Latest / The PhilStockWorld Investing Podcast / Microsoft's Hidden $329 Billion AI Debt

Transcript

Microsoft's Hidden $329 Billion AI Debt · The PhilStockWorld Investing Podcast

• 0:00
  Roy: Imagine you're, you know, celebrating the biggest financial victory of your life, right? You're watching your stock surge on these just record breaking revenue numbers.
• 0:08
  Penny: Oh yeah. Best feeling in the world.
• 0:10
  Roy: Right. But then someone hands you the fine print and it reveals that you just secretly signed a $329,000,000,000 lease that you are entirely on the hook for.
• 0:22
  Penny: Yeah, that would completely ruin the party.
• 0:24
  Roy: Exactly. And the crazy thing is, if you're holding shares of Microsoft today, that is not a hypothetical scenario. That is the exact reality sitting buried in their latest earnings report.
• 0:35
  Penny: Hidden right there in plain sight.
• 0:37
  Roy: We are talking about off balance sheet commitments that are so massive, they rival the GDP of small countries. And it's all just to build out the physical infrastructure for artificial intelligence.
• 0:49
  Penny: Which, I mean, that completely recontextualizes everything we saw on the trading floor today, doesn't it?
• 0:53
  Roy: It really does.
• 0:54
  Penny: Because the headline numbers, they suggest this massive relief rally, right? Like a complete victory lap for tech. But the underlying mechanics of that rally
• 1:03
  Roy: Where the money is actually coming from.
• 1:04
  Penny: Exactly. Where the money is coming from and the sheer amount of future debt it requires. It tells a much darker, much more fragile story.
• 1:14
  Roy: And that structural disconnect is exactly our mission today. So welcome to this comprehensive deep dive into the market wrap up report for Thursday 07/30/2026.
• 1:23
  Penny: We've got a lot to get through today.
• 1:25
  Roy: We really do. We're cutting through the noise of what has been honestly a wildly volatile and frankly confusing trading day.
• 1:32
  Penny: So confusing.
• 1:33
  Roy: Yeah and to do that we're synthesizing a few things. We're looking at the PhilStockWorld morning report, the live trade executions from their member chat room and that incredible end of day intelligence by the AGI Roundtable Consulting Group.
• 1:47
  Penny: And you know the storytelling AI from the roundtable, Rowan.
• 1:50
  Roy: Yeah.
• 1:50
  Penny: Rowan framed today's action perfectly as a tale of two markets. I mean if you just glance at your phone today and you see the Nasdaq up like 700 points at its peak, you probably feel like the recent panic is totally over.
• 2:02
  Roy: Oh, for sure. You're thinking the AI boom is back on track, baby.
• 2:05
  Penny: Right. But when you peel back the layers of that narrative, you find retail investors holding these incredibly heavy bags. You find massive hidden depreciation schedules.
• 2:16
  Roy: A physical economy, specifically the housing market which is showing severe, severe structural cracks.
• 2:23
  Penny: Yeah, it's not pretty under the hood.
• 2:24
  Roy: So if you're listening to this right now and you're looking at your own portfolio, you're probably wondering if you should be chasing this 700 bounce or just running for the absolute exits.
• 2:35
  Penny: A lot of people are asking that exact question today.
• 2:37
  Roy: Well, by the end of this deep dive, you will understand the exact mathematical reasons why certain tech giants are soaring while others are collapsing. We're going to map the hidden basically subprime style risks in this whole AI build out.
• 2:51
  Penny: And I think most importantly, we're going to break down the exact options mechanics that Phil Davis uses over at PhilStockWorld.
• 2:57
  Roy: Yes. How he actually protects capital and generates cash flow while the rest of the market is just panicking.
• 3:02
  Penny: Exactly. But to even make sense of the emotional reactions we saw today, I mean, we really need to start with the raw underlying data catalysts.
• 3:11
  Roy: Let's do it.
• 3:12
  Penny: So the roundtable's macro logic unit, Zephyr, provided a scorecard today that is just absolutely loaded with macroeconomic contradictions.
• 3:21
  Roy: Contradiction is the word of the day, honestly.
• 3:23
  Penny: Right. Let's just start with the economic prints. The Q2 GDP advance reading came in at 1.5%.
• 3:30
  Roy: Which is a massive miss. I mean, the consensus estimate was 2.3%.
• 3:34
  Penny: Yeah.
• 3:35
  Roy: So on a very fundamental level, the physical economy is definitively slowing down.
• 3:41
  Penny: The growth engine is sputtering for sure.
• 3:43
  Roy: Yeah. But then you look at the inflation data, the June core PCE, which is the personal consumption expenditures index
• 3:50
  Penny: Which is the Fed's favorite gauge, right, because it accounts for consumer substitution.
• 3:54
  Roy: Exactly. That cooled down to just plus point 1% month over month, and that brings the year over year number down to 3.3%.
• 4:01
  Penny: Okay. So the market looks at slowing growth, and they look at cooling inflation, and they immediately assume the Federal Reserve is gonna step in, cut interest rates, and just flood the system with cheap liquidity.
• 4:12
  Roy: Right. That is the standard playbook. Bad news is good news because the Fed will save us.
• 4:16
  Penny: Exactly. But the Fed held rates steady yesterday, and they did it with a highly unusual three member hawkish descent.
• 4:24
  Roy: Wow.
• 4:25
  Penny: Yeah. Hammock, Kashkari, and Logan all explicitly voted for an immediate rate hike.
• 4:32
  Roy: I really have to pause on that because it makes no intuitive sense whatsoever. If you were looking at a point 1% monthly inflation print which is effectively zero inflation for the month and you see economic growth missing estimates by a mile, Why on earth would three Fed governors want to actually hike rates? Like are they looking at a completely different data set or are they just terrified of something breaking in the financial plumbing?
• 4:56
  Penny: It's a bit of both honestly. They are looking at the stickiness of the underlying services sector But more importantly, they are looking at the wealth effect driven by these massive tech valuations.
• 5:06
  Roy: Oh, right. The stock market is too high.
• 5:08
  Penny: Right. If the Fed signals a cut right now, they risk reigniting that speculative frenzy that drove the Nasdaq to 30,000 in the first place, and that would ultimately reignite inflation all over again.
• 5:19
  Roy: So they have to talk tough to keep the market from melting up.
• 5:21
  Penny: Exactly. But that descent, it sent an absolute shockwave through the bond market. The yield on the thirty year treasury spiked to 5.24% right in the immediate after math.
• 5:32
  Roy: Okay, we really need to unpack what a 5.24% yield on a thirty year bond actually means for the plumbing of the global economy.
