SOFTBANK CORP.

Company Overview

SoftBank Corp. (TSE:9434) is the Japanese operating company built around mobile telecom, broadband, enterprise IT, Yahoo/LINE media and commerce, and PayPay-linked financial services. It is not SoftBank Group Corp. (9984), the parent investment holding company. That distinction matters. The operating economics of 9434 are still driven mainly by a domestic telecom franchise with large digital adjacencies, not by mark-to-market gains on Arm or OpenAI.

The latest clean official annual base is FY2026, the year ended March 31, 2026, from the audited annual securities report filed on June 19, 2026. More recent data are market prices and company guidance, not newer official operating results. The next scheduled earnings release is August 5, 2026, which is a market calendar item, not financial data.

Growth has been real, but it has not shown up cleanly in per-share earnings. Revenue compounded at a mid-single-digit rate, while earnings compounded only slightly. The concrete drivers have been Enterprise solutions and PayPay-linked finance, not the mature mobile core. Enterprise revenue rose from ¥702 billion in FY2022 to ¥1,003 billion in FY2026. Financial segment revenue rose from ¥51 billion in FY2022 to ¥405 billion in FY2026. PayPay reached 73 million registered users and ¥19.4 trillion of consolidated GMV in FY2026. By contrast, the consumer segment remains stable rather than explosive.

At the current market cap, that implies an owner-earnings yield of roughly 5.7%-6.7%. That is only modestly better than the P/E-based earnings yield, which tells you something important: this is still a capital-intensive telecom-plus-platform business, not a software business hiding inside a telco shell.

Capital efficiency is good, but not cleanly high quality. ROE has ranged around 19%-27% in recent years, but that is leverage-assisted; the equity ratio is only 16% in FY2026. ROIC looks more like 12%-13%, which is healthy, not exceptional. Incremental capital in enterprise solutions and fintech appears attractive. Incremental capital in the consumer telecom core looks adequate, not outstanding. This is a durable franchise, but I would not call it a high-return compounder.

How the Company Makes Money

The economic engine is still the access network. Consumer mobile and broadband provide the stable cash flow. Everything else either raises customer lifetime value, broadens the ecosystem, or creates growth options.

Profits actually come from the consumer telecom business. The newer segments matter because they change the mix: enterprise solutions are higher value, media and commerce deepen user engagement, and PayPay increases wallet share. But the base case still stands or falls with telecom cash generation.

This has been a good business for familiar reasons. Spectrum, network density, billing systems, retail and service distribution, and enterprise relationships are expensive and slow to replicate. SoftBank also has a useful multi-brand structure: SoftBank for premium, Y!mobile for value, LINEMO for online-only. That lets it segment price without fighting a one-brand price war across the entire base. On top of that, it has unusually large digital touchpoints for a telecom operator: 32.01 million smartphone subscribers, 8.50 million broadband subscribers, 83 million average monthly Yahoo-related users, 100 million LINE monthly active users in Japan, and 73 million PayPay registered users.

The enterprise side is stronger than many investors give it credit for. The company says it does business with 92% of Japanese listed companies with revenue above ¥100 billion. That distribution advantage matters. It gives SoftBank a real channel for cloud, security, AI, and data products that a pure startup does not have.

The caveat is equally important: the non-telco growth engines are not all fully owned. That means consolidated revenue can grow faster than the value that actually accrues to common shareholders.

Why the Stock Fell

This is a de-rating, not a business collapse. As of June 26, 2026, the stock closed at ¥203, about 18% below the FY2025 high of ¥247.9 and effectively at a fresh 52-week low. The market has not been reacting to a broken balance sheet or a collapse in audited earnings. It has been taking out a premium multiple.

The basic pattern is straightforward. FY2026 audited results were solid but not dramatic: revenue rose 7.6%, operating income 5.4%, and owner net income 4.7%. For a stock still trading around 18x earnings, that was not enough. The FY2027 company guide is the bigger issue: revenue is guided up 6.6% and operating income up 5.5%, but net income attributable to owners is guided up only 1.7%. That tells the market that growth is either arriving slowly, being absorbed by costs, or leaking out to minority interests.

The May earnings release also disappointed on the quarter relative to analyst expectations. Market-data estimates had quarterly EPS around ¥1.60; actual quarterly EPS was about ¥1.31. That is not thesis-breaking, but it reinforced the idea that the stock had been carrying too much optimism about AI, PayPay, and margin progression.

One nuance matters a lot for a non-specialist reader: many late-June headlines using only the word “SoftBank” were about SoftBank Group, the parent holding company, not SoftBank Corp. 9434. OpenAI timing and holding-company AI leverage are direct issues for 9984. They matter for 9434 only indirectly. Some of the sentiment spillover is real, but it should not be confused with the operating economics of this company.

