Rakuten Group (4755.T) — Re-Valuation Using Japanese Sources Only: Methodology Lesson

by psind707

Published: May 5, 2026 | Price: ¥760 (May 4, 2026) | Rating: HOLD (revised down from LONG) | PWFV: ¥725 | Time horizon: 12-24 months

TL;DR — Why I Revised My View

Reading only Japanese sources (Nikkei, IR documents in Japanese, Minkabu, Yahoo Finance Japan, Kabuyoho, IRBank, Commerce Pick, Business Network, ICT Research, K-tai Watch) reveals a materially more skeptical picture than English-language coverage of Rakuten Group.

My prior PWFV of ¥908 (English-source view) revises down to ~¥725 (Japanese-source view).

The stock at ¥760 looks fairly valued to slightly overvalued, not undervalued. Rating revised from LONG to HOLD. This is a methodology lesson as much as a stock call.

Critical Facts Emphasized in Japanese Sources but Downplayed in English

1. Net Loss WIDENED to ¥177.8B — 7th Consecutive Year of Net Losses

Per Nikkei and Minkabu:

• FY2025 net loss: ¥177.8B (vs ¥162.4B prior year — loss grew ¥15.4B)
• This is the 7th consecutive year of net losses since 2019
• FY2024 had a one-time remeasurement gain of ¥106.9B (from rebalancing of subsidiary stakes); without this special item, comparable Non-GAAP OI grew from ¥7B to ¥106.3B (15x), but bottom line still bled

Most English coverage led with "record EBITDA" and "operating profit second year." Japanese coverage led with "29 years of revenue records but 7 years of losses" (29期連続増収も7年連続赤字) — completely different framing.

2. Q4 Alone Deteriorated Severely

Per Minkabu's earnings analysis:

• Q4 net loss: ¥26.5B (vs ¥12.0B prior year Q4 — doubled)
• Q4 operating margin: 1.8% vs 15.7% prior year — collapse
• Q4 was the bad quarter that drove the -11% post-print stock reaction

3. NO FY2026 Numerical Guidance Disclosed

Per Minkabu (regulatory filing language):

25年12月期の業績見通しは未開示

Management refused to give specific FY2026 revenue/profit numbers, only directional language ("significantly expand operating profit"). This is a meaningful signal of management uncertainty about Mobile capex absorption and refinancing costs.

4. NO Dividend — for FY2025 or FY2026

前期の期末一括配当を見送るとし、今期の年間配当は未定

No dividend declared for FY2025 (skipped year-end). FY2026 dividend "to be determined" — i.e., probably zero again. This was barely mentioned in English sources.

5. 2026 Capex Jumps to ¥200B+ vs ¥62.9B in 2025 (~3.2x increase)

Per Business Network and Rakuten IR (Japanese):

2026年度は、2,000億円強の設備投資を計画（前年実績629億円）

This is a massive jump — Rakuten Mobile is shifting back to heavy investment phase, not harvest. The "self-funding" promise + ¥200B capex implies meaningful cash drag in 2026 even if EBITDA grows.

6. Nikkei Explicitly Criticized "Rural Neglect" as Enabling Mobile Profitability

Per Nikkei:

楽天モバイル、設備投資少なく黒字に寄与「地方を軽視」との指摘も

Translated: "Rakuten Mobile achieved profitability via low capex — but criticized for 'neglecting rural areas.'" This is an embedded competitive risk — KDDI/SoftBank/Docomo can argue regulatory leverage if Rakuten's coverage gaps persist.

7. MNO Net Adds Reality Check

Per Commerce Pick analysis:

MNO純増ペースは月間13〜15万回線程度

Real organic MNO net adds: only 130-150K/month. The headline "+1.71M YoY net subs" was inflated by BCP (business continuity plan B2B lines) and MVNE/MVNO consolidation. Organic individual subscriber growth is much slower than the headline implies.