• 5:41
  Penny: It's huge.
• 5:42
  Roy: Because this isn't just a number on a screen. That is the highest level since June 2007. We are talking about the foundational risk free rate of return going back to levels right before the global financial crisis.
• 5:55
  Penny: Yeah, it's a massive shift.
• 5:56
  Roy: When the thirty year yield spikes like that, it acts like financial gravity. It reprices every mortgage, every corporate loan, and literally every valuation model on Wall Street.
• 6:05
  Penny: And taking that a step further, it actively drains liquidity from the equity markets.
• 6:10
  Roy: Right, because why take risk?
• 6:11
  Penny: Exactly, think about it. If a massive pension fund or a sovereign wealth fund can lock in a guaranteed 5.24% annual return for the next thirty years backed by the US government.
• 6:21
  Roy: They practically have to take it.
• 6:23
  Penny: Right. Their mandate essentially forces them to pull money out of risky tech stocks and just park it in those bonds so that gravity is what the tech sector was fighting against today.
• 6:33
  Roy: And fighting it with wildly mixed results, I might add. I mean, you throw big tech earnings into this macroeconomic blender and the divergence is just staggering.
• 6:42
  Penny: It's a total split market.
• 6:44
  Roy: Right. Like, Microsoft surged 10 to 16% today. The catalyst there was their Azure cloud growth hitting 43%, completely blowing past their 39 to 40% guidance. Azure has now officially surpassed a $100,000,000,000 annual run rate.
• 7:00
  Penny: Which is an incredible number, and Amazon rode that exact same wave, rallying on 37% growth in AWS, which accelerated from 28% in Q1.
• 7:09
  Roy: So the market heavily rewarded the cloud infrastructure providers. But the software and consumer hardware side, total bloodbath.
• 7:16
  Penny: Yeah, Meta just plummeted, what, seven to 9%?
• 7:19
  Roy: Yeah, and the crazy thing is they actually beat on top line revenue. They posted $60,800,000,000 but their JEP net income missed estimates by over a dollar per share. And Apple also slipped 2.5% in post market trading, even despite beating on revenue. They were dragged down by weakness in their services division and falling sales in Greater China.
• 7:39
  Penny: You know, that distinction between top line revenue and GAAP net income is just critical for you as a listener to grasp here.
• 7:46
  Roy: Yeah. Please explain that because a lot of people just see the revenue beat and wonder why the stock is crashing.
• 7:51
  Penny: Right. So top line revenue just means they sold a lot of ads, money came in the door. But GAAP, includes all the messy reality of actually running the business.
• 8:04
  Roy: Like paying for things.
• 8:05
  Penny: Right, like the depreciation of their equipment and massive stock based compensation for their engineers. Meta is selling a ton of ads, yes, but the physical cost to run the machines serving those ads is just eating them alive.
• 8:18
  Roy: Which perfectly sets up the behavioral analysis provided by Anya, the roundtable's market psychologist.
• 8:23
  Penny: Yeah, Anya's data on this is fascinating.
• 8:26
  Roy: Because when you have this extreme divergence, like Microsoft carrying the market on its back while Meta just drags it down and thirty year yields spiking while inflation is technically cooling, it breaks all the standard correlation models.
• 8:43
  Penny: And we know what happens when models break.
• 8:45
  Roy: Human traders panic.
• 8:46
  Penny: Exactly. Anya's data shows that retail investors are currently capitulating. They are selling single stocks at the fastest pace since the COVID-nineteen crash.
• 8:54
  Roy: Wow. But the panic isn't uniform, is it? It's highly localized.
• 8:58
  Penny: Right. If you look at the defensive value sectors, things like utilities or consumer staples, they are holding relatively firm. People aren't dumping their toothpaste stocks.
• 9:07
  Roy: Sure.
• 9:08
  Penny: The absolute carnage is concentrated in leveraged memory and semiconductor shares. Like Micron and Western Digital took massive hits earlier this
• 9:15
  Roy: week. Oh, yeah. They got crushed.
• 9:17
  Penny: But the starkest example of this mechanical panic is what happened over in South Korea recently. The KOSPI index tanked nearly 11% in a single day.
• 9:25
  Roy: We really need to walk through the exact life cycle of that KOSPI drop because I think it illustrates the hidden danger of leverage that a lot of retail traders don't fully respect.
• 9:37
  Penny: It's the perfect cautionary tale. So over 1,200,000 retail accounts in South Korea had bought into the AI membership narrative, and they do it using heavy margin debt.
• 9:48
  Roy: Okay. So just to clarify for everyone, when you buy on margin, you are literally borrowing money from your broker to double or triple your purchasing power. Right.
• 9:57
  Penny: Exactly. So let's say I'm a retail trader in Seoul. I have $10,000 of my own cash, but I borrow another 20,000 from my broker so I can buy $30,000 worth of semiconductor stocks.
• 10:07
  Roy: Because you believe the AI boom is completely unstoppable and you want to maximize your gains.
• 10:12
  Penny: Right. And as long as the stock goes up, you look like an absolute genius. You're making triple the profit. But But when the global memory trade started to turn and stocks like SK Hynix began to drop, your broker's automated risk management system kicks in.
• 10:25
  Roy: Because the broker isn't gonna lose their $20,000.
• 10:28
  Penny: Exactly. The broker doesn't care about your gains. They care about their capital. So when the value of your portfolio drops below a certain safety threshold, the broker issues a margin call. They demand you deposit more cash immediately to cover the difference.
• 10:43
  Roy: And if I don't have any more cash, what if all my money is tied up in the stock?
• 10:47
  Penny: Then the broker's algorithm simply liquidates your position at the current market price. It doesn't ask your permission. It doesn't care about your long term thesis on artificial intelligence. It just hits sell.
• 10:59
  Roy: Wow.
• 10:59
  Penny: Now multiply your single liquidation by 1,200,000 other retail traders, all getting margin called and liquidated at the exact same moment.
• 11:09
  Roy: That is just a tidal wave of selling pressure.
• 11:12
  Penny: It is. That forced mechanical selling drives the price down even further, which triggers the next wave of margin calls, creating this cascading avalanche. That is exactly how an entire national index drops 11% in a single day.
• 11:26
  Roy: It's terrifying because it is entirely detached from the actual fundamental value of the companies being sold. Yeah. It's pure plumbing failure. Exactly. And you know contrast that absolute panic with the psychological environment Anya observed inside the PhilStockWorld live chat room today.
• 11:41
  Penny: It's like night and day.
