Q2 Earnings: The 'AI Bubble' Will Accelerate As We Enter Definite Optimism

The article argues Silicon Valley is in a new “definite optimism” phase, prioritizing AI investment and higher CapEx over share buybacks, which could support a constructive read-through for US equities. Investors are expected to shift focus from near-term ROI to AI efficiency roadmaps, unit economics, and deployment plans in forthcoming earnings updates. Net impact is likely more sentiment- and positioning-driven than a specific immediate earnings catalyst.

Analysis

This is a regime shift from financial engineering to industrial buildout. The first beneficiaries are the supply chain bottlenecks: semis, interconnect, networking, liquid cooling, and power infrastructure should see the cleanest order-flow leverage because buyers will prioritize delivery capacity over price. In the next 1-3 months, the market will likely pay up for names with the best evidence of conversion from AI capex into booked revenue and margin hold—think NVDA, AVGO, ANET, VRT, ETN, and CEG—while low-growth cash-return stories in large-cap tech risk multiple compression if buyback support fades.

Second-order effects matter more than headline AI demand. If hyperscalers and platform companies redirect cash from repurchases into depreciation-heavy projects, reported free cash flow can lag even as operating income appears strong; that creates a trap for investors chasing near-term EPS beats without checking unit economics per inference token, per rack, or per deployed watt. The winners over 6-18 months are likely the companies controlling scarce inputs—GPU supply, power interconnect, grid equipment—because the capex race raises barriers to entry and reduces the odds of a quick price war.

The contrarian risk is that this becomes a crowded narrative before monetization is visible. If next earnings season shows capex accelerating but no corresponding improvement in AI revenue, attach rates, or productivity metrics, the market can flip from rewarding spend to punishing it within a quarter. The tell will be whether management teams keep guiding to larger AI budgets while shrinking buybacks; if so, investors may eventually demand proof of return, and the current rotation could reverse sharply in 1-2 reporting cycles.

Key Decisions for Investors

• Overweight a basket of AI infrastructure winners (NVDA, AVGO, ANET, VRT, ETN) versus XLC on a 1-3 month horizon; thesis is that capex visibility outruns monetization skepticism, with upside if order commentary stays firm into earnings.
• Pair trade: long SMH / short AAPL or MSFT as a relative-value expression of the capex-over-buyback regime; best entered on any post-earnings pullback, with the short leg protected if management re-accelerates repurchases.
• Buy call spreads on VRT or ANET for the next earnings cycle if backlog and margin guidance remain tight; risk/reward is attractive because infrastructure names should see valuation re-rating before software monetization is proven.
• Set a falsification trigger: if hyperscaler capex guides flatten for two consecutive quarters or AI revenue disclosure fails to improve, reduce exposure to the capex complex and rotate back toward cash-return/quality mega caps.