What the Market Is Assuming

(a) One-time / cyclical / sentiment-driven factors.

• ASKUL’s ransomware-related outage may be a sign of broader execution weakness, not just a one-off hit.
• The Q4 EPS miss may signal deteriorating earnings quality.
• “SoftBank” AI headlines may keep dragging the stock even when they relate more to the parent than to 9434.
• Electricity and device-driven revenue growth may be masking weak underlying service growth.

(b) Medium-term business headwinds.

• Consumer mobile can grow revenue only by spending more on commissions, promotions, and device support.
• Enterprise AI and sovereign cloud are promising, but earnings contribution is later than the market hoped.
• PayPay and LY growth may raise consolidated profits without proportionately lifting profit attributable to common shareholders.
• Interest costs and growth investment can keep net income growth below operating-income growth.
• The stock still carries a premium to NTT and KDDI despite slower per-share earnings growth.

(c) Potential long-term structural threats.

• Japanese mobile becomes a permanently low-growth utility with weak pricing power.
• The best growth assets sit in partly owned subsidiaries, limiting per-share compounding.
• Repeated security or governance failures at LY/LINE damage trust in the ecosystem.
• AI and data-center investment earns low returns and drags group ROIC.

Temporary or Structural?

The numbers say the business is not deteriorating in the way the stock chart implies. Audited FY2026 revenue rose to ¥7.04 trillion, operating income to ¥1.04 trillion, owner net income to ¥550.8 billion, and operating cash flow to ¥1.39 trillion. The main time issues are visible and quantifiable: ASKUL’s outage hurt Media & EC, and consumer mobile margin is being squeezed by customer-acquisition costs. Neither is the same thing as franchise impairment.

Reality check on the headline worries:

• Consumer revenue rose from ¥2.95 trillion in FY2024 to ¥3.02 trillion in FY2026. Mobile revenue still grew 1.1% in FY2026.
• Smartphone subscribers reached 32.01 million by March 2026. Underlying mobile revenue has been positive year on year since FY2023 Q3.
• Media & EC revenue still rose to ¥1.67 trillion in FY2026 despite ASKUL. The disclosed ASKUL H2 hit was about ¥81.5 billion of revenue and ¥32.6 billion of operating profit.
• Enterprise revenue rose from ¥702 billion in FY2022 to ¥1.00 trillion in FY2026. Business-solution revenue alone rose 13.2% year on year in FY2026, with recurring revenue up 10.0%.
• Financial revenue rose from ¥325.5 billion in FY2025 to ¥404.5 billion in FY2026, and operating income more than doubled to ¥86.3 billion.
• Operating cash flow has remained strong for years: ¥1.22 trillion in FY2021, ¥1.16 trillion in FY2022, ¥1.37 trillion in FY2025, and ¥1.39 trillion in FY2026.

The one issue I would treat as genuinely structural is value capture, not moat collapse. The fastest-growing parts of the group are not fully owned, so common shareholders do not fully receive the upside. That does not break the operating business. It does cap per-share compounding.

Is the Market Wrong? By How Much?

Time-as-a-moat test. If I had today’s market cap in cash, roughly ¥9.7 trillion, I still could not rebuild this business quickly.

The moat is real. The harder question is whether the stock is cheap enough. Even after the selloff, SoftBank Corp. still trades at a premium to major domestic peers on the latest official annual numbers: NTT is around 12.5x earnings and KDDI around 14.8x, versus roughly 18x here. So the market has not thrown away all optionality. It is still paying for some combination of PayPay, enterprise, and AI upside.

This valuation is FY2026-audited based. I use current market price only for the comparison, and I use FY2027 company guidance as a cross-check, not as the primary valuation input.

No separate net-debt adjustment is needed in that bridge because the owner-earnings base is already after interest and attributable to equity holders. The debt still matters for risk, but not as an additional deduction from equity value in this framework.

At today’s price, the stock implies an owner-earnings yield of roughly 6.3% on my base estimate. I think fair is closer to 5.9% for a stable telecom-platform operator with real moat but only moderate per-share growth. That means the market is probably a little too pessimistic, but only by about ¥0.6 trillion, or roughly ¥13 per share, in the base case. That is not a major dislocation. It is a modest one.

This is an intrinsic value estimate, not a price target.

Key Facts, Estimates, and Judgments

Moat & mispricing score: 5/10. The moat is real: spectrum, network density, multi-brand segmentation, enterprise reach, and large digital touchpoints are hard to replicate. The market is probably too negative if it thinks ASKUL noise, current customer-acquisition cost pressure, or late-June “SoftBank” headlines mean the franchise is impaired. But the market is right about two things: this is not a clean high-return compounder, and the best growth assets are not fully owned. My bottom line is TIME, not ESSENCE—but the valuation only gives you a modest margin of safety, not a glaring one.