Wintermute and Galaxy Digital bought USDC on exchanges at about 99.9 cents and redeemed it through Erebor, the bank backed by Palmer Luckey that is raising at an $8 billion valuation, which had offered stablecoin-to-cash redemptions with no fees.
The trade itself was small. Ram Ahluwalia's interest is in what it says about pricing a private bank at all, because banks are valued on tangible book value and private markets have not been asked to care.
"No one will care about that while it's in private markets, but in public markets they would care about that."
Ahluwalia runs Lumida Wealth, which puts client money into alternatives and digital assets; he saw the Erebor deal early at a lower price and passed, and his firm invested in CoreWeave before it became the first neocloud to go public.
The full episode is covered here so you can skip it. 19 minutes of audio, 12 minutes of reading.
Here are the 11 takeaways that matter.
👤 Speaker: Ram Ahluwalia, CEO and Founder of Lumida Wealth Management, a registered investment adviser focused on alternative assets and digital assets
📰 Published: 16 September 2026 on the Lumida Wealth YouTube channel and podcast feed
🔴 YouTube | ⏱️ 19 min | ✅ Time saved: 7 min
Key Takeaways
A fee-free stablecoin redemption at a private bank turned into an arbitrage for two trading firms buying USDC at about 99.9 cents
His concern is the exit: a bank that public markets would value on tangible book is raising at $8B privately
He passed on Anduril for the same reason, at $60B going on $100B against $200B for the incumbent primes
The stranded-gas trade in Alberta turns shut-in wells into data-center compute and then frees the oil for trucking at $40 a barrel
Mark Carney's accelerated depreciation is a good policy that may not be enough, because the founders already left for Texas
He says the Fed oversteered by cutting last year and is now raising into five years of above-target inflation
The trade he wants someone to do is rescue financing for small-cap companies with five-year debt they cannot refinance
Magnetar's warrants on CoreWeave's credit facility are his template
He told a university sophomore not to start a prediction-market fund or join a long-short fund, and to get into Y Combinator instead
1. The Erebor Arbitrage
The episode opened on Erebor, the crypto-focused bank, and a redemption offer that two trading firms took apart.
The mechanics were simple. "Wintermute and Galaxy Digital bought USDC on exchanges at around 99.9 cents." The bank had offered stablecoin-to-cash redemptions with no fees, so the fraction of a cent was free money.
Ahluwalia's summary of who won: some trader at Wintermute got a nice bonus.
The company itself is a fundraising story. He said it is raising at an $8 billion valuation, backed by Palmer Luckey — whose attention, he noted, is substantially on Anduril and other things — with a group of Luckey's contacts putting in $100,000 each through a WhatsApp group.
His view of the valuation path is that it keeps going up, and that this is not the question. The question is whether it is priced well.
The problem is the exit. Banks are priced at tangible book value, and "No one will care about that while it's in private markets, but in public markets they would care about that."
He had an early look at a lower price and passed.
2. Why He Passed on Anduril
The same reasoning kept him out of the defense-technology company most private investors want to own.
"So, same similar reason why I did pass on Anduril."
The arithmetic he gives is the valuation ladder. Anduril went from $60 billion to talk of $100 billion, while Northrop Grumman, Lockheed Martin and Raytheon — with more revenue relative to their valuations, and growth of their own — sit at around $200 billion.
3. Raytheon and Defense Tech
Having named the incumbents, he said one of them is now on his own list.
"Looking at Raytheon very closely, by the way." The reason he gives is that they build the interceptors.
His explanation for weakness in the incumbent defense stocks is the new theme, not the business. They have been selling off in part because of the rise of defense technology as an investment story.
"But that's not going to impact near-term earnings at all when Pentagon is ramping up defense spending." He put the intended increase at 50%.
"It's not a zero-sum game. It's incremental growth to these new areas, but defense tech remains a hot, attractive theme, very exciting theme." That, he said, is the theme to focus on now — the way artificial-intelligence infrastructure was a few years ago.
4. Too Many Neoclouds
The counterpart to what he wants to own is a long list of what he does not.
"We don't need another neo cloud." His firm invested in CoreWeave, the first neocloud to go public. "There's too many neoclouds."
"We don't need another LLM model." He named Ilya Sutskever's Safe Superintelligence, which he expects to raise again after reaching unicorn status on its first round, and said he has no interest in the deal.
