SoftBank: Get In Before The OpenAI IPO

SoftBank’s net asset value reached a record ¥40.1T (~$260B) last quarter, driven by AI-linked holdings including Arm and OpenAI. The article highlights SoftBank’s unique exposure to generative AI, its aggressive infrastructure investments, and the potential catalyst of an OpenAI IPO. Despite recent share weakness, the setup is presented as an attractive entry point if AI sector momentum continues.

Analysis

The market is still mispricing SoftBank as a levered conglomerate when the real economic exposure is now a concentrated call option on the AI capex cycle. That matters because the next leg of value creation is not just model adoption; it is the monetization of compute, packaging, edge silicon, and infrastructure bottlenecks, where Arm sits closest to the toll booth. If AI spend continues to compound, ARM can rerate not merely on unit growth but on a higher terminal royalty multiple as investors extend the duration of its cash flows.

The second-order winner is anyone upstream of AI deployment, but the more interesting dynamic is who gets squeezed: cloud and software incumbents that must fund massive capex without matching near-term revenue uplift. That creates a split market where infrastructure-linked names outperform while application-layer names with weak monetization get punished. SoftBank’s private-market exposure also creates a feedback loop: a successful OpenAI listing can refresh the valuation of adjacent private AI assets, which could trigger mark-to-market gains beyond the obvious headline beneficiary.

The main risk is timing, not thesis. The OpenAI IPO catalyst can be delayed for quarters, and AI infrastructure enthusiasm can cool fast if hyperscaler spend moderates or if investors start demanding proof of ROI from model deployment. In that scenario, SoftBank’s NAV story becomes less a compounding machine and more a financing story, with public holders left owning an illiquid basket at a premium to realizable value.

Consensus seems to assume the trade is already crowded, but positioning in the public proxies may still be under-owned relative to the structural shift. The underappreciated angle is that weakness in SFTBY can be an opportunity to buy optionality on ARM-linked AI monetization with less direct private-market opacity. The setup is strongest over 6-18 months, where multiple expansion can outrun earnings revisions if AI capex remains elastic.

Key Decisions for Investors

• Long ARM on weakness over the next 2-4 weeks; use a staged entry and target a 3-6 month rerating if AI capex guidance stays firm. Risk/reward is attractive because downside is capped by existing licensing economics, while upside comes from multiple expansion if investors extend AI duration.
• Buy SFTBY as a medium-term levered AI basket only on pullbacks, with a 6-12 month horizon. Size smaller than a direct operating company because the discount/premium to NAV can widen before narrowing; the trade works best if OpenAI IPO timing becomes clearer.
• Pair long ARM / short a slower-growth mega-cap software name with heavy AI spend but weaker near-term monetization over 3-6 months. The idea is to own the infrastructure toll collector and fade the spenders whose returns on AI capex are less visible.
• Consider a call spread on ARM into the next AI infrastructure update cycle, 3-6 months out. This expresses upside participation while defining risk if the market temporarily de-rates AI beta.
• If OpenAI IPO chatter intensifies, look to take profits on any SFTBY pop into that event rather than hold for the listing itself. The first move is usually anticipation-driven; the better entry for the next leg is after underwriting and valuation details emerge.