Turkey stocks lower at close of trade; BIST 100 down 0.24%

Nvidia shares were supported by strong AI demand and a bullish earnings/outlook narrative, but the broader Turkey market finished weaker as the BIST 100 fell 0.24%. Trade was mixed at the index level: Pasifik Eurasia Lojistik (+6.61% to 258.00) led gains while Destek Finans Faktoring (-9.98% to 2,183.00) was the biggest decliner. Energy and FX were slightly firmer—October crude rose 0.19% to $82.39/bbl and Brent gained 0.53% to $87.40/bbl, while USD/TRY edged up 0.06% to 48.14.

Analysis

The real signal is that AI spending remains elastic enough to support upside revisions even after a huge installed-base run. That matters because NVDA is still trading more on forward supply allocation than on near-term unit growth; if demand is broadening, incremental upside should flow through mix and pricing rather than just higher volume. That keeps gross margin and operating leverage more durable than the street usually assumes after a strong print.

Second-order winners are the adjacent bottlenecks: HBM memory, advanced packaging, and foundry capacity should retain pricing power longer than the market expects if hyperscaler capex stays sticky. That argues for relative strength in supply-chain proxies such as MU, TSM, and ASML over the next 1-3 months, while legacy server CPU exposure remains the weak link. If buyers continue prioritizing accelerators, AMD’s near-term upside is more capped by ecosystem lock-in than by headline AI demand.

The contrarian risk is that consensus may be extrapolating demand growth while underweighting digestion risk. Over 6-18 months, the market will care less about “AI demand” and more about monetization per dollar of capex, export constraints, and whether hyperscalers keep funding the buildout at the same pace. The thesis breaks if sequential guidance slows, gross margin compresses, or capex commentary from MSFT/AMZN/GOOGL/META turns cautious.

Key Decisions for Investors

• Buy NVDA on post-earnings weakness or a 2-3% pullback; target a 1-3 month continuation move, but trim if the stock loses its 20-day moving average and next-quarter guide stops implying upward revisions.
• Pair trade: long NVDA / short AMD for 1-3 months to express continued platform-share concentration; best risk/reward if hyperscaler capex commentary stays firm and AMD fails to show faster ecosystem traction.
• Overweight second-order winners MU and TSM versus the broader SOXX basket for the next quarter; this is a cleaner way to express persistent AI supply tightness if you want less single-name gap risk than NVDA.
• Use call spreads rather than outright calls on NVDA for a 2-4 month horizon after the first post-event fade; upside remains, but implied volatility likely overstates follow-through after a strong print.