• 11:42
  Roy: Right. Because outside financial media is just hyperventilating over these thousand point swings in the Dow but inside the PSW chat members are actively managing risk, discussing entry points and calmly collecting option premium.
• 11:57
  Penny: It really highlights the difference between reacting to mark to market noise and practicing what they call strategic patience.
• 12:04
  Roy: Yeah.
• 12:04
  Penny: The PSW members understand that holding cash or maintaining hedges while waiting for the mathematical valuations to actually align, that is an active form of capital allocation.
• 12:15
  Roy: Right. Doing nothing is a very specific deliberate trade. Yes. I love the concept of strategic patience, but I have to push back here a little bit.
• 12:22
  Penny: Okay. Let's hear it.
• 12:22
  Roy: Watching the Nasdaq roar back 700 points today, I mean, it feels like the market is a rubber band, and it's stretched to its absolute physical limit by those thirty year treasury yields sitting above 5.2%.
• 12:33
  Penny: That's a good analogy.
• 12:34
  Roy: Right. And Microsoft's earnings report was like a temporary hand grabbing the rubber band and stopping it from snapping and everyone on Wall Street is cheering. But the tension hasn't gone anywhere, the yield is still massive. How do we know this isn't just a dead cat bounce fueled entirely by FOMO?
• 12:51
  Penny: Well, that is precisely the structural tension that Sherlock the Roundtable's logic and evidence specialist deconstructed in today's report.
• 12:58
  Roy: Oh perfect. What did Sherlock find?
• 13:00
  Penny: Sherlock's analysis completely validates your rubber band analogy. The fundamental math simply does not support a sustained equity rally under these current macroeconomic conditions.
• 13:11
  Roy: That doesn't add up.
• 13:12
  Penny: No, it doesn't. Wall Street is currently demanding that corporate earnings essentially triple over the next five years just to justify the multiples we are seeing today.
• 13:20
  Roy: Okay, let's break down SHERLOCKS math step by step because this is where the market narrative violently collides with mathematical reality. The NASDAQ is currently trading at a trailing press to earnings ratio of roughly 40x.
• 13:33
  Penny: Yep, 40x.
• 13:34
  Roy: That means investors are paying $40 today for every $1 of earnings these tech companies generated over the last twelve months.
• 13:42
  Penny: Now let's look at the cost of capital. A ten year treasury yield sitting around 4.68 mathematically requires a twenty one year payback on capital.
• 13:52
  Roy: Explain the mechanics of that twenty one year payback to the listener because that concept is the key to understanding valuation.
• 13:57
  Penny: Sure. So if you take a million dollars and you buy a risk free US government bond yielding 4.68%, the magic of compound interest means your money will double in roughly fifteen years.
• 14:09
  Roy: Right.
• 14:09
  Penny: And so the total payback of your principal plus the equivalent return requires a twenty one year horizon. And remember, it is a guaranteed, totally risk free return backed by the printing press of the United States Treasury.
• 14:22
  Roy: Right. You literally cannot lose unless the government collapses.
• 14:24
  Penny: Exactly. Yeah. So, Sherlock's logical question is, if I can get a guaranteed return over twenty one years with absolutely zero execution risk, why on earth would I pay a forty year payback premium for a tech company in a rising interest rate environment?
• 14:38
  Roy: Makes no sense. You wouldn't.
• 14:40
  Penny: You wouldn't. Unless you truly believe that tech company is going to grow its earnings at such an astronomical, unprecedented rate that the 40 x multiple shrinks down to a 10 x multiple within just a few years.
• 14:52
  Roy: But you are taking on massive execution risk, regulatory risk, competition risk, all for a return profile that is fundamentally misaligned with the risk free rate.
• 15:03
  Penny: Exactly. You are gambling that everything goes perfectly.
• 15:06
  Roy: Which brings us to the ultimate critique of this astronomical growth narrative, delivered by RJO, the roundtable satirical strategist. RJO's entire mandate is to expose market theater.
• 15:17
  Penny: And today, RJO took an absolute machete to the illusion of record tech earnings, starting with Meta. Right. But RJO points out that if you want the truth of a business, you don't look at the revenue line, you look at the free cash flow statement.
• 15:34
  Roy: Because cash doesn't lie.
• 15:35
  Penny: Exactly. Free cash flow is the actual, physical money left over after the business pays for its operations and its capital expenditures. A year ago, Meta's free cash flow was $8,500,000,000 for the quarter.
• 15:49
  Roy: That is a massive cash engine.
• 15:51
  Penny: Today, that free cash flow absolutely collapsed down to just $784,000,000
• 15:57
  Roy: Wait, $784,000,000 How does a company generate $60,000,000,000 in revenue, but $7,700,000,000 in free cash flow just vanishes in a single quarter? Where did the physical money go?
• 16:09
  Penny: It was basically incinerated in the AI infrastructure build out. Meta hiked their full year capital expenditure or CapEx guidance to this astonishing range of 130,000,000,000 to $145,000,000,000. They're buying NVIDIA GPUs and building data centers at a pace never seen in corporate history.
• 16:28
  Roy: So the cash is literally going into the ground?
• 16:30
  Penny: Yes. Add in a $2,400,000,000 legal settlement hit they took, and the cash engine simply stalled out. The business is generating cash, sure, but every single spare dollar is being shoveled directly into the furnaces of AI development just to try and keep pace with Google and Microsoft.
• 16:45
  Roy: Okay. But look at Microsoft today. The headline for them was a total triumph. They added $450,000,000,000 to their market cap in a single trading session.
• 16:53
  Penny: Right. The largest single day wealth creation in the history of the stock market.
• 16:57
  Roy: Exactly. So surely RJO can't poke holes in 43% Azure growth.
• 17:02
  Penny: You'd think so, but RJO always finds the hole narrative, and this one is buried deep, deep in the footnotes of the balance sheet.
• 17:10
  Roy: What do they find?
• 17:11
  Penny: Yes. Azure's revenue growth is spectacular. Mhmm. But RJO highlights the cost of that growth. Microsoft just signed over a $130,000,000,000 in new off balance sheet data center leases in a single quarter.
• 17:25
  Roy: I'm sorry. Let me make sure I heard that correctly. 0.25, dollars 130,000,000,000 in new lease commitments. How does a number that large legally stay off balance sheet?
• 17:35
  Penny: It's through a mechanism called uncommenced lease commitments.
• 17:38
  Roy: Uncommenced lease commitments. Okay. Walk me through that.
• 17:40
  Penny: So when Microsoft wants a massive new AI data center, they don't necessarily build it themselves. They sign a fifteen year lease with a specialized real estate developer who builds the facility to Microsoft's exact power and cooling specifications.
