Microsoft Rises, Google Tanks: Why Investors Are Polarized On Big Tech

Alphabet

Across the shows: 23 speakers in other episodes, 20 bullish and 4 bearish calls, past 30 days

• Gil Luria · D.A. DavidsonBullish

  Google Cloud's doubling growth, driven by TPU sales and Anthropic, is strong; the problem is elsewhere.

  Google Cloud revenue roughly doubled, to a run rate Luria calls very large. Part of that comes from TPU sales and a lot from momentum at Anthropic, which is decelerating but still adding substantial revenue. Much of that flows to Google Cloud as referral and compute fees. Luria says that if Google had only reported cloud results, the stock would be up. The stock drop was caused by the core advertising business instead.

  “It's not the cloud business. Cloud business is doing great... Part of it is TPU sales now. A lot of it is the momentum in Anthropic... So that part of the business was great.”at 7:43

• Gil Luria · D.A. DavidsonUncertain

  Google's abrupt search ad slowdown may reflect chatbot substitution; one quarter, so no need to panic.

  Google's core search ad business, the largest source of its profit, slowed abruptly and unexpectedly this quarter, despite management pointing to tough comparisons. Investors are now asking whether shifting user behavior toward ChatGPT, Claude and Gemini is hurting search, which was the original fear when ChatGPT launched. Luria says it is too early to tell from one quarter and that investors don't need to panic. But the concern is top of mind because search ads drive most of Google's profit, and Google's AI-era ad products may not be enough to offset lost search share in the long run.

  “If people stop Googling and start GPTing, that means less ad sales. And they've been able to assuage those concerns for the last couple of years... But this quarter, it appeared to abruptly slow down.”at 8:32

• Gil Luria · D.A. DavidsonBearish

  Google's rising AI capex and negative free cash flow will force more capital raises, pressuring its yields and share price.

  Google's free cash flow is negative again at about -$9.1B this quarter, with capex above $51B. Luria reads Google's guidance that capex will rise next year as implying a move from about $200B to $300B, which is a lot of capital to raise. Google will therefore need to return to capital markets if it keeps this pace. Investors can still fund it, but borrowing and share issuance add pressure on the yield Google must pay and on the share price. Google's move to offload capacity because it can't meet demand also means lower margins, which feeds concerns about the core business.

  “Yes. Because it means that they'll have to go back to the capital markets if they keep it this way.”at 12:04

Microsoft

Across the shows: 22 speakers in other episodes, 18 bullish and 1 bearish calls, past 30 days

• Gil Luria · D.A. DavidsonBullish

  Microsoft's clean beat-and-raise, with Azure accelerating to 48%, explains the stock's jump and reflects strong fundamentals.

  Microsoft beat and raised. Azure grew 48%, up from 43%, on a business of more than $120B, and margins are better than peers' even at that scale. The non-AI business also performed well enough to drive the overall revenue beat, earnings beat and above-expectation guidance. Microsoft is also selling more enterprise, infrastructure and application software and raising Windows and Office tiers, using AI to do so. That clean beat-and-raise is why the stock is up.

  “So not only was the Azure business great, it accelerated from 43 to 48%... This is a $120,000,000,000 plus business. Those are tremendous growth rates at better margins.”at 3:04

• Gil Luria · D.A. DavidsonBullish

  Microsoft's neocloud overflow capacity and long-term commitments let it limit capex and stay free-cash-flow positive.

  Unlike Amazon, Google and Meta, Microsoft does not need to build all the data center capacity its customers need. It makes long-term commitments to a significant portion of its build, then rents extra compute from neoclouds on short-term contracts when demand exceeds that. This keeps capex lower. Combined with a more profitable core business, Microsoft stays free-cash-flow positive. That means it doesn't have to raise money in capital markets, which Luria says makes it look better to investors. He adds that Microsoft's approach is more responsible even though all the hyperscalers are spending heavily.

