Welcome Curious Minds to the deep dive. Today we're, kicking things off with a number that might just make you do a double take. Dollars $225,000,000,000.

Billion with a B.

Exactly. That's how much Oracle, yes, Oracle gained in market value overnight.

Which is pretty wild.

And here's the twist, right? It happened after they slightly missed their earnings report?

Yeah.

Okay. Let's unpack this. A $225,000,000,000 overnight game despite missing earnings. I mean, that just doesn't sound right on the surface. What is the secret sauce here?

Well, what's fascinating here is that this isn't really about the traditional quarterly earnings game. You know?

Okay.

Today, we're diving deep into some unseen currents shaping the tech market, specifically AI infrastructure, cloud computing.

Right.

We've been looking through a stack of recent financial analyses, market reports, and our mission really is to cut through the noise. Mhmm. Understand this seemingly contradictory market behavior and, you know, truly grasp what's driving the AI revolution's picks and shovels economy.

Picks and shovels. I like that.

Yeah. And, the core concept you really need to get your head around is called remaining performance obligations, RPO.

RPO. Okay. Haven't heard that one much before.

Right. It's usually not a headline grabber. But Oracle's surge in RPO is well, it's truly unprecedented.

So, what is it exactly?

Simply put, RPO represents the total amount of contracted future revenue, money they've secured but haven't yet recognized on the income statement.

Ah, okay. So, future business locked in.

Exactly. For cloud and subscription services, it's a key forward looking indicator. It tells you about future revenue streams, their backlog. It shows what's coming down the pipeline, not just what's already arrived.

And, what's coming down Oracle's pipeline is, well, nothing short of astronomical based on these reports.

Absolutely.

Let's talk numbers. Oracle's RPO absolutely blew out. It soared by a staggering 359%

to

an astonishing $455,000,000,000

Wow!

To put that in perspective for you listening, there were total revenue last year around $57,400,000,000

This is huge compared to that.

This $455,000,000,000 is more than six years worth of their current annual revenues.

Six years?

It's like a like a pizza chain suddenly getting six years of confirmed orders when they normally only deliver one year at a time. It really highlights the sheer scale.

That's a great analogy.

So what exactly is Oracle Cloud Infrastructure, OCI, that's driving this? Right. Because it's clearly not just plain old cloud storage. Right?

Exactly right. OCI is far more than just generic cloud storage. You should think of it as the specialized computing foundation that's really powering the AI revolution.

The engine room?

Kind of, yeah. Yeah. The high performance engine room for the most demanding AI workloads. The core products driving this massive RPO include things like AI optimized compute instances

Okay.

Specifically designed for machine learning. Their autonomous database services, which have AI and ML capabilities built right in, and their Exadata Cloud service that's engineered for extreme data processing.

Gotcha.

And this RPO, it also includes multi year enterprise software subscriptions, you know, across their core suites like ERP.

Enterprise Resource Planning.

Right. ATM.

Human capital management. Yep. NSCM. Supply chain management? Okay.

The big stuff.

Exactly. And the contracts themselves, they're huge. Oracle signed four multi billion dollar deals in just one quarter.

Wow. Any names we know?

Well, there's a major partnership with OpenAI for their GPT infrastructure.

Okay. That's

big. Significant integration deals with Microsoft Azure. Tesla's doing a full stack Oracle implementation. Yeah. And then there are additional undisclosed contracts with other hyperscalers.

Hyperscalers.

Yeah.

You mean the really big cloud players like Amazon, Microsoft, Google, the Amazon.

Exactly. And the key point here is these aren't like speculative deals. They are legally binding multi year commitments for immense computational capacity.

That's an important distinction you made earlier contracted reality. But how can Oracle, I mean a company with decades in enterprise software, sure, but how can they actually deliver on these massive multi year commitments? Building out this kind of infrastructure at that scale, it seems like an impossible task for almost any company.

It's a critical question, absolutely. And what's fascinating, I think, is that Oracle isn't just selling a dream here. They're selling contracted reality, like you said, but it's backed by decades of scaling experience. This isn't, you know, your traditional software company anymore. It's become a specialized infrastructure A bit like how Taiwan Semiconductor, TSM, is for chip manufacturing.

Ah, interesting comparison.

Yeah.

The picks and shovels provider for AI infrastructure.

Precisely. And they have several key advantages. First, just raw financial firepower. Oracle generates over $20,000,000,000 in annual cash flow.

$20,000,000,000.

Yeah. Which allows them to spend, say, $1,520,000,000,000 dollars yearly on building new infrastructure directly from profits. They don't need to wait for external financing.