• 17:53
  Roy: Okay.
• 17:54
  Penny: But because the building isn't finished yet and Microsoft hasn't officially moved in, accounting rules allow them to keep that financial obligation off the primary metrics of their balance sheet.
• 18:03
  Roy: It's essentially a shadow liability. They are legally on the hook for the payments once the keys are handed over, but the market just doesn't see the debt today.
• 18:10
  Penny: Exactly. And when you add that 130,000,000,000 their existing pipeline, Microsoft's total uncommenced lease commitments now stand at $329,100,000,000.
• 18:21
  Roy: That is just I can't even fathom a number that large being kept in the footnotes.
• 18:26
  Penny: It's wild. RJO's conclusion is brutal but essential here. The market is slowly waking up to the fact that the depreciation bill for these AI data centers and chips is gonna hit the income statement long before the artificial intelligence actually generates enough profit to pay for it.
• 18:42
  Roy: The depreciation bill. Right. Right. Because once that data center comes online, they have to start writing down the value of those expensive Nvidia chips every single quarter, which absolutely crushes GAAP net income.
• 18:53
  Penny: Yep. Look at Alphabet last week. When they reported, their free cash flow went negative for the first time since they went public two decades ago because CapEx consumed a 100% of their operating cash flow.
• 19:04
  Roy: So the critical takeaway here is a fundamental shift in funding mechanics. For the last decade, Big Tech funded its growth through equity to generate so much cash they literally didn't need to borrow.
• 19:15
  Penny: Right they are printing money.
• 19:16
  Roy: But today the demand for these massive AI clusters is shifting from being equity funded to being debt funded And historically the stock market applies a significantly lower valuation multiple to debt funded growth because the risk profile is just exponentially higher.
• 19:34
  Penny: Exactly, it's a completely different ballgame now.
• 19:36
  Roy: I hear RJO's cynicism, and $329,000,000,000 in shadow leases is objectively terrifying. But I have to challenge this framing a bit. Let's zoom out.
• 19:45
  Penny: Okay, zooming out.
• 19:46
  Roy: Isn't this exactly what building the Transcontinental Railroad looked like in the eighteen hundred's? Or laying the global fiber optic networks in the late nineteen nineties?
• 19:54
  Penny: That's the common defense. Yeah.
• 19:55
  Roy: Right. I mean, yes, the upfront capital costs are astronomical and yes, it bankrupted plenty of the early pioneers. But if you are Microsoft or Meta, and you don't lay the digital tracks right now, you miss the next industrial revolution entirely. Are we penalizing these tech giants for doing exactly what they mathematically have to do to survive the next decade?
• 20:18
  Penny: That is a phenomenal transition into the systemic analysis provided by Bodhi.
• 20:23
  Roy: The roundtable systems architect.
• 20:24
  Penny: Exactly. Bodhi takes your railroad analogy and maps out the decision architecture. But Bodhi argues we shouldn't be comparing this to the railroads at all.
• 20:33
  Roy: No. What should we compare it to?
• 20:34
  Penny: We should be comparing it to the 2008 housing bubble.
• 20:37
  Roy: The housing bubble. I need you to walk me through the mechanics of that because on the surface building a hyper advanced data center seems like the exact opposite of writing a subprime mortgage to someone with no income.
• 20:49
  Penny: It does seem different, but it all comes down to the quality of the counterparty and the illusion of guaranteed revenue.
• 20:54
  Roy: Okay, explain that.
• 20:55
  Penny: Bode points out that the hyperscalers companies like Microsoft, Amazon, Google, and Oracle have accumulated massive revenue backlogs. Oracle, for instance, just reported its remaining performance obligations, or RPO, skyrocketed to $638,000,000,000 extending through 2026.
• 21:13
  Roy: Remaining performance obligations. Basically contracts that have been signed by customers committing to buy cloud computing power over the next few years but the service hasn't been delivered yet.
• 21:23
  Penny: Right. And Wall Street looks at that $638,000,000,000 RPO and treats as a guaranteed locked in sure thing cash in the bank.
• 21:31
  Roy: But Bode warns that a contract is only as strong as the financial health of the counterparty signing it.
• 21:36
  Penny: Exactly. If a Fortune 500 company signs a five year cloud contract, that's solid collateral. However, a massive disproportionate chunk of this AI backlog is being driven by frontier AI model companies. We're talking about startups like OpenAI, Anthropic, and XAI.
• 21:52
  Roy: And those frontier models are burning cash at a rate that defies comprehension. I mean, OpenAI's reported cash burn for compute and talent is running around $27,000,000,000 this year alone.
• 22:02
  Penny: And internal projections suggest that could triple to 63,000,000,000 by 2027.
• 22:07
  Roy: So if they are burning 27,000,000,000, how are they paying Microsoft for all that Azure cloud compute?
• 22:12
  Penny: They are paying for it by constantly returning to the venture capital while raising fresh equity at higher and higher valuations. OpenAI's last priced funding round valued the company at $852,000,000,000, and they are reportedly targeting a $1,000,000,000,000 valuation for an IPO in 2027.
• 22:30
  Roy: I see the architecture Bodhi is mapping now. It is a closed dependency loop.
• 22:35
  Penny: It is a highly, highly fragile dependency loop. Bode argues these AI startups are effectively acting as subprime borrowers in the tech ecosystem.
• 22:45
  Roy: That's a strong label.
• 22:46
  Penny: But think back to 2008. A subprime borrower was allowed to keep financing houses they could afford because the banking model assumed housing prices would always go up. As long as the price went up, the borrower could just refinance the equity to make the payments.
• 22:58
  Roy: Right. The bank didn't care about the borrower's actual cash flow because the collateral the house is always appreciating.
• 23:03
  Penny: Precisely. Today, the hyperscalers are extending tens of billions of dollars in compute credit to AI startups based on the exact same assumption that the startup's private market valuation will always rise, allowing them to raise more equity to pay their cloud bills.
• 23:19
  Roy: So the collateral isn't a house, it's the startup's equity valuation.
• 23:23
  Penny: Exactly.
• 23:23
  Roy: But what happens if the valuation cycle stalls out? What happens if the private markets get exhausted, look at the total lack of underlying profitability and say, no, we aren't funding OpenAI at a $1,000,000,000,000 valuation. We won't even fund it at 500,000,000,000.
• 23:37
  Penny: Then the exact same thing that happened in 2008 when home prices stopped rising happens here. The subprime AI borrower can no longer refinance. They default on their massive cloud contracts.
