Capital & Industrial Strategy

JPMorgan has revised its 2026 tech bond issuance forecast upward to exceed half a trillion dollars, driven by hyperscalers — Microsoft, Google, Amazon, and Meta — continuing to finance data centre and GPU capacity expansions through debt rather than equity dilution. Bloomberg notes that investor fatigue is emerging but has not yet translated into spread widening sufficient to slow issuance. The structural preference for debt financing at this stage of the cycle reflects hyperscalers' confidence in long-run AI revenue but also their desire to preserve equity for strategic M&A optionality.

Simultaneously, Switch Inc.'s confidential IPO filing — confirmed by people familiar with the matter per Bloomberg — represents the equity market counterpart to this debt story. Data centre operators are packaging AI infrastructure exposure for public market investors who cannot access hyperscaler capex directly. The IPO pipeline in this segment is crowded; Switch's valuation premium will depend heavily on whether it can demonstrate contracted hyperscaler tenancy rather than speculative capacity.

The hedge fund Situational Awareness, whose forced unwind was characterised by CNBC as a clearing event for the broader AI trade, simultaneously committed $400 million to Source Foundry, a stealth-mode startup targeting chip manufacturing process innovation, according to WSJ. The strategic logic is a bet that the current AI chip architecture — dominated by Nvidia's CUDA ecosystem and TSMC's fabrication process — is vulnerable to disruption at the manufacturing layer rather than the design layer. This is a less-crowded thesis than the many AI chip design startups competing with Nvidia on architecture.

Details on Source Foundry's technology remain undisclosed — the company is described as stealth — so the $400 million commitment cannot be independently validated in terms of valuation or deal structure beyond what WSJ has reported. The bet is high-conviction and concentrated, which is notable given the fund's recent losses. It suggests Situational Awareness is doubling down on a long-horizon infrastructure thesis rather than retreating to liquid positions.

The week's software earnings created a stark bifurcation. Airbnb surged 15% after CEO Brian Chesky credited AI — specifically agentic and customer service automation — with reigniting growth, and explicitly committed to materially higher AI spending, per CNBC. Twilio similarly beat on record profitability and free cash flow, with CEO Khozema Shipchandler framing agentic AI as a structural tailwind for its communications infrastructure, even while acknowledging AI is still a modest direct revenue contributor today, per Bloomberg. Both stocks were rewarded for demonstrating that AI was changing the unit economics of their core business, not merely being piloted.

The broader SaaS cohort swung violently in both directions, as CNBC reports, reflecting investor difficulty in distinguishing which platforms have defensible AI integration from which are exposed to displacement by AI-native competitors or direct model interfaces. The key dividing line emerging is whether a company's data assets and workflow integrations create switching costs that AI agents reinforce — as Twilio argues for communications infrastructure — or whether AI commoditises the company's core value proposition.

Gulf sovereign and institutional investors are actively re-routing AI allocations toward Asian markets, citing relative political stability compared to the Middle East and Europe and supply-chain resilience, per Semafor. This is a meaningful capital flow signal: Gulf investors — led by Saudi Arabia's PIF, Abu Dhabi's MGX and Mubadala, and Qatari vehicles — have been among the largest non-US sources of AI infrastructure and model company funding over the past two years. A pivot toward Asia diversifies their exposure away from US-centric AI infrastructure and toward Southeast Asian data centres, Japanese robotics AI, and potentially Chinese open-source model companies.

This re-routing also intersects with the China AI story. CNBC reports that Chinese AI companies have made measurable progress closing the performance gap with US frontier labs. Semafor separately notes that US companies are quietly incorporating Chinese open-source models into commercial products. Together, these signals suggest the capital case for Asian AI exposure is strengthening precisely as US export controls and geopolitical restrictions make direct investment in Chinese AI more legally complex for Western investors — creating an arbitrage opportunity for Gulf capital operating under different regulatory constraints.