That's a huge advantage.

It is. Second, modular construction. They use these prefabricated data center module, I think giant Lego blocks. It enables automated deployment, standardized designs, they can replicate them quickly anywhere in the world.

Makes sense. Faster build out.

Third, smart technology. They have something called Oracle RAC, which lets them add competing power basically by just plugging in more servers. And their autonomous database, it manages itself, which reduces the human admin needed.

Self managing databases. Sounds efficient.

Very. Plus, their systems use AI driven optimization to automatically adjust resources. Fourth, there's been a strategic shift in their sales focus. How so? Instead of just hunting for new customers, their sales team is now primarily focused on delivering services to these existing contracted customers and expanding those relationships.

So more like account management than cold calling.

You got it. And finally, just proven scaling experience. Oracle's been deploying data centers for decades. They predict capacity needs months ahead, handle massive demand spikes. They've consistently deployed new data centers in under twelve months.

Okay. So they have a track record.

Exactly. This proven ability to deliver is precisely why these huge trillion dollar companies are signing on the dotted line.

It sounds like Oracle is really, thinking several steps ahead here. You also mentioned, a new initiative they're pursuing. Something that goes beyond just providing the raw computing power.

Yes.

Something that could, you know, truly change the game for businesses trying to adopt AI. What is this AI in a box strategy and what does it mean for companies, especially those struggling with AI adoption because of say, data security and privacy concerns.

Right. This upcoming initiative from Oracle, it could represent maybe the most transformational business model shift we've seen in enterprise AI deployment yet.

Okay. Intriguing.

The core concept is this. Instead of companies sending their proprietary sensitive data out to external AI models, like a public chat GPT

Which is a huge concern for many.

A massive concern. Oracle is bringing the AI models directly to the data. They deploy custom trained large language models, LLMs

Got it. LLMs.

Inside the enterprise's secure environment, right where that valuable data already lives.

Ah, so the data doesn't leave their control.

Exactly. This directly addresses the biggest barrier to enterprise AI adoption, those security and privacy worries that stop companies using public AI with sensitive corporate info.

Makes perfect sense. It is. So how does it work technically? Magic?

Well, the technical magic involves their autonomous database storing and managing all the company data, then OCI generative AI services deploy these customized LLMs within that same secure bubble.

Right.

They offer things like Oracle blueprints for, pre configured deployments and even agentic AI capabilities.

Agentic AI. What's that?

It means AI agents that can proactively perform tasks, make decisions, essentially create autonomous business workflows based on the company's data.

Wow. Okay. That sounds like a truly transformative shift for businesses. It's not just renting servers anymore. It's like renting an AI powered intelligence that knows your company inside out.

That's a perfect way to describe it.

Can you elaborate a bit on the, the business model implications of this? Cause that sounds like where the real value is.

Absolutely. This is where the business model truly shines. It shifts, like you said, from merely rent our servers to rent our AI powered business intelligence that knows everything about your company.

Right.

This creates exponentially higher value per customer. It creates massive switching costs because the AI gets deeply embedded, it learns your entire business.

Hard to rip out once it's in there.

Very hard. And it creates continuous data dependency because the AI gets smarter the more data it sees. We're potentially talking about our revenue multiplication effect here, maybe increasing revenue per customer five, ten times.

Five to 10 times.

Potentially. Turning, say, a $13,000,000 enterprise customer into a $320,000,000 one.

Okay. Now that $455,000,000,000 RPO makes a bit more sense if they can convert customers like that.

It helps validate it. Yes. It transforms it from just infrastructure rental to these comprehensive AI transformation contracts.

However There's always a however.

There is. And this raises an important question for you, the learner. If Oracle can't execute on this massive backlog, what happens then? What stands out to you about these potential challenges?

It does seem like a huge undertaking, even for a company like Oracle. What? What are the potential pitfalls they face? I mean, what could go wrong?

Indeed. While the outlook seems bullish, there are definitely potential pitfalls to consider. One is simple technology disruption risk. Like? Well, what if quantum computing emerges faster than we think?

Or entirely new chip architectures come along that make Oracle's current infrastructure obsolete sooner than planned?

Right. The tech landscape changes fast.

Very fast. Another risk is economic reality. What if the promised AI productivity gains don't materialize as quickly? Or maybe customers just overestimated their compute needs? That could lead to budget pressures, maybe even contract renegotiations down the line.

Okay. So the demand might not be as solid as the contracts suggest long term.

It's a possibility, and critically there's execution failure. What if Oracle simply can't deliver the promised capacity on schedule?