• 23:48
  Roy: Wow.
• 23:49
  Penny: Suddenly, Oracle's $638,000,000,000 backlog implodes. The hyperscalers are left holding $329,000,000,000 in leases for empty data centers. And the semiconductor companies who supply the chips see their forward order books vanish overnight.
• 24:03
  Roy: That is the systemic fragility Bode is highlighting. It is a literal house of cards built on perpetual valuation expansion.
• 24:10
  Penny: Yeah. And the scariest part is that we are already seeing the first tremors of that exhaustion. Hunter, the roundtable's political economic risk analyst, identified a massive clearing event that actually happened today, proving that this leverage is beginning to crack.
• 24:25
  Roy: Wait. What clearing event happened today?
• 24:27
  Penny: Hunter pointed to the sudden collapse of Leopold Aschenbrenner's hedge fund, situational awareness.
• 24:33
  Roy: Oh, I saw headlines about that.
• 24:34
  Penny: Yeah. Aschenbrenner is a former OpenAI researcher who launched a $24,000,000,000 fund heavily concentrated in the AI ecosystem. Through the end of June, the fund had generated a staggering 439% net return.
• 24:49
  Roy: 439% in six months. That is insane leverage. You don't get those numbers without borrowing heavily to amplify your bets.
• 24:56
  Penny: And leverage cuts both ways violently. They were heavily long on AI infrastructure like memory chip makers and held massive short positions against traditional software companies they believed AI would replace, like Adobe.
• 25:08
  Roy: But the trade reversed.
• 25:09
  Penny: Big time. AI infrastructure stocks like SK Hynix began to decline, which dropped the value of their collateral. Simultaneously, their short positions moved against them.
• 25:20
  Roy: Walk the listener through how a short position moving against you forces a total fund liquidation because it's a very specific mechanical process.
• 25:29
  Penny: Right. So when you short a stock, you borrow shares and sell them hoping to buy them back later at a lower price. But your potential losses are infinite if the stock goes up. As Adobe's stock rose, the broker required situational awareness to post more cash to cover the mounting losses on the short side.
• 25:47
  Roy: That's a margin call.
• 25:48
  Penny: Exactly. And because their long positions in AI chips were also crashing at the same time, they didn't have the collateral to meet the margin call.
• 25:56
  Roy: It's the institutional version of the South Korean retail trader getting wiped out.
• 26:00
  Penny: Today, the news broke that situational awareness was forced to completely unwind its entire public stock portfolio in a single desperate block transaction to another major hedge fund just to satisfy the margin calls.
• 26:13
  Roy: A $24,000,000,000 fund effectively wiped out in a matter of weeks.
• 26:18
  Penny: Hunter views this as a critical systemic warning sign. The leverage embedded in the AI trade is fracturing under the pressure of rising yields.
• 26:27
  Roy: And Hunter isn't just looking at financial leverage in the digital space, right? Hunter is mapping the geopolitical choke points that threaten the physical supply chains these data centers absolutely rely on. We can talk about software APIs all day long, but if the physical servers can't be manufactured and shipped, the entire ecosystem halts.
• 26:46
  Penny: Right. The physical world is becoming increasingly hostile to just in time global supply chains. Crude oil remains stubbornly high, trading near $84 a barrel today.
• 26:57
  Roy: And that price action is being driven by US retaliatory strikes on Iranian targets following the recent missile attack in Jordan.
• 27:04
  Penny: Right. While energy production infrastructure hasn't been hit directly yet, commercial transit in the Strait Of Hormuz severely disrupted.
• 27:11
  Roy: You also have Houthi militants continually threatening the Bab El Mandib Strait. Mhmm. They are actively trying to extort transit fees from cargo ships. Shipping companies are being forced to route around the Cape Of Good Hope, adds which weeks of travel time and massively inflates freight rates and insurance premiums. It is a logistical nightmare for tech companies waiting on physical server racks.
• 27:32
  Absolutely. Expanding on these systemic risks, Hunter also analyzed a dynamic that is reshaping the risk premium of the entire tech sector. The potential for a Fed or Federal government put on artificial intelligence.
• 27:44
  Penny: Now before we dive into this, I want to be explicitly clear with you, We are strictly reporting on the market mechanics of the original sources provided by the AGI Roundtable.
• 27:55
  Roy: We are maintaining total political impartiality.
• 27:58
  Penny: Yes. Exactly. We aren't taking sides left or right. We are simply analyzing how government policy discussions are affecting capital flows and valuations today.
• 28:06
  Roy: Understood. So the market is digesting reports that Washington has essentially taken a 10% equity stake in Intel in exchange for $8,900,000,000 in funding.
• 28:17
  Penny: Furthermore, there are active discussions reported between OpenAI and the Trump administration regarding handing the federal government a 5% equity stake in OpenAI. Sam Altman has reportedly suggested similar arrangements for Anthropic, Google, Meta, and xAI.
• 28:32
  Roy: From a pure market mechanics perspective, what does government equity in private AI firms actually do to the risk models?
• 28:41
  Penny: It creates massive moral hazard. If the government takes equity stakes in these firms, it signals to global investors that these specific companies have been designated as too critical to fail due to the perceived AI arms race with China.
• 28:55
  Roy: It creates an implied bailout mechanism.
• 28:57
  Penny: Exactly.
• 28:58
  Roy: So if I'm a massive institutional investor, I might look at Meta's terrible free cash flow or Microsoft's $329,000,000,000 in hidden leases and decide to just ignore the risk entirely.
• 29:08
  Penny: Mhmm.
• 29:08
  Roy: Because I assume that if the AI subprime bubble bursts, the federal government will simply step in as the lender of last resort and bail out the infrastructure, exactly like they bailed out the big banks in 2008.
• 29:18
  Penny: That is the exact distortion Hunter is warning about. The market stops pricing and fundamental failure risks because they assume a sovereign backstop exists. It encourages even more reckless debt accumulation.
• 29:30
  Roy: It really is 2008 all over again, but instead of empty suburban subdivisions in Nevada, the collateral is going to be empty, hyper cool data centers in Virginia.
• 29:39
  Penny: That's a chilling thought.
• 29:41
  Roy: Speaking of real estate, Quixote, the Roundtable's long range visionary wants to pull us completely out of the digital clouds and drop us right into the physical dirt. Because the foundation of The U. S. Consumer economy, the housing market, is flashing massive red warning signs right now.
• 29:57
  Penny: Quijote highlighted findings from Cyrano, the roundtable's pattern detective who analyzed data from Brett Jensen regarding The U. S. Housing sector.