  “So what Microsoft does slightly different, it's different than Amazon and slightly different than Google, is they've been relying on overflow capacity from Neo Clouds. So they don't need to build all the data center capacity to meet their customer needs.”at 4:14

Meta Platforms

Across the shows: 60 speakers in other episodes, 47 bullish and 33 bearish calls, past 30 days

• Gil Luria · D.A. DavidsonBullish

  Meta's ad growth accelerating to the high twenties shows the ad market is healthy and Meta is gaining share from Google.

  Meta accelerated growth to the high twenties last week. Its digital advertising business is growing more than twice as fast as Google's, and it is gaining a lot of share. That shows the overall ad market is not the problem. The concern about slowing ad growth is specific to Google's search advertising.

  “Meta accelerated growth to the high twenties last week. Meta is now growing more than twice as fast in the digital advertising business than Google. They're gaining a lot of share.”at 9:18

Oracle

Across the shows: 20 speakers in other episodes, 3 bullish and 19 bearish calls, past 30 days

• Gil Luria · D.A. DavidsonBearish

  Oracle's cut to its planned raise and turn to vendor financing signals that capital for AI buildouts is less available.

  Oracle changed its financing plan to raise $20B in total, split between $10B of equity and $10B of debt. Luria argues this is not because Oracle needs less capital, since it still needs a lot. It is because Oracle must go to plan B: bring-your-own-chip arrangements and vendor financing. He reads this as a sign that capital is not as available as it used to be, and that the pressure reaches past the hyperscalers to smaller players such as CoreWeave.

  “Oracle talked about the fact that they've changed their plans... That's not because they don't need a lot of capital. It's because they have to go to plan b.”at 15:53

US rates & the Fed

• Ed Elson · Prof G Media

  The Fed's hold may be political, since its stated strategy of raising rates to fight inflation contradicts the case that hikes won't help.

  Some argue the Fed held rates the week before the midterms for political reasons, given the president's opposition to hikes. Elson hears the economist argument that raising rates won't reduce oil prices, which are driven by Iran, so the Fed should not respond to energy-driven inflation. But the Fed had already raised rates, and it had told markets that this was its strategy. Elson says that argument doesn't hold up, which makes him suspect the hold was political. He acknowledges he is not certain.

  “But they already did raise rates. And so they've already told us we think this is gonna be our strategy... I don't fully buy it, and it makes me think maybe this was a political decision.”at 24:52

• Michael Gapen · Morgan Stanley

  The Fed's urgency message clashes with holding rates; if it wants its 2% target taken seriously, it must act.

  Gapen says the Fed frames the 2% target as central and says it will hit it, but he notes that actions matter more than words. The Fed describes the economy as growing strongly, with inflation above target, yet it waited. Gapen thinks there are reasons to wait, but the chair did not explain them. If the target and the urgency are as important as the Fed says, Gapen asks why it waited, and says this creates some conflict in the message.

  “Actions have to speak. Words are important, but actions ultimately are most important... If all of this is that important and that urgent and the economy is this strong, why did you wait?”at 28:17

• Michael Gapen · Morgan Stanley

  The Fed will hike again in December and March, with the pace of tightening slowing over time.

  Gapen expects the Fed to raise rates in December and then for a third time in March. He thinks the Fed could also reach the same restriction by early next year, and that this would deliver about the same tightening into 2027 and 2028. Growth is solid, the labor market isn't overheating, and inflation is sticky but has come down. Given that, the committee is making a modest recalibration, with rates going higher but not necessarily at consecutive meetings. The Fed has time to assess whether energy prices are feeding into other prices through second-round effects.

  “We think they will raise rates again in December and then for a third time in March.”at 30:18

• Michael Gapen · Morgan Stanley

  Warsh's reformer approach and reticence may cause messaging conflict, but new chairs usually settle in within six to twelve months.

  Warsh speaks less and gives little explanation, which Gapen finds frustrating for markets, although his restraint is skillful. Gapen also says Warsh sees himself as a reformer who believes the institution has drifted from its core principles over the past decade or two and wants to steer it back. That creates a recipe for messaging conflict with the 15 to 18 FOMC members who do things the old way, and could make for mixed messages. Gapen adds that every new chair has indigestion, and it usually takes six months to a year to develop a style and relationship with markets.

