AIToday
AI Stocks & MarketsAI Business & IndustryFortune AIPublished: Aug 16, 2026, 19:01 JST3 min read

AI isn't one bubble—it's a 'rolling sequence' of them popping

AI isn't one bubble—it's a 'rolling sequence' of them popping

Key takeaway

  • Rather than a single AI bubble destined to pop, strategist Dhaval Joshi argues the market is experiencing a rolling sequence of sector-specific booms and busts—software, semiconductors, and silver have each experienced sharp rallies followed by crashes as investors repriced their assumptions about which companies can sustain high margins.

  • The pattern has held so far because capital has simply cycled from one inflated sector to the next, but a tightening in monetary policy, a sharp capex downturn, or a non-mild recession could break the cycle and trigger a broad selloff.

3 Key Points

  1. What happened

    Strategist Dhaval Joshi argues the AI market is experiencing rapid boom-and-bust cycles across sectors rather than a single bubble. Software stocks rallied on productivity hopes, then crashed as investors feared AI agents would destroy the SaaS subscription model; silver spiked as a data-center conductor but fell when investors realized the price surge was unjustified; semiconductor stocks surged on perceived pricing power, but are now unwinding as investors question whether chipmakers can sustain their high margins.

  2. Why it matters

    Joshi describes this as a "profit margin bubble" rather than an earnings bubble—the question is whether companies can maintain their stratospheric margins, not whether valuations fit current earnings. Hyperscalers' capital expenditure is already eating into free cash flow, with Google posting its first free cash flow negative result and Microsoft, Alphabet, Amazon, Meta, and Oracle projected to see capex overtake free cash flow by 2027. The rolling pattern has so far prevented a correlated market crash, but if monetary policy tightens or a non-mild recession hits, the entire sequence could unwind at once.

  3. What to watch

    Joshi forecasts AI capex will likely peak in late 2026 or the first half of 2027. He cites DDR3 RAM—a 20-year-old chip—as an example of the pattern: it has surged 600% in less than a year. He flagged crypto as a possible next candidate if AI and blockchains produce synergies, and emphasized the discipline is spotting which narrative inflates next and which supposed competitive "moat" turns out to be unsustainable.

Ask the AI about this article →

Context & Analysis

Joshi's framework rejects the binary question—"Is AI a bubble?"—in favor of a pattern-recognition approach: which sector's narrative is inflating today, and which will deflate tomorrow? The evidence he marshals is concrete: software crashed after investors realized AI agents threaten recurring revenue; silver prices nearly tripled before being reassessed as unjustified by fundamentals; semiconductor margins are now under scrutiny after investors initially priced in seemingly unlimited pricing power. His distinction between a "profit margin bubble" and an earnings bubble is central: the worry is not that current earnings don't justify valuations, but that the margins underpinning those earnings are unsustainable and will normalize sharply.

The structural condition enabling this rolling sequence is that each deflation has been isolated enough to avoid triggering a broad selloff; capital has simply rotated to the next inflated narrative. Joshi points to the real constraint on this cycle: highly accommodative monetary policy. If that changes—if real interest rates or real bond yields rise sharply, or if the capex cycle unwinds violently—the rotation could reverse into a correlated exit from risky assets. His forecast of capex peaking in late 2026 or early 2027 also hints that the end of the sequence may already be visible on the horizon.

FAQ

When does Joshi think AI capex will peak?
He forecasts the peak of AI capital expenditure would most likely be late 2026 or the first half of 2027.
What could unravel the entire rolling bubble sequence at once?
Three main risks: if real interest rates and/or real bond yields rose sharply, if the capex cycle unwinds very sharply, or if a non-mild recession hits. A fourth risk is a lack of what he calls market "complexity," a metric he adapted from mathematician Benoit Mandelbrot's research into complex adaptive systems.
Which sectors has Joshi identified as part of the rolling bubble pattern?
Software-as-a-service stocks (rallied on AI productivity hopes, then crashed), silver (spiked as a data-center conductor but fell when the price surge could not be justified), and semiconductors (rose on pricing-power expectations but are unwinding as investors question sustained margins).

Get the latest AI Stocks & Markets news every morning

For example, today's edition would include:

  • NVIDIA Leads 5 AI Stocks to Watch in SeptemberYahoo Finance AI · 1h ago
  • Palantir stock near fair value; AI data debate weighsYahoo Finance AI · 7h ago
  • CrowdStrike Stock Climbs as AI Security Demand SurgesYahoo Finance AI · 10h ago

AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.

Free · takes 30 seconds · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. Q&As are published on this page for other readers too.

Related Articles

Next articleAWS Growth Surge Gives Amazon Edge Over Microsoft in AI Cloud Race