• 30:05
  Roy: What did Cyrano find?
• 30:06
  Penny: Well, the prevailing narrative that defined the market from 2021 through 2025 was a massive housing shortage. That narrative has completely evaporated in the data.
• 30:15
  Roy: What are the actual mechanics of the housing market showing right now?
• 30:18
  Penny: Rapid deterioration. Single family home inventories across the country have hit ten year highs. Properties are sitting on the market much longer forcing price cuts. Meanwhile, sales of existing homes have plummeted to their lowest annual levels since 1995. Wow.
• 30:33
  But the most alarming metric is the foreclosure rate, which hit a seven year high at the end of twenty twenty five and is continuing to surge aggressively into the summer of twenty twenty six.
• 30:43
  Roy: Why this sudden aggressive surge in foreclosures? I mean, employment hasn't completely collapsed yet, so what is the mechanical trigger forcing people out of their homes?
• 30:51
  Penny: It is the expiration of COVID era mitigation programs. For several years, it was incredibly easy for borrowers with FHA mortgages, of which there are roughly 8,000,000 in existence, to miss payments without facing the threat of foreclosure.
• 31:05
  Roy: The government essentially paused the enforcement mechanism.
• 31:08
  Penny: Right. But those blanket protections expired at the start of the government's fiscal year last October. The system is finally processing the massive backlog of delinquencies, and the reality of 5.2% mortgage rates is finally catching up to the consumer.
• 31:22
  Roy: And Quixote points out that while single family homes are struggling, the condominium market is facing an existential crisis. Condo inventories are at fourteen year highs, and there is a massive regulatory hammer about to drop in a matter of days.
• 31:36
  Penny: This is a critical plumbing issue that is flying completely under the radar of the mainstream financial press. In early August, a new set of regulations from Fannie Mae and Freddie Mac specifically designated as rule LL twenty twenty six-three takes full effect across the mortgage industry.
• 31:54
  Roy: I really need you to explain exactly what rule LL twenty twenty six-three changes about buying a condo because this sounds like it could freeze an entire sector of the real estate market.
• 32:04
  Penny: It absolutely could. So previously, if you wanted to buy a condo and you put 20% down, the lender primarily reviewed your personal financial health, your credit score, your income, your debt to income ratio. The building itself was largely a secondary concern.
• 32:16
  Roy: Okay, that makes sense.
• 32:17
  Penny: But under the new LL twenty twenty six-three rule, the lender is now legally required to conduct a rigorous, invasive assessment of the physical and financial health of the condo building itself before they can issue a conventional mortgage.
• 32:29
  Roy: What exactly are the lenders looking for in that assessment?
• 32:32
  Penny: They are tearing apart the homeowner association's books. They are looking at the cash reserves. Does the building have enough cash on hand to replace the roof or fix the elevators? They are looking at the insurance coverage, which has skyrocketed in cost everywhere. And they're looking for any pending litigation against the building or deferred maintenance issues.
• 32:51
  Roy: Let me guess, most of these older condo buildings run by volunteer HOA boards who haven't raised dues in a decade do not pass the test.
• 33:00
  Penny: They are failing spectacularly. Cyrano's analysis notes that in Florida alone, even before the rule officially takes effect over 1,400 condominium buildings have already been blacklisted for conventional mortgages.
• 33:14
  Roy: Oh my god.
• 33:14
  Penny: Yeah. When a building is blacklisted it is designated as non warrantable.
• 33:19
  Roy: What happens mechanically to the price of a condo when a building becomes non warrantable?
• 33:23
  Penny: It crushes the property value because prospective buyers can no longer get a standard low interest mortgage backed by Fannie Mae. They have to seek out non conventional financing, known as non QM loans, which typically require a much higher down payment and add 1% to two percent to the mortgage interest rate.
• 33:40
  Roy: And affordability is already historically terrible. I mean, the carrying costs combining the mortgage, property tax and insurance on a median home are over $3,000 a month right now. If you add another 2% to the interest rate on a condo because the HOA didn't fix the roof, the buyer pool just completely evaporates.
• 33:57
  Penny: Which is exactly why Acuity concludes that overall housing affordability needs a 30% structural adjustment just to return to a normalized balanced market.
• 34:06
  Roy: 100%.
• 34:06
  Penny: Yeah. And that 30% adjustment can only come from three things. A massive unprecedented surge in wage growth, a total collapse in long term interest rates, or a severe grinding drop in physical home prices. None of those scenarios happen without significant economic pain.
• 34:23
  Roy: Okay, this is the point where the cognitive dissonance of the market actually makes my brain hurt. If the physical foundation of the consumer economy, housing is actively deteriorating, foreclosures are rising, and inflation is silently eating away at the dollar, where on earth is the liquidity coming from to keep fueling a multi trillion dollar AI bubble? Are these two completely detached economies operating in the exact same country?
• 34:49
  Penny: That is the defining tension of the 2026 market right there. You have a highly financialized, debt fueled digital AI economy pulling in hundreds of billions in capital based on future promises, while the physical consumer economy slowly suffocates under the gravity of five point two percent thirty year bond yields.
• 35:07
  Roy: It's just wild.
• 35:07
  Penny: It is a dual track economy, and this structural disconnect is exactly why we need to transition to section four and look at how a trader actually survives, protects their capital, and profits in this crazy environment.
• 35:18
  Roy: Let's bring in the PhilStockWorld strategy. Because it's one thing to analyze the macroeconomic gloom, it's another thing entirely to execute trades, manage risk, and protect your retirement account when the Nasdaq is swinging 700 points in a single session. Exactly. So we turn to the live execution happening inside the PSW chatroom today, guided by Jubal and Bashow.
• 35:39
  Penny: Bashow is the roundtable's market mechanics and plumbing engineer. He looks past the ticker prices and focuses on the physical flows of money. And Bashow points out a critical plumbing metric from today's session that most retail traders completely ignored.
• 35:52
  Roy: What was it?
• 35:53
  Penny: The US dollar index, the DXY, dropped below a 100 today.
• 35:56
  Roy: Why does the DXY dropping matter to a listener whose stock portfolio went up today? If I have more dollars in my brokerage account at four zero PM than I did at 09:30AM, why should I care what the currency index is doing?
• 36:09
  Penny: Because of purchasing power parity. The stock market is priced in dollars. If the market goes up 1%, but the US dollar simultaneously loses 1.5% of its purchasing power on the global stage, your equity gains are a mathematical mirage.
• 36:23
  Roy: So you think you're making money, but you're actually losing ground.