8. Rakuten Bank (5838) — Rakuten Owns Only 49.26%, Not Majority

Per IRBank and Yahoo Finance Japan (5838):

• Rakuten Bank market cap (April 30, 2026): ¥1,100.2B
• Rakuten Group ownership: 49.26%
• Rakuten's stake value: ~¥542B (vs ¥800B I had assumed in English-source analysis)
• Bank stock has declined from ¥1,388B mcap in February to ¥1,100B in April — bank itself derated ~21% while parent holds a partial stake

9. Rakuten Card After Mizuho Deal — Only ~85% Owned

Per Rakuten IR (Japanese):

2024年12月に楽天カードの普通株式14.99%をみずほフィナンシャルグループに譲渡

Already disposed 14.99% of Rakuten Card to Mizuho Financial Group in December 2024. So FinTech segment no longer owns 100% of Card.

10. Old Perpetual Subordinated Bond Cliff — April 2026 First Calls

Per JTG Securities filings (Japanese):

• USD perpetual subordinated 5.125% coupon — first call April 22, 2026
• USD perpetual subordinated 6.250% coupon — first call April 22, 2026
• Both required refinancing in March-April 2026 via domestic perpetual subs at materially higher coupons in JPY terms (given JGB rose from 0.5% to 2.51%)
• These were issued when rates were near zero — the rolling cost increase is real

Updated WACC Inputs

The cost of capital is rising as cheap debt rolls off.

Updated SOTP — Using Japanese-Source Adjustments

Internet Services

• Revenue ¥1,370B, EBITDA ~¥130B
• Apply 6-8x EBITDA = ¥780-1,040B (no change from prior view)

FinTech (Revised Down)

The base case ¥1,500B is at the lower end of my prior range — the Bank stake is meaningfully lower at current market prices.

Mobile + Symphony (Revised Down)

Two valuation lenses:

A. Current EBITDA basis:

• ¥28.8B EBITDA × 8x growth multiple = ¥230B
• Symphony first-time profitability optionality: ¥80-120B
• Today's value: ¥310-350B

B. Forward-looking (3-year):

• 2028E EBITDA: ¥120-180B (assumes 12-13M MNO subs, ARPU stable around ¥2,500)
• × 6x KDDI multiple = ¥720-1,080B
• Discount back 3 years at 10% = ¥540-810B today

But: ¥200B 2026 capex is a real cash drain that limits forward EBITDA build. Use weighted midpoint: ¥400-550B

Other (AST SpaceMobile + Medical + Minorities)

• AST SpaceMobile stake (~3% of $13B mcap) = ¥60B
• Rakuten Medical (cancer treatment) = ¥50-100B
• Other minorities/options = ¥50-100B
• Total: ¥160-260B

Net Debt (Non-FinTech) — Revised Higher

• ¥1.0T+ given continued cash burn and refinancing at higher rates
• 2026 capex ¥200B+ implies further debt buildup before harvest

Holding Company Discount — Revised Higher to 20%

Reasons:

• 7 consecutive years of net losses
• No FY2026 guidance disclosed
• No dividend
• Continued capex burn cycle
• Multiple partial subsidiary disposals create complexity
• "Rural neglect" creates competitive/regulatory tail risk

Revised SOTP Table

Probability Weighting (30/45/25 — Revised More Bearish)

PWFV = ¥442 × 0.30 + ¥725 × 0.45 + ¥1,010 × 0.25 = ¥725

vs spot ¥760 = ~5% downside to PWFV (was +20% upside in English-source analysis)

Reconciliation with Japanese Analyst Consensus

Per Minkabu (April 29, 2026):

• 14 analysts covering: 2 strong buy, 5 buy, 6 hold, 1 sell
• Consensus target: ¥984 (+26.83% upside vs ¥760 spot)
• Yahoo Finance Japan shows similar ¥996 consensus

My PWFV of ¥725 is below the analyst consensus of ¥984 by ~26%. Why the gap?

The Japanese sell-side appears to be pricing my bull-case scenario as base case:

• Mobile reaching 12M+ subs by 2027
• Successful refinancing without spread blow-out
• FinTech compounding at 15%+
• Record EBITDA narrative dominates

I'm pricing in:

• 30% probability of bear case (refinancing pressure + Mobile capex absorption issues)
• More conservative Mobile EBITDA trajectory
• Higher hold-co discount given losses persistence
• Lower Bank stake value at current market

Both views are defensible. The disagreement is about whether Mobile's EBITDA breakeven will compound or stall under capex pressure.