  “the other backdrop is he's coming into the position as a reformer. He views himself as a reformer. He thinks the way that the institution has worked over the past decade or two is incorrect, and he wants to change things.”at 31:40

US inflation

• Michael Gapen · Morgan Stanley

  Much of current inflation is supply-driven by energy and tariffs, so rate hikes cannot fix it directly.

  Gapen argues the Fed's hold reflects the view that much of today's inflation comes from supply shocks it cannot fix, because raising rates doesn't bring more oil to market. Without the tariffs, the Middle East conflict and the Russia-Ukraine crisis, inflation was decelerating and could have fallen close to 2% in Europe and the US. Monetary policy is poorly suited to every supply shock. The Fed's aim is to lean against second-round effects, where energy costs spill into other goods, so it hikes less than headline inflation would suggest.

  “a lot of this inflation is driven by supply side forces that we can't really fix. In other words, the Fed raising rates doesn't bring more oil to markets.”at 22:38

• Ed Elson · Prof G Media

  Inflation may never return to the Fed's 2% target, since it has been far off for years.

  The Fed emphasized doubling down on its 2% inflation target, but inflation has been well above that for several years. Elson is starting to believe the target will never be reached and that the economy has settled into a world where inflation stays far from it. He notes the Fed keeps saying the target is important but has been miles off for years. He asks whether the target is realistic and whether the Fed takes it seriously, and leans toward doubting it.

  “I'm starting to think that that's just never gonna happen... I'm starting to think we just live in a world where inflation is never close to the target... we've been miles off for years.”at 26:52

• Michael Gapen · Morgan Stanley

  Core PCE should drift down to about 2.6% by end of 2027, still above the Fed's target.

  Gapen treats core inflation as the best measure of underlying price trends, since headline is being driven by energy. He expects core inflation of about 3.1% year on year in the near term and around 3% by year end 2026. Residual seasonality, meaning firms raise prices early in the year, will keep inflation elevated and sticky through roughly the end of Q1 2027. After that, a clearer downward trend should bring core PCE to about 2.6% by the end of 2027. That is directionally toward the target but not all the way to it.

  “We think core PCE finishes next year at around 2.6.”at 30:02

AI infrastructure & data centers

• Gil Luria · D.A. Davidson

  External funding for the AI buildout is finite, and some capital markets may already be seizing up.

  Capital for the AI buildout is finite. Meta said in its filing that it needs to raise capital before it runs out, and SpaceX, Anthropic and Google have all raised large sums recently. Equity supply is limited, and debt investors are the other key source of funding. Not everyone will get a piece, and some parts of the capital markets may already be seizing up, as seen in higher CDS spreads and very high yields on some outstanding bonds. Microsoft is fine since it doesn't need capital, and Amazon, Google and Meta are still investment grade and could probably raise as much as they want up to a point. Luria's bigger worry is everyone else.

  “but it is finite. In fact, Meta pointed this out in their filing last week. They said, hey, look, we need to go get capital before it's all gone... So at some point, not everybody's gonna get a piece, and some parts of the capital markets may actually already be seizing up.”at 14:51

AI & jobs

• Ed Elson · Prof G Media

  Microsoft's 27% buyout would leave headcount below pre-pandemic levels, a structural AI effect rather than an overhiring correction.

  Microsoft is offering voluntary buyouts to 27% of its global workforce, after nearly 5,000 layoffs in July, while its capex keeps rising. Microsoft added almost 60,000 employees in the two years after COVID, so a cyclical overhiring correction is one explanation. But if 27% of staff actually leave, total headcount would fall below pre-pandemic levels. That looks structural and tied to AI, whether AI is replacing workers directly or through the cost of AI spending. Elson concedes AI's effect on the overall US job market has been negligible, but says it is beginning to have an impact in tech.

  “If Microsoft actually eliminates 27% of its workforce like they're suggesting, then their total headcount will actually be lower than what it was before the pandemic... That actually looks like a structural consequence of AI.”at 34:43