How Masayoshi Son's Nvidia Exit Reveals His Bold AI Investment Strategy

When SoftBank divested its entire Nvidia position last month, netting approximately $6 billion in proceeds, market observers initially questioned whether the legendary Japanese investor Masayoshi Son had lost faith in the chipmaker. But recent comments from Son himself tell a completely different story—one that illuminates a far more ambitious and calculated investment thesis.

The Reluctant Exit: Son Breaks His Silence on Nvidia

At the Future Investment Initiative summit in Tokyo earlier this month, Masayoshi Son was asked directly about SoftBank’s unexpected exit from its Nvidia holdings. His response was surprisingly candid and revealing. Son emphasized his deep respect for Nvidia CEO Jensen Huang and the company’s technological leadership. Perhaps most tellingly, Son expressed that he hadn’t wanted to sell “a single share,” even joking that he was “crying” over the divestment decision.

These comments suggest that SoftBank’s exit wasn’t driven by skepticism about Nvidia’s long-term prospects, but rather by strategic necessity—a painful but deliberate choice to reallocate capital toward even more transformative opportunities across the AI ecosystem.

The Grand Vision Behind SoftBank’s Portfolio Reshuffling

To understand why Masayoshi Son made such a counterintuitive move, it’s crucial to examine the sweeping AI infrastructure investments SoftBank has undertaken over the past year. Following President Trump’s January announcement of Project Stargate, SoftBank emerged as a key partner in what promises to be a transformative venture. This joint initiative with Oracle and OpenAI targets $500 billion in AI infrastructure investment across the United States over the next four years alone.

Beyond Project Stargate, Masayoshi Son’s investment footprint has expanded dramatically across the AI value chain:

• Direct OpenAI Investment: SoftBank committed up to $40 billion to OpenAI, with $22.5 billion expected by year’s end, bringing the total to $30 billion
• Chip Manufacturing: A $2 billion stake in Intel, coupled with Nvidia’s subsequent $5 billion commitment to the chipmaker
• Semiconductor Design: Approximately $12 billion deployed to acquire Ampere Computing and ABB’s robotics division
• Infrastructure Breadth: Strategic positions spanning GPUs, data centers, CPUs, custom-designed application-specific integrated circuits, and robotics

This diversified approach reveals Masayoshi Son’s fundamental investment philosophy: don’t concentrate bets on a single company or technology, no matter how dominant. Instead, build exposure across the entire AI ecosystem to capture value at multiple nodes of the industry stack.

Why the Nvidia Divestment Made Strategic Sense

The irony is that SoftBank’s exit from Nvidia wasn’t a vote of no confidence—it was a resource allocation decision. By crystallizing gains from its Nvidia position, Masayoshi Son effectively rebalanced a portfolio that was becoming too concentrated in a single chipmaker at a critical inflection point for the broader AI industry.

This move allowed SoftBank to:

1. Lock in substantial returns from years of Nvidia appreciation
2. Deploy capital into complementary AI infrastructure plays
3. Reduce portfolio concentration risk at a time when competition in the chip market is intensifying
4. Position itself as a diversified AI investor rather than a single-stock bet

The Shifting Landscape: Why Competition Matters More Than Ever

As of mid-December, Nvidia traded at a forward price-to-earnings ratio of 24—a level the company hadn’t sustained since its January plunge. This valuation compression reflects genuine structural shifts in the competitive landscape:

Alphabet’s Tensor Processing Units represent a direct challenge to Nvidia’s GPU dominance, with major cloud providers increasingly developing custom chips to reduce dependency on external suppliers. Meanwhile, Advanced Micro Devices continues to gain ground in GPU markets, and various tech giants have deployed proprietary application-specific integrated circuits for specialized workloads.

Yet despite these headwinds, the underlying fundamentals remain compelling. Nvidia maintains substantial moats through software ecosystem depth, performance leadership, and established relationships with hyperscalers investing heavily in AI infrastructure. Recent alliances with Anthropic, Palantir Technologies, and Nokia suggest the chipmaker continues to expand its addressable market while adapting to emerging competitive pressures.

The Masayoshi Son Lesson: Discipline Over Momentum

Masayoshi Son’s decision to exit Nvidia while simultaneously praising the company embodies a sophisticated investment principle: the best investments sometimes require knowing when to take profits and redeploy capital strategically. Rather than holding indefinitely or selling due to loss of confidence, Son recognized that SoftBank’s capital could generate greater value by participating across multiple layers of AI infrastructure buildout.

For investors watching this strategy unfold, the lesson is clear: the AI infrastructure thesis remains compelling, but success requires flexibility, diversification, and willingness to make difficult portfolio decisions in service of a larger strategic vision. Masayoshi Son’s recent moves suggest that the most profitable path forward may involve exposure not to any single “AI winner,” but to the foundational platforms and infrastructure enabling the entire ecosystem’s growth.