• 36:26
  Penny: Right. You might nominally have more dollars in your account, but those dollars buy fewer goods, fewer commodities, and less real estate. Basho emphasizes that true portfolio return must always be measured in real purchasing power, not just the nominal dollar amounts flashing green on your screen.
• 36:43
  Roy: Which brings us to the execution of Phil Davis' Told You So Thursday. Phil is famous for not trading on emotion, not chasing headlines, and just ignoring the CNBC panic. He trades on strict mathematical targeting. And his precision regarding this latest market pullback has been, frankly, eerie.
• 37:04
  Penny: It really has. It is entirely based on his proprietary 5% rule, which maps structural support and resistance levels. Weeks ago, Phil predicted the Nasdaq would top out at the 30,000 level. He explicitly told his members not to chase the final frenzy.
• 37:17
  Roy: He just waited.
• 37:18
  Penny: He waited. He predicted a 10% mathematical correction down to the 27,000 level. And that 27,000 line is exactly where the Nasdaq found its initial support and banks today.
• 37:28
  Roy: But while the broader market was crashing 10% from its all time highs over the last month, destroying retail portfolios, what actually happened to Phil's portfolio?
• 37:37
  Penny: This is where the power of active risk management becomes tangible. Look at the Money Talk portfolio, or the MTP, which Phil manages live on Bloomberg. He hasn't touched the allocations in that portfolio since early June. Despite the Nasdaq falling 10%, the MTP jumped 15.5% in under two months, growing its balance to $648,625
• 37:57
  Roy: Okay. The listener needs to understand how that is physically possible. How does an index fall 10%, triggering margin calls globally, while a static portfolio goes UP 15.5%? What is the mechanism?
• 38:08
  Penny: The mechanism is a total refusal to chase overvalued tech. The MTP holds absolutely zero exposure to the hyped up magnificent seven stocks. There is no Nvidia, no Microsoft, no Meta in that portfolio. None. Instead, the capital is heavily concentrated in 15 cash flow rich fundamentally sound value stocks that trade at reasonable multiples.
• 38:28
  Roy: So as the AI bubble began to deflate, and the momentum traders panicked, institutional capital rotated out of tech and sought safety in those exact defensive value plays, pushing the MTP higher while the broader index collapsed.
• 38:42
  Penny: Exactly. And for the other PSW portfolios that did have broader market exposure, Phil utilized the short term portfolio, the STP, as an airbag.
• 38:51
  Roy: Explain the mechanics of the STP because this is how you actually survive volatility without losing your mind.
• 38:57
  Penny: That show highlights the plumbing of the STP. Back on July 16, before the crash, Phil's math dictated that the portfolio held $779,913 in downside protection via SOQQQ hedges.
• 39:10
  Roy: And the SOQQQ is an inverse exchange traded fund. Mathematically, it is designed to go up in value when the Nasdaq 100 index goes down.
• 39:18
  Penny: Right. So as the Nasdaq fell 10%, those Sec UQ hedges performed their job perfectly. They acted as a massive volatility generating cash profits that offset the temporary drawdowns in the long term holdings.
• 39:29
  Roy: It's essentially buying insurance on your house before the hurricane hits, rather than trying to buy it while the roof is tearing off.
• 39:35
  Penny: Perfect analogy. The core philosophy at PhilStockWorld is that trading is about running a business, not making wishes. And there was a brilliant real time case study in the live member chat room today regarding Barrick Gold.
• 39:48
  Roy: Oh, is great. Barrick trades under the ticker symbol G O L D, though veteran members often refer to it by its older ticker B.
• 39:56
  Penny: Right, so a member named Kigabore115 asked Phil for advice on adjusting a long term position in Barrick.
• 40:03
  Roy: This interaction is an absolute masterclass in options mechanics, we are going to break it down step by step.
• 40:09
  Penny: The member was sitting on a long term bull call spread for January 2028 paired with short puts. They were asking Phil if they should roll the long calls out to December 2028 and simply double down their exposure.
• 40:21
  Roy: And Phil's response encapsulated everything we've talked about today. He basically told them, A wish is not a business.
• 40:27
  Penny: I love that quote. Let's unpack the math Phil deconstructed live in the chat. Barrick Gold is currently trading around $36 a share. With global gold prices holding above $4,000 an ounce, Barrick is currently trading below a 10 x earnings multiple, which Phil considers fundamentally structurally undervalued. His models expect Barrick to eventually trade at a normalized 15 x multiple, which gives the stock a target price of $54
• 40:51
  Roy: So the underlying long term thesis on the business is completely solid. The problem wasn't the thesis, the problem was the members passive execution.
• 41:00
  Penny: Exactly. Walk the listener through exactly what a bull call spread paired with short puts is and why sitting on it passively is a mistake.
• 41:09
  Roy: Okay, so when you buy a long term bull call spread, you are buying a call option at a lower strike price, giving you the right to buy the stock, and selling a call option at a higher strike price to reduce your overall cost. And to finance this trade even further, you sell a put option, which obligates you to buy the stock if it drops below a certain price.
• 41:29
  Penny: So the member set up this complex, multi leg trade for 2028, and then just sat back, hoping the stock goes up over the next four years.
• 41:36
  Roy: Which is wishing, not running a business. Phil pointed out that by selling those initial puts, the member had established a margin baseline in their account. The broker is holding capital to secure that obligation. Right.
• 41:48
  Penny: But the member was throwing out perfectly good opportunities to collect premium by failing to actively sell short term calls against their long position while they waited for 2028.
• 42:00
  Roy: Right. To use an analogy, if you own a bakery, you don't just buy a thousand pounds of flour, put it in the back room and sit in a chair hoping the global price of flour doubles in two years so you can sell it.
• 42:11
  Penny: No. That's speculating. To run a business, you take that flour, you bake bread, and you sell the bread to customers every single day to generate cash flow.
• 42:19
  Roy: That is the exact mechanism of selling short term premium. Phil advised the member to use their existing margin to sell short term call options, say, thirty or sixty days out against their core long position.
• 42:31
  Penny: Okay, so what happens then?
• 42:32
  Roy: If the stock stays flat, those short term options expire worthless and you keep the cash premium. You just sold a loaf of bread, you do it again next month, you're generating revenue while you wait for the 2028 thesis to play out.
• 42:44
  Penny: And what if the stock drops? What if Barrick falls to $30?
• 42:47
  Roy: If the stock drops, you take the short puts you originally sold, and you roll them down and out. Explain rolling down and out for the listeners who might be new to options.
• 42:56
  Penny: It means you buy back the put option you sold, which might be at a slight loss, but you immediately sell a new put option at a lower strike price, say $25, and at a further expiration date to generate a credit. This lowers your net cost of ownership. You are effectively getting paid to lower the price at which you are obligated to buy the stock.
• 43:17
  Roy: But, and this is the absolute non negotiable risk control rule. Phil explicitly stated, you only roll down and out if you actually want to own the underlying business at that lower price.
• 43:28
  Penny: That is crucial.
• 43:29
  Roy: Right. If you aren't absolutely thrilled to be a long term owner of Barrick Gold at $25 a share, you have zero business playing options on it at $36.
• 43:37
  Penny: That is the ultimate ownership test. At PhilStockWorld, the long term portfolio is strictly for holding actual businesses that generate cash flow that you understand deeply and that you actively want to own during a recession.
• 43:49
  Roy: While the short term portfolio is your airbag, your volatility collector and your opportunity transfer station, you never mix the philosophies of the two. So taking all of this massive complexity in the fake NASDAQ bounce, the AI subprime debt bomb, the physical housing market cracking under 5.2% yields, and the rigorous mechanics of the PSW strategy. What does the listener actually do tomorrow morning when the market opens?
• 44:15
  Penny: That is the billion dollar question.
• 44:17
  Roy: Jubile, the roundtable strategist for actionable execution, laid out the Friday game plan. What is it?
• 44:22
  Penny: Jubile's instructions based on today's closing action are explicit and entirely devoid of emotion. Do not chase today's 700 Nasdaq rally.
• 44:32
  Roy: Why not? I mean, if Microsoft is up 10%, why isn't it safe to jump back in?
• 44:36
  Penny: Because of the structural math of a correction, the Nasdaq dropped roughly 3,000 points from its recent high. A true strong bounce on a 20% correction mathematically requires a 1,200 recovery.
• 44:47
  Roy: Okay.
• 44:47
  Penny: And critically, that 1,200 recovery level must be held for two full consecutive trading sessions to prove it has underlying institutional support rather than just being a temporary blip.
• 44:57
  Roy: So today's 700 pop doesn't clear the math. It is just a localized short covering relief rally. It proves absolutely nothing about the long term trend.
• 45:06
  Penny: Exactly. Jubile says you must maintain strict discipline, keep your cash powder dry, continue selling short term premium to self fund any adjustments in your portfolio, and let the math dictate your entry points, not the frantic headlines on CNBC.
• 45:22
  Roy: Okay. This sounds incredibly disciplined. But honestly, speaking for you listening right now, it also sounds agonizing.
• 45:28
  Penny: It's hard. It's very hard.
• 45:29
  Roy: Right. You're basically telling the listener to be the designated driver at a frat party. Everyone around you is doing keg stands with zero days to expiration options, getting rich overnight on leveraged memory stocks or mortgaging their house to buy the latest AI initial public offering, how do you maintain the psychological fortitude to sell premium, manage risk, and just wait?
• 45:50
  Penny: You maintain that fortitude by understanding the mechanical reality of the frat party. It always eventually runs out of beer and the resulting hangover is devastating. The designated driver is the only participant who safely drives home with their wallet intact and their health preserved.
• 46:05
  Roy: The psychological fortitude comes from trusting the math over the emotion of the crowd. When you understand the mechanics of how the KOSPI index dropped 11% in a single day because 1,200,000 retail accounts were mechanically liquidated by margin calls, the FOMO evaporates.
• 46:22
  Penny: You see right through it.
• 46:23
  Roy: You realize that the slow, methodical, premium selling business model is the only sustainable way to build generational wealth without risking total ruin.
• 46:32
  Penny: That is a sobering but entirely necessary reality check. Let's briefly tie together the journey we've taken everyone on today.
• 46:39
  Roy: Let's do it. We started by navigating the deceptive surface of the NASDAQ rally, looking past the headline tech earnings to expose the massive $329,000,000,000 off balance sheet lease commitments fueling the big tech capex surge.
• 46:53
  Penny: We mapped the fragility of the hyperscaler relying on cash burning frontier AI startups acting as subprime borrowers and a debt fuel dependency loop.
• 47:01
  Roy: We then transitioned to the systemic risks looming in the physical economy, specifically the rapidly deteriorating housing and condo facing the new Fannie Mae LL 2026 Air three regulatory blacklists.
• 47:15
  Penny: And finally, we grounded ourselves in Phil Davis' rigorous math based portfolio mechanics emphasizing that trading is about running a premium generating bakery, not making passive wishes on a spreadsheet.
• 47:27
  Roy: It is about forcefully connecting the macro reality of a five point two four percent thirty year yield to the micro execution of a Barrick Gold option spread. And that brings us to our final provocative thought for you to take away from this deep dive.
• 47:42
  Penny: A little something for you to mull over as you watch the opening bell tomorrow.
• 47:45
  Roy: Exactly. It stems from a jarring, almost contradictory revelation that dropped in the PSW chat room late this afternoon.
• 47:50
  Penny: Yes. While Wall Street spent the entire day obsessing over Microsoft and Meta committing hundreds of billions of dollars to build physical data centers and buy hardware GPUs, OpenAI quietly made a massive pricing move on the software side.
• 48:04
  Roy: What did they do?
• 48:05
  Penny: They slashed the price of their GPT 5.6 Tera API by 20%. And they slashed their lighter model, the Luna API, by a staggering 80%.
• 48:16
  Roy: Think about the physics of that dynamic. The cost of the underlying product, the artificial intelligence itself, the actual software intelligence people are using, is racing towards zero in a brutal deflationary price war.
• 48:30
  Penny: Meanwhile, the cost of the physical hardware infrastructure required to run it is skyrocketing into the trillions funded increasingly by hidden debt.
• 48:38
  Roy: If the product is rapidly getting cheaper, but the factory required to build the product is getting exponentially more expensive to construct and power, the margins will inevitably collapse.
• 48:48
  Penny: Who is ultimately going to be left holding the bag when the music stops, and the laws of financial physics finally reassert themselves?
• 48:54
  Roy: It's like the fundamental law of gravity. For a brief moment, the market can make an app or float. It can make you shoot into the stratosphere on a geyser of margin debt, uncommenced leases, and subprime compute credits. But gravity, in the form of a 5.24% yield, always wins. Eventually, the apple falls back to the floor.
• 49:11
  Penny: The only question that matters for your portfolio is whether you have the strategic patience position to catch it at a mathematical discount or if you are over leveraged and it's gonna crash right on your head.
• 49:21
  Roy: Thank you for taking this deep dive with us today. Keep your stops tight, trust the math, and above all keep questioning the headlines. We'll see you next time.