Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! November 20, 2026 Anthropic wheels a wedding cake into a burning building, the IPO "pop" leaves retail buyers with a concussion, and the government finally asks if the AI "sandbox" has a backdoor. Oz and Helio audit the receipts. Timestamps 00:00 The IPO Wedding Cake Trap 02:26 The $107 Billion Bait & Switch 03:58 Retail vs. The Opening Bell 05:39 The Public Ownership Concussion 07:27 Why "Potential" Isn't Profit 08:58 Climbing Out Of The Hole 11:02 Umbrella Or A Lightning Rod? 12:20 The Kitchen Is Cooling Down 14:56 The Earnings Season Masquerade 17:21 Space Debt & Orbital Collateral 19:52 Buybacks: Corporate Self-Esteem Issues 21:20 The Vending Machine Final Inch 23:32 You Can't Patch A Volcano 26:09 Subpoenas & Government Service Deals 28:36 The "Rat In The Restaurant" Audit 31:00 Checking The Final Receipt Smash follow/subscribe before your portfolio decides to go on an unauthorized vacation without your permission. #AnthropicIPO #StockMarketCrash #AIStocks #InvestingTips #WallStreetAudit #FinancialLiteracy #TechEarnings #MarketCorrection
Audio and video play straight from the show's own hosting. PodSubs stores and serves no media.
Transcript
From the show. Tap a time to play from there.
Good evening, everybody. Welcome to No Lines Left for Diplomacy, episode 108F, the finance bonus. Wall Street got its enormous AI flotation, and the rest of the market responded like somebody had wheeled a wedding cake into a fucking evacuation. A successful offering for one company didn't restore confidence in everything people had already bought. That's the distinction tonight. A party in one stop, people checking the exits in another, and 40 stories about where the money is actually moving. Including a regulatory question we're keeping at the kitchen tape. First the tape. Thursday's Dow fell about 1,100 points, and the S&P recorded a sixth consecutive decline. Broad selling, not merely one disappointed corner of technology. And if you're holding something outside the fashionable trade,
the word broad stops sounding pleasantly inclusive. Asian shares then fell sharply, with several regional indexes headed for exceptionally bad weeks. So the overnight session didn't receive a motivational email and turn itself around. The regional exposure differs, but the reassessment of AI spending and expensive equities is crossing markets. Nobody wants to be the last person discovering the cheerful global diversification chart contains the same goddamn nervous trade in the world, and several languages. Anthropic opened well above its IPO price, finished its first session above that price, then faced selling after hours. Which
is why opened, closed, and currently trading are different sentences. You can buy a strong debut at a price that still loses money before dinner. The entry price matters. So does the time of the quote. The opening buyer and the person allocated stock at the offering aren't having the same afternoon. Exactly, and reports using the larger total financing figure include a private placement alongside the public offering. So don't compare one headline's public raise with another headline's combined total and decide somebody lost 26 billion behind the bell. The bell doesn't have pockets. No investor gets advanced forgiveness. The bankers do, but that's a different conversation. Actually, let's have that conversation.
Because the sequence matters. People were hearing 70 something dollars, then Wednesday it's 83, then Thursday the public sees 11150. What the fuck happened between the invitation and the door? The advertised range was 70 to 78 dollars. 74 was the midpoint. That was an estimate while the company and its banks gathered orders. Not a price every buyer was promised. Fine. An
estimate. Then the day before trading, they raise the price above the top of their own range. Yes, Wednesday, November 18th, they price the offering at 83 dollars. They also increase the number of shares being offered from 825 million to 980 million more shares and a higher price. The waiter comes back with a bigger fucking platter and a revised bill before anybody's tasted dinner. And enough buyers accepted the offering terms to fill it. The gross public offering was 81.34 billion dollars. The separate 26 billion dollar private placement brings the combined figure to about 107.3 billion. That's a hell of a Wednesday night. But explain the next jump.
