The Philadelphia Semiconductor Index has entered a technical bear market, erasing $3.3 trillion in value and exposing a rare rift between JPMorgan and Morgan Stanley on what comes next.
The Philadelphia Semiconductor Index has entered a technical bear market, erasing $3.3 trillion in value and exposing a rare rift between JPMorgan and Morgan Stanley on what comes next.

The Philadelphia Semiconductor Index has entered a technical bear market, erasing $3.3 trillion in value and exposing a rare rift between JPMorgan and Morgan Stanley on what comes next.
The Philadelphia Semiconductor Index's 20% plunge into bear territory has split Wall Street's two biggest banks, with JPMorgan calling a summer buying opportunity while Morgan Stanley argues chip stocks won't regain market leadership.
"The selloff has been driven by momentum unwinding and sector rotation, not a deterioration in fundamentals," Mislav Matejka, head of global equity strategy at JPMorgan, said. His team sees the relative strength index approaching oversold territory, suggesting chip stocks may soon attract buyers.
The SOX fell 10% last week alone, its worst weekly performance since April 2025, closing at about 11,674 after touching a record 14,655 in late June. The decline has wiped $3.3 trillion from the global semiconductor sector. Memory-related names suffered the steepest losses, with Micron Technology Inc., Samsung Electronics Co. and SK Hynix Inc. all posting significant declines, while Nvidia Corp. showed greater relative resilience.
The divergence matters because it frames the second-half playbook for the $670 billion AI infrastructure spending cycle. If JPMorgan is right, the chip sector's 20% correction offers an entry point before hyperscaler capital expenditure — expected to total hundreds of billions this year — reignites demand. If Morgan Stanley is correct, the market's center of gravity has shifted permanently toward cloud giants that can monetize AI investments more directly.
Matejka's team argues that effective semiconductor capacity additions remain limited through 2028, meaning supply constraints should underpin pricing power even if demand growth moderates. The strategists also note that chip company earnings remain strong, with current valuations still supported by fundamentals. However, they caution that momentum-factor unwinding typically amplifies short-term volatility, and they remain skeptical about how quickly cloud giants can convert massive AI capital spending into profit growth.
Morgan Stanley strategist Mike Wilson's team takes the opposite view. While acknowledging that a technical bounce after a 20% decline would not be surprising, they argue that the chip sector's leadership position is unlikely to return in the second half. "Market gains are broadening to a wider range of industries," Wilson's team said. They see consumer discretionary and transportation stocks as potential new drivers of market upside. Within the AI trade, Morgan Stanley prefers hyperscale cloud operators — companies like Microsoft Corp., Amazon.com Inc. and Alphabet Inc. — which they say have stronger core businesses and greater ability to improve profitability through cost optimization.
The debate has been complicated by two developments. First, hyperscalers have outperformed chip stocks by about 30 percentage points over the past three weeks, reducing the relative value of rotating into cloud names now, Morgan Stanley acknowledged. Second, the emergence of Moonshot AI's Kimi K3 — an open-weight model with 2.8 trillion parameters priced at $15 per million output tokens versus $50 for comparable Western systems — has raised questions about whether the AI industry's reliance on expensive proprietary hardware will persist. Bank of America noted that Moonshot is improving model performance through better training methods rather than more advanced chips, a dynamic that could pressure semiconductor pricing assumptions if replicated at scale.
For investors, the split creates a tactical dilemma. Broadcom Inc., which Morgan Stanley rates Overweight with a $502 target, represents the custom ASIC thesis — the firm expects it to retain about 80% of Google's TPU business. Nvidia, trading at a premium multiple, faces questions about whether its data center dominance can withstand both competitive custom chips and the emergence of cheaper open-source AI models. The upcoming earnings reports from Alphabet, Microsoft, Amazon and Meta Platforms — the hyperscalers that buy the majority of advanced chips — will provide the next major test. Any sign of slowing orders or delayed deployments could deepen the selloff; confirmation of robust spending could validate JPMorgan's call.
This article is for informational purposes only and does not constitute investment advice.