucked away in This Day in Market History, a reader would have learned that on July 7, 1993, an individual named Charles Knapp “was found guilty on three counts of conspiracy ... costing US taxpayers $2 billion.” The business Mr. Knapp owned?

A savings and loan.

Tea Leaves #3: THE NEVER-ENDING HANGOVER

Phillip Stegner — Thu, 20 Aug 2026 01:15:28 GMT

The WSJ last week reported a surge in US imports, driven in part by retailers building holiday inventory ahead of expected tariff increases.

The expectation of inflation begets … inflation? An economist would yawn.

So too might a historian. In Adam Fergusson’s When Money Dies, he cites V.W. Germains’ account of the hyperinflation that gripped Germany in the early 1920s. As Germains noted, “Women bought in stocks of … goods against a rise in prices … Others squandered money recklessly;

“wine would be dearer tomorrow!”

Dearer wine tomorrow fuels drunkenness today.

But what happens when we wake to find that wine remains dear? What happens when it’s etched into our brain that the price of wine will continue to rise, day by day?

One of the more popular measures of inflation expectations comes from the University of Michigan’s monthly Surveys of Consumers. At last release, the Wolverines pegged 12-month inflation expectations at 4.8% – a figure meaningfully hotter than recent prints of just about any popular price change index.

Since 2020, those expectations have remained stubbornly elevated – Americans see nothing but expensive wine on the horizon. This, despite the fact our bar tab has only increased spectacularly in recent memory.

In other words: Despite the already elevated real-time readings, consumers expect inflation to persist, at an even higher level. And they’re not backing down.

As the yawning economist would remind us, inflation expectations aren’t simply a tool for predicting the future. They can change behavior in the present.

For proof, look no further than retailers building inventory ahead of tariff hikes – a self-fulfilling trend that, if maintained, becomes incredibly difficult to buck.

Exogenous shocks dissipate. Businesses find ways to re-direct supply chains. A war, in theory, ends.

But until the American consumer believes wine will be cheaper tomorrow, it remains happy hour today. And the hangovers may never end.

Tea Leaves #2: THE BIGGIE SMALLS DILIGENCE MODEL

Phillip Stegner — Thu, 20 Aug 2026 01:12:43 GMT

The WSJ recently reported of a Texas rancher who raised $170mm from lenders and investors to buy and sell cattle. When all was said and done, the rancher was found to be a fraud – his 80k-head operation actually had fewer than 10k animals.

The discovery came during the rancher’s first collateral inspection.

The first collateral inspection. In four years. After the lender had built a $50mm exposure.

Collateral fraud isn’t surprising. What is surprising is how often lenders get surprised by it.

In Manias, Panics, and Crashes, Charles Kindleberger tells the story of Tino De Angelis, a 20th century commodities trader. In the 1960s, De Angelis borrowed heavily against tanks purportedly filled with salad oil. As Kindleberger notes, “Tino knew that … salad oil was less dense than water…

“[so] he floated a six-inch layer of salad oil on top of twenty feet of water [to maximize his loan proceeds].”

Nothing new.

So why do we keep getting surprised?

I’ve long argued a strong credit program conforms to the Biggie Smalls Diligence Model. Where weak credit programs only scratch the surface (“what’s your name, what’s your sign”), a strong credit program – like Biggie – goes deeper. A strong credit program asks “what your interests are.”

A strong credit program asks, “Who you be with?”

Until lenders approach collateral the way the Notorious B.I.G. approached dating in the club, history will repeat itself.

Whether the collateral is salad oil or cattle. The (right) question remains the same.

Who you be with?

Tea Leaves #1: SPACEX AND MISSISSIPPI

Phillip Stegner — Thu, 20 Aug 2026 01:08:03 GMT

Seeing Cursor’s acquisition by SpaceX – paid for in SpaceX shares – reminded me of John Law’s infamous Mississippi Company of the early 18th century.

At the height of the Mississippi Mania, shares traded in the Rue de Quincampoix at a furious pace. Charles MacKay, in Extraordinary Popular Delusions and the Madness of Crowds, told of the poor and indigent renting out their backs as writing desks for the execution of sales, minting new “millionaires” (a term reportedly, contemporaneously coined) overnight.

No, what struck me about the Cursor acquisition isn’t the AI frenzy itself. It’s the medium of exchange – SpaceX paying in SpaceX shares.

When an asset is used to pay for other assets, it plays the role of money – it becomes a medium of exchange.

Much like John Law’s Mississippi shares did.

Edward Chancellor cites, in Devil Take the Hindmost, an anonymous pamphleteer writing on John Law’s bubble:

“The additional rise of [Mississippi shares] above the true capital will be only imaginary; one added to one, by any rules of vulgar arithmetic, will never make three and a half; consequently, all the fictitious value must be a loss to some person or other, first or last. The only way to prevent it to oneself must be to sell out betimes,

“and so let the Devil take the hindmost.”

Every great mania is rooted in some nugget of value. The Mississippi Company finally did finance (at least a few) real ventures. Railroads transformed the world. The Internet re-wrote the laws of commerce, of life. AI is supposed to change everything, everywhere, all at once.

The question isn’t the innovation – it’s who pays for it, at what price, and how.

Three hundred years after John Law's Mississippi Company, the question remains the same: Who will take the hindmost?