He was careful to separate the person from the trade. "I think he's an extraordinary individual." He said the same of Mira Murati, calling her an extraordinary talent, and credited Sutskever as seminal to the transformer paper.
His complaint is that the researchers are fundraising rather than researching, which is why he thinks the innovation is no longer compelling.
What is happening instead is scale: compute, data, scaling laws and reinforcement learning. "These are linear breakthroughs, but they're not explosive." He corrected himself mid-sentence to make the point sharper: "It's not a breakthrough is the main point."
His count of the field is about 60 large language model companies at unicorn valuations, most of them not making money, and he said open-weight models are now the thing.
5. Carney's Tax Move
The Canadian policy that prompted the second half of the episode is accelerated depreciation on investment, including oil and gas pipelines and artificial-intelligence infrastructure.
"Very good policy move. Excellent policy move. I approve this message Mark Carney."
He explained the mechanism for a general audience. "When you have the accelerated depreciation, it means that you get immediate year one expensing instead of amortizing the cost over time." Expensing the asset in year one lowers the tax bill, which makes it pro-investment.
His hesitation is whether tax policy can do the rest of the job. "Is it enough to have that policy? Will it help businesses? Yes." Whether it ignites what he called animal spirits is the open question.
He said Canada has neglected its energy infrastructure for a long time, with not enough midstream pipeline and, now, high energy costs.
6. Gas Into Compute
His worked example of the pipeline shortage is an investment his firm holds in Alberta, in a company he named as New Western Data.
The constraint is physical and regulatory at once. There are thousands of scattered wells with no pipeline to move the natural gas out, and rules that prohibit flaring it except in an emergency. Wells that cannot vent or ship their gas are shut in.
"These are productive wells with reserves that are economically profitable that if they were in the Lone Star State, the great state of Texas, they'd be making money, but they don't."
The workaround converts the stranded gas into a different product. "So, what these guys do is very clever. They metabolize the natural gas with ASICs. They turn it into data center compute."
The second leg is the oil. "And then because they've released the natural gas, which is commingled, dissolved into the oil, they can liberate the oil, put the oil on trucks and sell it at $40 a barrel."
He explained flaring by pointing at a gas stove and at the television series Landman, and stressed that the Canadian limits are very strict.
7. Canada's Missing Founders
The diligence trip behind that investment is what turned the policy question into a talent question.
"So much talent has left Canada." He reached out to contacts in Alberta and got the same answer about the people he was looking for: "They're in the Permian Basin. They're in Midland. They're not here anymore."
"So, people move to where the opportunities are." That mobility, he said, is part of what has driven American economic success, on cultural roots that go deep.
He attributed animal spirits to John Maynard Keynes and traced the underlying idea to Joseph Schumpeter, whom he called one of the most underrated economic philosophers and who coined the term creative destruction.
His counterweight is the Canadian university system, which he praised: he noted that Sutskever graduated from the University of Toronto, called the University of Waterloo one of the best engineering schools in the world, and said one of the best engineers he ever hired came out of it.
8. The Fed Oversteered
The macro section was set against a rate decision the next day, with the market pricing a 92% chance of an increase.
His historical claim is that the 2024 election was itself a monetary event. When Trump was elected he argued it was equivalent to roughly a 50 basis point rate cut, and at the same time he was arguing that the Fed should not be cutting, because the economy was strong, robust and hot.
"They oversteered. They oversteered. They should never have cut rates last year."
The consequence he points to is duration. "And now you got 5 years of inflation running above target", with the consumer price index at 3.5% to 3.6%.
His read on the corporate impact is mild. "And if rates go up, it's not going to change the fundamental earnings power of these companies. The CFOs are smart. They all refinanced, most of them."
9. The Small-Cap Debt Trade
The exception to that calm is the small-capitalization index, and it is where he spent the most time describing an opportunity he is not taking himself.
"The small firms in the IWM index, many of them have not. They have way too much debt. Five-year money's coming due and they can't refinance." The IWM is the exchange-traded fund that tracks the Russell 2000 small-company index.
His instruction to listeners was to go and find those companies and put together what he called a structured solution — rescue financing on attractive terms.
The precedent he cites is Magnetar, the credit firm founded by former Goldman Sachs partners, which provided a credit facility to CoreWeave and took equity warrants alongside it. "They essentially became venture investors. They made venture returns on this debt offering."