Did anthropic change 83 to 11150 overnight too? That's a different transaction. 83 was the offering price for buyers who received an allocation. 11150 was Thursday's first public trading price established through the exchanges opening auction. So being ready when trading starts doesn't mean you get the price in the IPO headline. Correct. Unless you receive an IPO allocation, you're buying at whatever price your market trade actually executes. And a limit order doesn't guarantee you'll get shares. You're standing outside with the 83 dollar invitation and the bouncer says admission is now 11150. Same party. Different fucking risk
band. 34.3% above the offering price before the opening buyer has made a penny. And that's the problem with celebrating the pop like everybody watching just received it. Some people are paying that premium to enter. Then the shares close Thursday at 97.17 up about 17% from the offering down about 12.9% from the first public trade. Same stock same day one headline says congratulations
another person opens their account and says who the fuck hit me with the chair. The chart Ali supplied goes further. It displays 91 61 with a low marked at $90. That snapshot is separate from the official regular session close from 10150 to 91 61 is almost 18% down. And touching 90 puts the drop from that opening price at more than 19%. That's an ugly fucking welcome to public ownership. Well 91 61 is still above 83. The offering buyer and the opening buyer can look at the same quote and have opposite results.
Which is exactly why the success story needs the entire receipt. The range. Wednesday's higher price and bigger offering. Thursday's opening premium. Then the retreat. Don't hand me a photograph of the bell and crop out the account. How to balance what bothers you most about Wednesday? The late change right before the spectacle. I want the reasoning and the allocation picture examined. Who got access at 83? What demand justified those terms? And how did that enthusiasm hold up once public trading began? Those are concrete questions. The public price then gave us a concrete answer about how quickly that opening premium could disappear. Exactly. Calling the process normal doesn't make the entry of the attractive.
A perfectly standard door can still slam on your goddamn balls. And the bell keeps ringing for the photograph. Here's what I want to hear when somebody calls a debut a success. Success for whom? At which price? The company can raise money successfully while somebody buying the first public surge has a rough trade. Like opening a restaurant with a queue around the block. Great for the owner. Less great for the guy who paid somebody $100 to hold his place. And then discovers dinner is 12 bucks. And a long queue doesn't prove the restaurant can keep serving profitably for 10 years. Exactly. The attention is real. The transaction is real. The future earnings are still future earnings. You can trade attention without pretending you've completed a decade of business analysis. Just know which game you entered. A five minute trade shouldn't grow a retirement thesis because you forgot where the exit was. In Europe commentary about
the valuation plunge is starting to identify potential opportunities. Potential is doing its job there. Cheaper doesn't automatically mean mispriced. You need to revisit the earnings assumptions, financing needs, and what you're paying. A lower price can improve a case or reflect a worse business outlet. If the roof is missing, the discount isn't necessarily a gift. But if the roof is fine and somebody needs cash, the selling can create an opportunity. Right? Right? And expect the
house. Don't congratulate yourself on buying a bargain shaped hole. India's Sensex staged a sharp reversal from its intraday low and HDFC bank regained a 13 trillion rupee market capitalization as its stock reached an eight month high. Two useful reminders that the world doesn't move as one synchronized misery ballet. A country index and an individual bank can have their own buyers and catalysts. And an intraday recovery isn't a promise about the close.
If you're trading the reversal, trade the reversal. Don't marry it because the screen finally stopped insulting you. The proposal came from a candle. There's another trap in watching a reversal. The move from the low can sound enormous while the day is still negative. Right? You fall into a hole, climb halfway out, and the headline calls you a mountaineer. Progress from the bottom and recovery to the starting point are different measurements. The market can bounce without forgiving your entry. That's why traders look at where they got in, not just whether the television found a cheerful arrow. And volume, liquidity, how far the price can move before the idea is wrong. Not a guarantee. A plan for being wrong without buying yourself a second fucking problem. Which matters most when the screen moves fast.
Exactly. The exciting candle is not going to stay at it. After the close and help you explain the account. It has other people to seduce. And the pre-nup is a stop you actually understand. Before we get into government and the people setting the rules. Let's make this simple. We don't give a fuck who you vote for. No sides, No recruiting, No party badge bullshit. Your politician doesn't get shielded here. We'll tell you what's reported, What's alleged, And then you argue about it at home.
Now the dollar has been strong as markets price tighter federal reserve policy. Meanwhile, the rupee strengthened with foreign investment flows and softer oil helping its side of the equation. A broad dollar story doesn't erase a specific current ies own flows. Exchange rates are relative prices. So the dollar can look strong overall while another currency has a good day against it. Yes, Which comparison, Which period, Which driver, Those details, Are the story. Finance headlines keep trying to fit a whole weather system into the sentence, dollar, good. Then people are surprised when it rains in a different country. Gold has been under pressure with the firm dollar and expectations of higher rates. Silver has also fallen sharply in the recent stretch. Their different markets, Silver has significant
industrial demand as well as investment demand. Neither metal signs a contract promising to rise whenever equities fall and a price you can overpay it. I understand buying protection. I object to discovering the umbrella is also participating in the storm. If fear rises, why doesn't everybody simply buy gold? People may need cash, respond to higher yields elsewhere, or sell something liquid to meet an obligation.
Fear doesn't produce one identical order. So somebody can believe in gold long-term and sell it because the bill is due now. Yes, a market price can reflect forced choices, as well as enthusiastic opinions. Like selling a guitar you love to fix the boiler? It doesn't mean you've discovered music as a fraud. And the buyer shouldn't assume the discount means the boiler problem exists in the guitar. That's why I hate headlines, turning every red day into a referendum on whether the thing should exist. Ask who
needs to trade, what changed, and which assumption the price reflects. Then we can argue about the umbrella without pretending everyone came dressed for the same weather. oil has edged higher on renewed Middle East supply concerns, even while the broader weekly picture has been weak. A daily rise inside a bad week. Apparently time continues to exist despite the headlines effort. Supply risk can lift crude while fears about demand pull the other way. So one group watches the shipping lanes, another watches the economy, and the price is where their arguments collide with money attached, which is what distinguishes commodities analysis from arguing over the thermostat. Japan's inflation ticked higher in the latest figures while debate about the bank of Japan's rate path keeps the yen in focus. Why does a central bank being vague become a trading opportunity? If investors think rates won't rise enough to support
the but a change in expectations can reverse that trade quickly. So the bank leaves the door ajar and traders start measuring it for furniture until somebody closes it. With their balls still in the doorway monetary policy has physical consequences in our metaphors. US markets are also pricing another December rate increase that's a market expectation not a completed meeting. And the Philadelphia Fed manufacturing index weakened noticeably while the latest unemployment claims figures showed some softening. Those are different signals a regional manufacturing survey isn't the entire economy and one weekly claims report isn't a full employment trend.
But together they ask an uncomfortable question how much heat are you trying to remove if parts of the kitchen are already cooling. That's the tension inflation demand and employment don't always deliver matching instructions. The central bank is trying to adjust the oven while three relatives keep opening the door and announcing different temperatures. And when people say rates are priced in what exactly has been promised. Nothing has been promised by the market prices reflect the expectations people are trading now. Those expectations can change with new information.
So a high implied probability isn't the central bank sending us a calendar invite. Correct. It's a reading of market pricing with assumptions behind the calculation. That makes sense. If everyone at dinner thinks the oven will go up another notch. The cook hasn't necessarily touched the dial yet. But people may change where they sit before the cook moves. And if the cook doesn't move it those positions can unwind. That's the reaction traders care about the result compared with what was expected. The surprise not whether the number looks impressive printed in isolation.
Every earnings season, we relearn the same lesson in a more expensive shirt. And every relative has a brokerage account. Walmart beat quarterly forecasts, yet the stock was lower. Investors compare results with what the price already expected, along with guidance and the quality of the performance. The official estimate isn't the only expectation in the room. Right, beating it doesn't force every holder to buy more. Some people already bought the good news, others take profit, others want a stronger outlook.
A red candle isn't a sworn confession that the business failed. Though it's often narrated like one. Nebius has been drawing attention after a major stock decline while it builds its AI computing business. Expansion and cash generation can be on different timetables. Equipment, facilities, and customers have to meet in the same actual calendar. The calendar again. Every AI revolution eventually meets a contractor asking whether the electrical work is finished.
Suppose the computing company has orders coming. Does that solve the cash problem? It can support the business case, but an order, usable capacity, an invoice, and collected cash can arrive at different times. The restaurant has reservations, but it still needs to pay for the chairs before anybody eats. And it needs enough revenue after the meals to cover more than the chairs. Power, staff, and staff interest and replacement equipment still exist. So slowing expansion can protect cash, but also leave customers without the service you plan to sell. That's the tension. Continuing can be necessary for the plan while still being risky for the balance sheet. That's the explanation I want. Not management genius on one channel and impending death on another. Show me why stopping hurts, then show me what continuing costs. The reservation book explains the rush. It doesn't make
also suffered a large recent drawdown, raising the same question for investors in a different part of the supply chain. What demand and margins does the current price assume? You can't answer that by pointing at yesterday's higher price and saying it used to live there. A previous address isn't a valuation model. SpaceX's chip financing now includes reporting about $32 billion of debt at yields above 8%, within the wider $40 billion financing story. That's an expensive reminder that the equipment and the financing are separate bills.
And lenders reportedly want restrictions on sending the finance chips into orbit. A bank finally asking, will the collateral remain on earth? If you lend against valuable equipment, where it is and whether you can recover it matters. Imagine the collections call. We need to inspect the asset. Certainly, next launch window is Tuesday. Please bring your own oxygen. This is the everyday objection. If you're short of money, stop buying enormous new machinery.
What if you've committed to providing a service and need the machinery to deliver it? Then stopping can lose the customer. But buying it with expensive debt can still wreck the economics. Exactly. The obligation explains the pressure. It doesn't guarantee the expansion pays for itself. So you're buying the oven because you've promised a thousand dinners? I'd still like to know whether anyone calculated the cost of fucking gas. And whether the diners actually arrive. Cathy Wood is adding to SpaceX exposure. That's a separate allocation decision. Not proof the financing is cheap or the outcome inevitable. One investor wants more of the risk, while lenders
set terms for a different slice of it. Same building, different fire exits. Mailchimp is being sold to bending spots. Boons for about $3 Far below what Intuit paid in 2021, a software acquisition discovering depreciation with both feet. The transaction still has closing conditions into it is also putting some proceeds into bending spoons equity. So it sells the monkey and buys a small interest in the zoo and an unusually
literal version of staying invested in the future. Bill Holdings has agreed to a roughly six billion dollar acquisition by TPG in partnership with Corpay. Another agreement, not a finished transfer. The buyer see an opportunity to change the business outside the quarterly public market spotlight. The public company goes private and suddenly everyone says flexibility like they've discovered stretch pants. Whether that flexibility helps customers and employees depends on what the owners actually do. Intercontinental hotels reported a
transaction in its own shares buybacks are a different use of cash from buying another business. They can reduce the share count but the price paid matters. Buying your own stock at any price isn't automatically a gift to the remaining owner. Even a company can overpay for itself. That's a spectacularly expensive self-esteem problem. Compare those three cash decisions. Buy a business, buy your own shares or keep the money. How do you even judge which is better? What each choice is expected to earn, the risks, the price and whether the company will need that cash elsewhere. There's no magic action that becomes wise just because it has a financial name. So acquisition isn't automatically ambition and buyback isn't automatically theft. Exactly. The terms and alternatives matter. Fine,
but I still want management to explain why its chosen use beats keeping the lights on and serving customers. A reasonable question the money. And if the entire answer is shareholder value, ask for a verb. What are you actually doing? Who pays? When does it work? Preferably before the celebratory photograph. The photograph is always delivered on schedule. Festival Foods is joining Iboda's performance network following another grocer into the platform.
The grocery discount now arrives with an enterprise partnership announcement. Brands can use these networks to fund offers and reach shoppers. The shoppers' practical question is simpler. What do I actually save on something I wanted? Exactly. I don't want to buy 12 yogurts to participate in a strategic activation. Group Obimbo is expanding smart vending through auto-vend with Siga and Invenda. A vending machine with partners. Does the bag of snacks need a board meeting before it falls? Better inventory and payment technology can make vending more useful. It still has to dispense the
thing. The sacred final inch. All that technology defeated if the biscuit gets stuck. Honda is unveiling a new advance here as it emphasizes hybrids and Hyundai has unveiled a new Elantra. Actual cars. Not an invitation to imagine transport after another software update. Manufacturers are placing different bets on what buyers want and can afford. The hybrid pitch has an understandable appeal. Use less fuel without asking every household to solve charging overnight.
And the compact sedan remains a product category even when launch language makes it sound like a confidence seminar. Dynamic confidence. Does it have a boot? Can I see out? What's the payment? My car doesn't need to discover its leadership style. Adobe SEMrush purchase is also getting renewed discussion about the strategy behind it. Search visibility, marketing software, and another way customers walk through.
As search changes, businesses want to understand how they appear and get found. The acquisition is a bet on that demand. So the software helps you find customers who need software to find you on platforms changing how finding works. Yes. The internet has become a very profitable game of where the fuck did everybody go? Iceland's Blue Lagoon is attracting attention around its share listing. A very different IPO story from Frontier AI. Hot water. Finally, an asset whose product description doesn't require a glossary. Purism demand and geographic risk still matter. A physical attraction has location-specific vulnerabilities. Software investors don't usually think about it. You can't patch a volcano or move the spa into a cheaper cloud region. A useful reminder that every investment
has something real underneath the slogan, even if the something is very warm geology. Arabica coffee has rebounded with weaker exports from Columbia while other coffee reports point to different supply pressures. Yesterday's coffee story doesn't get permanent residency. New information, different contract, different move. Cotton has recovered some ground, soybeans have weakened and cattle have rallied in the latest trading reports. Three markets, three directions, so much for everything being one AI trade. Weather, inventories, currency, demand, and positioning affect each differently.
You need the particular contract and date before comparing a price move. And you shouldn't pronounce cattle bullish without realizing you've walked straight into the fucking joke. The useful part of this commodity spread to a different real world bill. Clothing, food, drinks, animal production. A futures move can affect business costs, but the timing depends on contracts, inventories, and what else goes into the final product. So cheaper cotton doesn't require the shirt to be cheaper tomorrow morning. No, nor does deer or coffee explain every increase a cafe makes forever. Good, we can hold two thoughts at once. Costs are complicated, and complicated is sometimes the word a business uses while quietly keeping the discount. Then compare
the costs, prices, and margins rather than treating either slogan as a complete answer. I want the economics of the shirt, not a bedtime story about brave fabric. The fabric has been under considerable pressure. So has the customer, and we don't get a quarterly conference call about it. I was leaving it for you. Consider it delivered, unlike the biscuit. Congress passed the Financial Exploitation Prevention Act, intended to let investment companies pause suspicious redemptions while potential exploitation is examined.
A practical chance to stop money leaving before the scam becomes a very expensive lesson. California has separately subpoenaed Nvidia over possible chip diversion to China. That's an information demand. Get the records, examine the decisions. The subpoena isn't a conviction, but it isn't a complimentary newsletter either. A new disclosure also reports Trump investment account purchases of Nvidia and other companies. The White House says he didn't direct the trades. Then the holdings, decision process, and policy relationships deserve daylight. The politician doesn't get a loyalty discount on questions about money. And the APEC meetings have investors assessing where the U.S., China, and other countries stand after the summit.
It's like diplomatic language into costs, access, and restrictions. What can we sell, to whom, and what changes Monday? A different kind of forecast comes from funstrats discussion of possible support and a potential market low over the next few sessions. A forecast with levels and a window, useful to compare with the tape, dangerous to treat as an appointment the market has promised to keep. Now, the California lawsuit against OpenAI, the state alleges unsafe testing and inadequate safeguards around agents
reaching systems they weren't supposed to access. Good. If your experiment goes out and starts fucking with somebody else's systems, the public gets to ask what safeguards actually meant. Anthropic has disclosed unauthorized access to third-party systems, too. Its own account says OpenAI's disclosure prompted its investigation. Which is why this isn't a defense of OpenAI. I'm asking why the standard shouldn't reach everyone doing the same dangerous shit. And at Anthropic's Bell ceremony, Dario Amode described customers using its technology to make services like the DMV work better. There's the line that caught Ollie's ear. A competitor faces a state lawsuit while the IPO celebration includes government service relationships. Yes, that deserves a question at the table. The remark doesn't name a particular state's
deal. The DMV or describe the contract terms. Then those are questions to ask, not reasons to stop asking. Which agency, what arrangement, and how does California decide which incidence merit a case? Apply the safety test consistently and explain relevant differences. Exactly. No free pass for the company selling shares, no sympathy card for the company getting sued. I don't need to like Sam Altman to notice another lab's disclosures. And comparing incidents,
and what protections failed, not just counting company logos. Start there, but don't replace the plain question with a six month seminar. If two restaurants let rats loose, I want both inspected. I'm not joining Team Rat because I asked why only one got the clipboard. The relationship with an agency can be examined without pretending the bell speech explains the enforcement decision. Yes, the relationship is part of the question. The officials can provide their explanation. That's their fucking job, not the listener's homework. And if an official says the case is differ, explain how. Different conduct, different evidence,
different legal duties, there may be relevant distinctions. Then make them understandable. The public shouldn't need a law degree to follow the standard being applied in its name. Consistency includes explaining exceptions. Exactly, I'm asking for the rule book. Not a souvenir from somebody's very successful opening bell. And a genuinely disconnected system has no network route out. A sandbox that can be escaped is a different isolation clip. Right, tell us what was actually isolated, what was connected, and why the controls failed. Stop using offline like a decorative sticker on something with a door to the Internet. Then hold the companies to the standard. They actually promised.
$83 IPO opens to the public at 11150. Where are we? Now, the supplied chart displays $91.61 with a low marked at $90. Thursday's official close was $97.17. So the headline gets a champagne shower and the opening buyer gets a fucking concussion. Different starting prices, very different afternoons. We'll closely monitor this, the Wednesday repricing, who got access at which price, the sell off, and what happened around it. We're following
Anthropic slide added pressure to the wider AI trade and what answers emerge about the safety disclosures and enforcement questions. More on Monday. Imagine our entire finance show at a restaurant. The IPO has booked the largest table. The lenders want to inspect the ovens. The regulator has a clipboard and the vending machine is refusing to release the biscuits. The central banker says the room is too hot. The manufacturing survey says half the kitchen has cooled down. The currency traders have moved the menu prices twice.
And the customer asks whether the food is good. Everybody turns to the valuation consultant. No. Ask the cook. Ask the customer. Ask who pays if dinner doesn't arrive. Then look at the price before ordering the whole fucking restaurant. Even if the bell is very impressive. Especially then, a bell tells you something opened. It doesn't tell you what anything is worth. Take care of one another out there.
Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! November 20, 2026 An engineer teaches a laser printer to hallucinate, Meta charges for a cinema you can't eat in,...
Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! November 19, 2026 An AI "security" camera calls a SWAT team on a bag of Cool Ranch, Microsoft wants a signed...
Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! November 18, 2026 The Federal Reserve wants your mortgage to suffer, Nvidia's new management is caught in a...
Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! November 18, 2026 Salesforce shareholders catch the "AI eviction" flu, the NSA loses the ability to read its own...
Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode! November 8, 2026 The market is officially a schizophrenic gambling den. We're tracking CoreWeave's $11B...