- Surmount Markets
- Posts
- Microsoft's $400B Wipeout, Silver's 1980-Style Crash, and the AI Spending Reckoning
Microsoft's $400B Wipeout, Silver's 1980-Style Crash, and the AI Spending Reckoning
The Russell 2000 just staged its longest winning streak since 1996, while a 175-year-old glassmaker is suddenly the hottest name in AI infrastructure.

The era of rewarding AI potential just ended. This week, Wall Street sent an unmistakable message: show us the money, or watch your market cap evaporate.
Microsoft lost $400 billion in a single session despite beating earnings. Silver crashed 30% in its worst day since 1980. And the entire tech sector learned that spending $37.5 billion per quarter on AI infrastructure doesn't impress anyone if Azure growth slows from 40% to 39%.
Welcome to the AI Spending Reckoning.
Microsoft's $400B Reality Check
Microsoft technically crushed it on Wednesday: $81.3 billion in revenue, $4.14 EPS, both above estimates. By Thursday, the stock had plummeted 10%—its worst day since March 2020—wiping out roughly $400 billion in market value.

The culprit? A $37.5 billion quarterly capital expenditure bill, up 66% year-over-year. That's $150 billion annualized—more than most Fortune 500 companies are worth. CFO Amy Hood called it building "AI Super Factories." Investors called it unsustainable.
Here's what broke the market's faith:
Azure growth decelerated to 39% from 40%—missing whisper numbers around 41%
45% of cloud backlog tied to a single customer: OpenAI
Capacity constraints will persist through 2026 despite record spending
Guidance disappointed with Azure growth projected at 37-38% next quarter
CEO Satya Nadella tried to spin it positively, noting they added "nearly one gigawatt of total capacity this quarter alone." But the market heard something else: even burning $150 billion annually, Microsoft can't build data centers fast enough to meet demand.
The second-largest value destruction in U.S. corporate history wasn't about Microsoft's fundamentals—it was about a narrative shift. Investors are no longer rewarding the "potential" of AI. They want proof of returns, and they want it now.
The Great Metals Massacre
Friday delivered one of the most violent reversals in commodity history. Gold, which had been trading near $5,600 per ounce—a record—plunged 9% intraday. Silver dropped from $120 to below $85, a 30%+ crash marking its worst day since March 1980.

The trigger? President Trump's announcement that Kevin Warsh would be his nominee for Fed Chair, replacing Jerome Powell when his term expires in May.
Warsh, a former Fed governor (2006-2011) known for his hawkish stance and criticism of quantitative easing, represents everything the "debasement trade" was betting against. For months, investors piled into precious metals on the thesis that Trump would politicize the Fed, keep rates artificially low, and effectively devalue the dollar.
The Warsh nomination killed that narrative instantly. Markets interpreted it as Trump prioritizing Fed credibility over political expediency. The dollar rallied sharply, and the entire debasement trade—which had been one of 2025's most crowded positions—unwound in hours.
Even accounting for Friday's carnage, silver is still up over 150% from this time last year. But the speed of the reversal reveals how fragile momentum-driven trades become when the underlying thesis breaks.
The AI Infrastructure Paradox
Microsoft wasn't alone in spooking investors with massive spending plans. Alphabet announced it expects to spend between $175 billion and $185 billion in 2026 on AI infrastructure. The stock fell 4% as analysts questioned whether the industry is building overcapacity.

This is the paradox tech giants now face: spend too little, and you risk falling behind in the AI arms race. Spend too much without immediate revenue acceleration, and Wall Street punishes you for prioritizing market share over profitability.
The comparison to the late-90s fiber optic buildout is increasingly common. Back then, companies laid thousands of miles of "dark fiber" that went unused for a decade. The difference? Today's spending is funded by free cash flow from the world's most profitable companies, not speculative debt.

But that distinction may not matter if investors lose patience before the AI infrastructure pays off.
Other market moves worth noting:
Bitcoin crashed below $75,000—its first time under $80K since April
Equal-weight S&P 500 outperformed cap-weighted, signaling continued rotation away from mega-caps
VIX spiked above 18, reflecting heightened volatility
Chip stocks got no relief as Qualcomm issued weak guidance
The tech sector is now trading at a crossroads. For three years, any mention of AI was enough to drive valuations higher. That era is over. Welcome to the show-me-the-money phase.
Strategy in Focus: Analyst Buys
While tech giants grapple with trillion-dollar infrastructure questions, sometimes the smartest move is following the analysts who've actually gotten it right.
That's the premise behind Surmount's Analyst Buys strategy—now live through our Quiver partnership. Instead of treating all Wall Street research equally, this strategy uses Quiver's proprietary scoring system to weight forecasts based on analysts' historical accuracy. Better track record? Higher weight. Consistently wrong? Lower weight (or even negative).

The strategy analyzes all price targets issued over the past year, scores analysts based on their predictive accuracy, then invests equal-weighted in the top 10 largest companies (by market cap) with the highest weighted consensus. Monthly rebalancing keeps the portfolio fresh.
Current positioning: The portfolio skews toward large-cap tech and financials—names like Nvidia, Oracle, Broadcom, Microsoft, and JPMorgan. These aren't contrarian picks; they're high-conviction calls from analysts whose past recommendations have beaten the market.
This week's tech selloff is exactly the environment where systematic, data-driven approaches matter. Analyst sentiment isn't infallible, but it beats guessing.
Also live: The Rob Bresnahan tracker and Tim Moore tracker mirror congressional portfolios, rebalancing when new trades are reported. Moore's strategy has shown a 74.4% CAGR since his 2025 entry to Congress, with concentrated bets in cyclicals like Harley-Davidson, Ford, and American Airlines.
What to Watch
The AI infrastructure debate isn't going away. Every earnings call from here forward will face the same scrutiny: are you spending responsibly, or just spending?
Microsoft's experience this week set the template. Beat earnings, raise guidance, add capacity—and still lose $400 billion if investors think you're prioritizing scale over returns.
For precious metals, the Warsh nomination changed the game. Whether he actually governs as a hawk remains to be seen (he's recently softened his stance to align with Trump), but markets are pricing in Fed independence and higher rates.
And for everything else? Watch the rotation. Equal-weight indices are outperforming, small caps had their best January in years, and cyclicals are catching bids. The Magnificent Seven era isn't over, but it's no longer the only story.
— Surmount Markets Team