The appeal of a public-market target is the diligence already done for you. A listed company has a known asset, audited financial statements, and has already solved the compliance burden of being public.
"So, find those opportunities that Blackstone can't get after cuz they're too small. And everyone's just focused on the AI trade."
10. Go to Y Combinator
The last third of the episode is a career argument, prompted by a conversation with a University of Chicago sophomore named Zach, an Exeter alumnus who started the university's poker club with sponsorship from firms including Citadel and Susquehanna.
He praised poker as the closest game to real life for dealing with genuine uncertainty, because it is multiplayer and both bet size and position in the hand matter.
Asked what he thought about Polymarket and Kalshi, and about building a hedge fund around them, his answer was flat. "No. No, no, no, no, no. These are just fads and trends. You got to be ahead of the curve." He noted the two companies are raising at very large valuations.
On joining a long-short hedge fund instead, he was harsher, saying that idea made sense 30 years ago when Julian Robertson was starting the Tiger Cub lineage. "This is a stupid thing to do right now."
The alternative he endorsed was the friends starting Y Combinator companies. "That is the exact thing you should be doing is develop these entrepreneurial skills." The list he gave was solving problems, forming teams, developing a vision, getting people rowing in the same direction, executing, shipping and using technology.
He applied the same logic to his own children's education, next to the 529 accounts he still contributes to. "But like the best move is get into Y Combinator. That's it. That is the best education you can get."
"You're going to get an MBA degree and practical skills in a extremely short period of time." His advice does not stop at students: if you are in your thirties, do it anyway. "Get up and move to SF, lower your cost structure."
His fallback case is the network. If the company fails, the founder has built a network around the hub and will find something else — the point being to be in the right opportunity set.
He described the thought arriving in three stages: first encourage his children to be entrepreneurs with some risk capital and guidance; then build alongside them, which he allowed could make for an unusual family dynamic; then, finally, send them to Y Combinator for the peer group.
11. The 81x VC Fund
The episode closes on a venture fund he met five or six years ago whose returns he found hard to believe.
The multiple is 81 times, not fully realized, though he said the fund also has real distributions to paid-in capital.
The question he says an allocator has to ask is whether that is skill or a coin flip. With thousands of venture funds, one of them will flip heads ten times in a row, and the meeting is with that one. "You have to say, is it skill or is it luck?"
His answer, in this case, is that they rode two bubbles and timed the first exit. "They rode two bubbles. And they sold the first bubble at the top. The first bubble they did was crypto. They got into crypto, and in 2021, they knew they were in a bubble, and they sold."
He declined to name the fund out of respect, and said he may try to get them on a podcast.
Bonus Insights
The episode was recorded on the way to pick up a birthday cake. It was his youngest daughter's birthday, and he and his wife settled the cake question between them: his wife wanted tiramisu, he argued that tiramisu requires refined adult sensibilities, and strawberry shortcake won.
His own childhood benchmark is a fourth-grade surprise party with water guns and his first sleepover.
He credited Grok for his preparation for the Erebor segment.
Ahluwalia's bottom line is that the crowded trades — another neocloud, another model company, a private bank at a venture multiple — are where the money is going and the returns are not, while the opportunities sit in unglamorous places: interceptors, stranded Alberta gas, and small listed companies with debt they cannot refinance.
Products, Companies & Tools Mentioned
Wintermute and Galaxy Digital (The two firms that bought USDC at about 99.9 cents and redeemed it through the bank's fee-free offer)
Anduril (The deal he passed on, at $60B going to a talked-about $100B)
Raytheon, Northrop Grumman and Lockheed Martin (The incumbent primes at around $200B; he is looking at Raytheon closely because it builds the interceptors)
CoreWeave (The first neocloud to go public, which his firm invested in, and the borrower behind Magnetar's warrants)
Magnetar (The credit firm whose CoreWeave facility produced venture-style returns from a debt deal)
New Western Data (The Alberta company turning shut-in wells' natural gas into data-center compute and then selling the freed oil at $40 a barrel)
Polymarket and Kalshi (The prediction markets he told a student not to build a fund around)
Y Combinator (His answer to what a young person should do instead of joining a fund)
Citadel and Susquehanna (Sponsors of the University of Chicago poker club the student started)
Blackstone (Too large to chase the small-cap rescue financings he describes)
Watch the full episode:
If this was worth your time, send it to someone who has to have a view on this.
Get the latest market chatter as it happens: