- Coiled Spring Capital Macro Report
- Posts
- Coiled Spring Capital MR 12/14/25
Table of Contents
Introduction
Everything looks fine in the market… until suddenly it’s not. No one ever said this game was supposed to be easy. Just when the stars seem to align for a clean run into year-end—an unexpectedly dovish Fed, liquidity tailwinds, and the promise of “more cowbell”—the market gods remind us who’s really in charge.
What looked like a smooth glide path quickly turned into turbulence as the week closed, with growth leadership—especially semiconductors, the emotional center of the AI bull market—taking a hit.
The sequence started innocently enough: a soft ORCL print. Then AVGO, priced for perfection, delivered something less than perfect. Add the predictable “AI bubble is bursting” narratives, and suddenly the stocks that carried this bull market were in a full-scale retreat.
Under the surface, the ORCL story is even more interesting. Credit markets aren’t ignoring it: ORCL CDS has surged above its 2022 bear-market highs, raising questions about whether the company can generate enough return to comfortably service its growing debt load. The market is concerned about free cash flow turning negative for the first time since 1992—a datapoint that doesn’t fit neatly into the AI-everything narrative.
Whether these credit concerns are truly justified is not our domain, and we’re not going to pretend to be distressed-debt analysts overnight. What we can say is this: betting against Larry Ellison has rarely been a winning strategy. The man has spent more than four decades proving skeptics wrong, reinventing Oracle multiple times, and outmaneuvering competitors in cycles far more hostile than today’s.
So if the new consensus is that now—after all that—Oracle has finally overstepped its bounds? Sure. Got it. Let’s see how that plays out.
Bubble fears around the hyperscalers’ astronomical capex budgets are completely normal—and, frankly, part of the price of admission for anyone wanting exposure to the upside of a major tech thematic. Every technology super-cycle—PCs in the ’80s, the Internet in the late ’90s, smartphones and cloud in the 2010s—began the same way: soaring claims of unlimited TAM, paradigm-shifting productivity, and permanently higher multiples for the winners.
AI is hitting those same emotional notes almost verbatim. We’re hearing the familiar language of “a new economic era,” exponential efficiency, and trillion-dollar market expansions. As always, the narrative is moving ahead of adoption. In prior cycles, the rhetoric peaked years before the economics caught up: internet penetration lagged its hype by nearly a decade; cloud profitability followed on a similarly long fuse. AI is behaving no differently—massive capex, murky monetization, and early winners pulling forward years of expected growth. The story is accelerating faster than the business models.
But this cycle isn’t a carbon copy of the past. This time, the AI boom sits inside an entrenched megacap oligopoly with unprecedented balance-sheet strength, distribution, and data advantages. That structure reduces existential risk—these companies can afford to make mistakes—but it doesn’t eliminate cyclicality. Every super-cycle includes violent boom–bust–boom phases before the durable trend asserts itself.
Yes, there will be winners and losers. Yes, the bubble will eventually pop. But are we really pretending this all ends in 2026? The technology is advancing far faster than real-world adoption. Most industries haven’t integrated AI into their core processes, revenue models remain early, and the infrastructure for mass deployment—compute, energy, model operations—is still being built. Productivity gains are still isolated, not systemic. Regulatory scaffolding is in its infancy. Historically, that gap between capability and commercial integration is what defines the early innings, not the peak.
The U.S. government’s expanding involvement reinforces that we’re still early in the cycle. Washington is now simultaneously investing in the AI ecosystem—via subsidies, defense contracts, semiconductor incentives, and large-scale federal adoption—while constructing a regulatory framework to cement long-term U.S. dominance. None of this signals the end of a trend; it signals the foundation of one.
With all that said, the poster child of the AI boom, NVDA, continues to wobble. Despite a strong earnings print and easing China restrictions, the stock broke its head-and-shoulders neckline on Friday. That technical break has reignited a wave of bear narratives and shallow media commentary that the AI bubble is “popping.”
Maybe. Or perhaps NVDA is simply testing the 23.6% Fibonacci retracement level off the April ’24 low—a perfectly normal pullback within a still-intact long-term uptrend.
Another concern entering the conversation is that rising depreciation expenses from the hyperscalers’ massive AI build-outs could start to weigh on the cadence of buybacks and dividends—an important pillar of the megacap equity story.
Taken together—slowing revenue growth layered on top of surging capex, rising depreciation, and questions around near-term monetization—the recipe for some valuation compression is legitimate. That said, these companies are nowhere near bubble territory. Current multiples sit well below the ~80x forward earnings the Nasdaq commanded at the height of the dot-com era, and the fundamental backdrop today is materially stronger.
But we’re not alone in this view—the largest asset manager in the world is making the same case.
So where does this leave us? We repeatedly remind clients that rotation is inherently bullish, and Friday’s growth-stock selloff looked far more like year-end repositioning than a structural unwind. Capital flowed into areas of the market that historically outperform during periods of economic acceleration.
Remember, the Fed is cutting rates and is likely to continue doing so—especially with a new Fed president next year. As liquidity improves, many of the softening economic indicators should begin to stabilize and then reverse, creating a more supportive backdrop for cyclicals.
Take a look at Friday’s sector performance. What stands out? Yes, Staples led the tape, but materials, healthcare, transports, consumer discretionary, and financials all finished in the green—classic pro-cyclical leadership that typically emerges ahead of economic expansion. Meanwhile, semiconductors and high-growth names absorbed the bulk of the damage, which is precisely what rotation looks like under the surface.
While the Nasdaq absorbed most of the damage on Friday, underlying breadth told a different story. SPX breadth was only modestly negative, and importantly, there were five times as many stocks making new four-week highs as four-week lows. That’s not what deteriorating market internals look like.
Rotational days can be painful—especially for portfolios overweight growth—but they are a healthy feature of a durable bull market. It’s entirely possible we are witnessing a passing of the baton toward more cyclically sensitive areas of the market, particularly the sectors that have lagged during the AI-led advance. We highlighted healthcare in our 10/5 report as an attractive, uncorrelated Q4 long candidate. That call continues to play out: healthcare remains the top-performing sector in the SPX, outperforming the other GICS groups by nearly 10x (ex-Communication Services).
Remember, it’s a market of stocks, and while the unwind in the AI trade may feel unsettling, there is far more beneath the surface to remain constructive on than meets the eye. Over the coming weeks, we will outline which sectors are best positioned to assume leadership into the new year. Stay tuned.
Now let’s turn to the charts.