That would be bad.

Very bad. Service level failures could trigger penalty clauses, Or customers might just switch to competitors mid contract if they can. You know, the precedent of companies like WeWork. Oh. Or even some .com era companies with massive contracts that didn't end well.

That serves as a kind of cautionary tale.

Though Oracle's customers are a bit more stable than WeWork's.

Generally, yes. We're talking trillion dollar entities, typically with strong cash flow. But still, the downside risks are significant if Oracle fails to deliver on this scale. It's not speculative.

Those are significant challenges and they actually lead us nicely into this brilliant analogy you mentioned from the analysis. You know how customers couldn't wait eight, ten years for a Boeing seven eighty seven Dreamliner?

Well

enterprises are kind of in a similar bind with infrastructure. They simply cannot wait three, four years.

The pace is just too fast.

Exactly. So Oracle sitting on over six years of contracted work, it's actually creating immense opportunities for its competitors, doesn't it? Almost like the Airbus to their Boeing.

That Boeing Airbus analogy perfectly captures the massive backlog reality we're seeing across the entire cloud infrastructure market. It's not just Oracle.

Right.

If you combine the backlogs of the big four cloud providers, you've got AWS at $195,000,000,000 Microsoft at $368,000,000,000, Google Cloud at a $106,000,000,000, and now Oracle at $455,000,000,000.

Add that all up.

You're looking at nearly $1,000,000,000,000 in contracted orders across just those four.

A trillion dollars.

So Oracle's sheer backlog, while impressive for them, creates a bottleneck for the industry. Enterprises simply cannot afford to wait that long in this rapidly evolving AI landscape.

So who are the Airbuses then? Who's stepping up to meet this demand that Oracle might be too backlogged to handle immediately?

Well, the Airbus opportunities are pretty clear. Microsoft Azure is certainly a clear Airbus here.

Okay.

They're already partnering with Oracle, ironically, using Oracle's infrastructure for Bing AI because even Microsoft can't build fast enough themselves right now.

That's fascinating.

Azure's growing at 39% annually. They have that huge $368,000,000,000 backlog, and their deep enterprise software integration creates strong incentives for customers to stick with them or choose them. Microsoft is often seen as like a solid defensive cloud play, reliable growth.

Okay. So Microsoft is one. Who else?

Then there's Google Cloud, kind of the underdog with AI advantage.

How so?

Well, they're growing at 32%. They have genuine native AI advantages from DeepMind and Gemini. And because they have a lower market share currently around 11%, they just have more room to grow.

More upside potential.

Exactly. Google can potential be more agile, capture some of this displaced demand, maybe leveraging their superior AI tech as the hook.

Interesting. Are there others besides the big three plus Oracle?

Yeah. Beyond the hyperscalers, we're seeing specialized providers emerge winners too, like CoreWeave. They focus specifically on GPU specialized infrastructure, critical for AI. They have a $30,100,000,000 backlog already.

$30,000,000,000, wow.

And another one is Nebius Group. They recently secured a $19,400,000,000 deal with Microsoft actually for AI infrastructure.

So the pie is big enough for more players.

Absolutely. And stepping back, if we connect this to the bigger picture, this whole dynamic means enterprises are increasingly adopting multi cloud strategies.

Makes sense. Don't put all your eggs in one basket.

Exactly. They're doing it to avoid vendor lock in and also just to hedge against these capacity concerns. Can provider x actually deliver what I need it? This isn't about one winner taking all. It's about a rapidly expanding market with multiple winners.

That's fascinating. So for those of you listening who might be wondering about investment implications, how does this play out? Any thoughts there?

Well, based on this analysis, we could see tier one plays, let's call them, like Microsoft and Alphabet, Google. They seem like direct beneficiaries of Oracle's capacity constraints. They offer substantial growth potential, arguably without Oracle's specific massive execution risk right now.

Okay.

Then there are maybe tier two plays, the infrastructure enablers that benefit, kind of regardless of which cloud provider wins the specific deal.

The picks and shovels suppliers again.

Exactly. Think Broadcom, which provides custom silicon chips for all the major cloud providers. Or, of course, NVIDIA, the undisputed GPU kingmaker. Their technology is required by basically every cloud provider building out AI infrastructure.

So they benefit from the whole industry build out.

Right. The rising tide lifts those boats.

Okay, so Oracle's ambitious plans, the sheer scale of the AI build out, it naturally brings up questions about the, the broader economic climate doesn't it? Can this tech euphoria, this AI boom really be sustained if other parts of the economy are showing signs of strain? What did that August PPI number for instance tell us about inflation and maybe the Fed's next move in this context?

