3 ways the pros are trading markets right now, including why JPMorgan downgraded semiconductor stocks

Trump’s China visit and reports that the U.S. approved Nvidia H200 chip sales to Chinese firms are supportive for tech, especially semiconductors, while Xi signaled a wider opening for business. BlackRock’s James Turner favored corporate credit over sovereign bonds, Standard Chartered downgraded semis after concentrated gains, and J.P. Morgan’s Madison Faller continued to favor a 60/40 portfolio with bonds, infrastructure, hedge funds and a barbell into U.S. and emerging markets. The piece is mainly strategic commentary rather than a direct catalyst, but it highlights rotation away from crowded chip names and toward income, diversification and EM exposure.

Analysis

The immediate winner is less the headline chip complex than the parts of the ecosystem that monetize regulatory easing without carrying single-name valuation risk. If H200 access expands beyond a narrow set of buyers, the first-order benefit accrues to NVIDIA, but the second-order trade is more important: Chinese AI capex can shift from “domestic substitute at any cost” to “best-available foreign silicon where allowed,” which pressures local accelerator vendors and may delay a full decoupling narrative. That said, this is still a policy-driven tradable window, not a durable regime shift; any renewed export-control tightening or licensing backlash could compress the uplift in days, while the broader demand benefit plays out over quarters.

BlackRock’s corporate-credit framing matters because the market has spent two years paying up for duration and quality while underpricing balance-sheet resilience in corporates. If growth holds up and disinflation continues, investment-grade credit can outperform sovereigns on a risk-adjusted basis through spread carry alone, but the real edge is that tighter underwriting and refinancings have reduced default convexity versus prior cycles. The main risk is that investors crowd into carry just as rate volatility returns; in that case, spread widening can overwhelm the yield advantage within weeks even if fundamentals remain sound.

The semiconductor downgrade is a useful contrarian signal: when a trade becomes a consensus proxy for AI and geopolitics at the same time, its beta tends to get misread as earnings strength. The cleaner way to express the theme is via selective exposure to supply-chain beneficiaries and market-structure names rather than the most crowded mega-cap winners, especially if Korean hardware names are vulnerable to a near-term mean reversion. Meanwhile, the 60/40-plus-shock-absorbers framework points to a barbell where short-duration bonds provide tactical carry and EM/resource exposure hedges any renewed inflation impulse, particularly if China stimulus or trade normalization lifts commodity-linked Asia and LatAm.

Key Decisions for Investors

• Trim NVDA into strength over the next 1-2 sessions; keep core exposure but reduce tactical overweight by 25-35% while the export-policy headline is fresh. Risk/reward favors taking profits before the market prices in a licensing bottleneck or renewed political pushback.
• Express a relative-value trade: long QQQ put spread vs short SMH call spread 4-8 weeks out. The goal is to monetize crowded semis positioning while protecting upside if AI remains intact but leadership broadens; best if semis underperform on any broader risk-on tape.
• Add investment-grade corporate credit exposure via LQD or short-dated IG ETFs over the next 1-3 weeks. Favor the front end of the curve where carry is attractive and duration risk is lower; stop if rates volatility re-accelerates or spreads gap wider by 15-20 bps.
• Rotate part of semiconductor beta into more diversified Asia exposure: long EEMA or VWO vs short a basket of the most extended semiconductor names for 1-3 months. This captures the potential shift from concentrated AI gains to broader EM participation if policy tailwinds persist.
• Use short-duration Treasuries or T-bill ladders as the portfolio shock absorber rather than adding more long-duration sovereigns. If the market is wrong on hawkish central bank expectations, the near-term upside is modest but the drawdown protection is meaningfully better.