Comparing the Two Valuation Approaches Side-by-Side

Investment Conclusion (Revised)

Rating: HOLD (revised down from LONG)

The Japanese-source analysis reveals enough material concerns that the risk-reward at ¥760 is now symmetric, not asymmetric:

• Bear: ¥442 (-42%)
• Base: ¥725 (-5%)
• Bull: ¥1,010 (+33%)

This is no longer the asymmetric long I had on English sources only. The structural concerns — 7-year loss streak, no guidance, no dividend, capex ramp, refinancing pressure, partial ownership in FinTech subsidiaries — meaningfully reduce the case for adding equity exposure here.

Sizing (Revised)

• Trim or pass at ¥760
• Only consider initiating below ¥650 (margin of safety vs ¥725 PWFV)
• Stop loss: ¥440 (bear case)
• Trim aggressively above ¥950

What the Q1 Print on May 14 Needs to Show to Support the Bull

• Mobile EBITDA Q1 ≥ ¥10B (sustained from ¥28.8B FY2025 run-rate)
• ARPU not declining (need ≥ ¥2,470)
• Net debt not building meaningfully (capex absorption visible)
• Some hint at FY2026 guidance even if directional
• Symphony continuing operating profit
• FinTech segment OI growth ≥ 20%

If any of these break, the bear scenario gains ground and ¥650 is the trade entry, not now.

The Honest Methodology Reflection

This exercise illustrates why language source diversity matters in equity research.

English coverage of Rakuten was largely:

• Translated press releases from Rakuten IR (which obviously emphasizes positives)
• TradingView/Investing.com superficial summaries
• Morningstar's outdated Fair Value (¥7,516 from May 2025 — pre-large reduction)

Japanese coverage included:

• Nikkei's critical reporting (rural neglect, 7-year losses framing)
• Minkabu's regulatory disclosure language (no guidance, no dividend explicit)
• IRBank's structural breakdown (49.26% Bank ownership clear)
• Commerce Pick's critical decomposition (true MNO net adds 130-150K/month)
• Business Network's industry-specific lens (¥200B capex jump in context)

The English coverage isn't wrong, but it's systematically more bullish because it's downstream of the company's English IR materials, which are designed to attract international investor capital. Japanese-language coverage benefits from local press scrutiny, regulatory disclosure norms in Japanese, and analyst commentary that's more critical (Japan has cultural deference but financial press is rigorous).

For Japanese names, always cross-check against Japanese-language sources before forming a view. The ~25% PWFV gap between my two analyses (¥908 vs ¥725) is the cost of skipping this step.

This applies broadly to non-English markets: Thai SET names need Kaohoon/Mitihoon/Thunhoon cross-checks, Hong Kong names need AAStocks and HKEx primary disclosures, Korean names need Korean-language coverage. The English-language summary is always downstream of company IR — and company IR is always promotional.

Sources Used (Japanese Only)

• 楽天グループ IR (corp.rakuten.co.jp) — FY2025/Q3/Q2 press releases in Japanese
• 楽天グループ FY2025 通期決算説明資料 (25Q4PPT_J.pdf)
• 日本経済新聞 (nikkei.com) — Mobile rural neglect article
• みんかぶ (minkabu.jp) — Stock screen, analyst consensus, earnings analysis
• Yahoo!ファイナンス (finance.yahoo.co.jp) — Stock data, growth/risk commentary
• 株予報Pro (kabuyoho.jp / kabuyoho.ifis.co.jp) — Target prices, ratings
• IRBank (irbank.net) — Rakuten Bank 49.26% ownership disclosure
• 株式会社アイフィスジャパン (kabuyoho.ifis.co.jp) — Analyst sentiment
• ICT総研 (ictr.co.jp) — Industry-specific decomposition
• Commerce Pick (commercepick.com) — Critical Q4 earnings analysis
• Business Network (businessnetwork.jp) — ¥200B capex contextualization
• ケータイ Watch (k-tai.watch.impress.co.jp) — Mobile-specific reaction
• JTG Securities (jtg-sec.co.jp) — USD perpetual subordinated bond first call dates

Research notes for personal use, not investment advice. Verify all figures independently before acting. This report supersedes my prior English-source-only Rakuten valuation published earlier today.