That's a critical question. The broader market context really reveals an interesting collision doesn't it? This AI euphoria meeting some significant macro doubt.

Yeah.

We saw that unexpected August producer price index, the PPI, actually declined by 0.1% month over month. The expectation was for a gain of 0.3% and core PPI was also negative.

Okay lower than expected inflation at the producer level.

Right. Year over year PPI slowed to plus 2.6%. Now this benign print as some called it gave the bull some comfort. Maybe tariffs weren't ripping through the supply chain as badly as feared.

And what did it do to bond yields and Fed expectations?

Bond yields slid quite a bit. And pretty dramatically, the odds of a Fed rate cut at the September FOMC meeting jumped. We saw odds hitting 92 for a quarter point cut.

25 basis points.

Right. 25 BPS. And even about an 8% chance priced in for a half point cut or 50 basis points.

Wow. The market really reacted. So on the surface, lower inflation, potential rate cuts. That sounds like positive news for the economy. Right?

Is that the full picture or are there underlying concerns we should maybe be aware of?

That's a fair point. Lower inflation, easier money. Generally good. However, beneath that surface, there are some pretty significant signals of a weakening economy showing up in the data.

Such as?

Well, consider the massive downward revision to US employment growth. They revised down the growth in the year through March 2025 by 911,000 jobs.

Minus 911,000. That's a huge revision.

It's a huge number. We're also seeing clear signs of a, well, a tapped out consumer.

Tough so.

The savings rate is down at 4.4%, which is well below the historical average. Credit card debt hit a new record, $1,210,000,000,000. Delinquencies on those cards and other loans are rising, and consumer sentiment just remains stubbornly low.

Not a great picture for the consumer.

Not really. And this raises that important question. Can the market sustain this AI fueled rally if the consumer, who powers what, 70% of GDP?

Yeah. The bulk of

it. If the consumer is buckling under pressure, the market seems to be betting on Fed cuts and AI innovation to carry growth forward. But, you know, if tomorrow's CPI report comes in hot, that delicate balance could snap pretty quickly.

It feels a bit fragile. And when we look at the recent Q two twenty twenty five earnings, they generally exceeded expectations. Right? But how much of that positive news was actually, broadly distributed across the market? Was it widespread?

Not broadly at all, actually. And if we look at the wider implications, this really highlights a significant concentration risk in the market.

Concentration risk.

Yes. Q two earnings beating expectations. That was largely driven by the magnificent seven tech giants.

Okay. The usual suspects.

The usual suspects. Get this. The Mag seven grew their earnings per share by a whopping 26% year over year.

26%.

Meanwhile, the other 493 stocks in the S and P 500, they only grew earnings by a modest 24%.

Wow. That's a massive difference.

It is. So it means that the headline S and P 500 growth really isn't indicative of broad market health.

Yeah.

You've got this handful of tech giants basically distorting the aggregate numbers.

Pulling the average way up.

Exactly. And we also just need to briefly mention the tariff headwinds. President Trump's expanded tariff regime, it is starting to show its impact on corporate performance. You had companies like General Motors reporting a $1,100,000,000 loss. Tapestry revising its profit outlook down.

Right. Real impacts.

Real impacts. And all of that affects how companies manage inflation too. So it's another cross current. So just to recap then, we've really explored this dual narrative playing out in the markets right now. On one hand, you have this undeniable AI infrastructure boom.

It's spearheaded by Oracle's, frankly, unprecedented RPO numbers and their potentially game changing data plus LLM strategy.

A massive wave of tech investment.

Truly massive. But that's juxtaposed with these broader economic current signals of a potentially weakening consumer and a federal reserve that seems poised to cut rates, all happening amidst a very concentrated tech rally.

It's quite a mix. Oracle's incredible backlog certainly proves AI infrastructure is becoming a contracted reality, not just, you know, hype or a dream. Mhmm. But that whole cloud wars dynamic we talked about, the Boeing Airbus thing. It suggests that even these market leaders face intense competition and significant execution risk.

Absolutely.

So as you, the listener, consider all this, maybe ask yourself. In an economy where AI promises immense productivity gains, but could also accelerate job displacement, and where the core consumer is clearly stretched thin, how much can a concentrated tech rally truly buffer against broader economic shifts? And maybe more fundamentally, what will be the ultimate AI dividend for the average person beyond just seeing some stock market gains?

A critical question for the future.

Indeed. Lots to think about there. Thank you for joining us on this deep dive.