Β₯1.17 Trillion in annual profit. Japan's largest corporate buyback. A 9% Berkshire Hathaway stake. And yet the stock trades at 10 times earnings.
There is a peculiar dynamic playing out in Japan's stock market right now. Warren Buffett β the world's most imitated investor β publicly declared his Mitsubishi Corporation stake as his favorite Japanese holding. Berkshire owns approximately 9% of the company. The position has not been reduced since 2020. Yet Mitsubishi trades at 10x forward earnings, implying the market expects essentially zero growth from a business generating 19% return on equity with Β₯300 billion per year in share buybacks.
That disconnect is the thesis.
The Business in One Paragraph
Mitsubishi Corporation is not Mitsubishi Motors or Mitsubishi Electric. It is the parent trading house of one of Japan's most powerful corporate networks β a sogo shosha β operating seven divisions across 90 countries: natural gas (LNG from Brunei, Indonesia, Australia), minerals, chemicals, machinery, food, and automotive. The natural gas division alone generates roughly 30% of total profit. In 2024, the company completed the full privatization of Lawson β Japan's second-largest convenience store chain with 14,000 locations β adding retail consumer data and cash flows to the mix.
FY2025: Near-Record, Highest ROE of the Group
Net profit came in at Β₯1.17 trillion ($7.8 billion) β within 5% of the all-time record set in FY2023. More telling is the quality of those earnings: return on equity hit 19%, the highest among Japan's Big Five trading companies. The dividend was raised to Β₯200 per share. The buyback was authorized at Β₯300 billion β again, the largest of the group.
To understand why 19% ROE matters at 10x earnings: a business that compounds equity at 19% per year and trades at only book-value-plus should, by standard valuation logic, trade well above that multiple. The implied growth rate priced into the stock at 10x P/E is near zero. With Β₯300B in annual buybacks retiring ~4% of shares, EPS grows even without any underlying profit expansion.
Five Things Analysts Are Watching
The buyback math is mechanical. At 10x earnings, each repurchased share is removed from the float at below-intrinsic-value pricing. At Β₯300B per year on a Β₯7 trillion market cap, roughly 4% of shares are retired annually. This creates a compounding EPS effect independent of business performance.
Lawson gives Mitsubishi something none of its peers have: direct retail consumer relationships. 14,000 stores process millions of daily transactions. Mitsubishi is integrating this data with its upstream food import and supply chain operations β creating feedback loops between retail demand signals and procurement decisions. The strategic optionality here is understated.
Diversification as genuine risk management. Unlike Marubeni (grain exposure to US tariffs) or Sumitomo (nickel disruption), Mitsubishi's seven-division structure means no single commodity cycle can derail the consolidated numbers. This structural resilience justifies a valuation premium versus peers.
The Sakhalin-2 overhang. Mitsubishi maintains its stake in Russia's Sakhalin-2 LNG project. Western sanctions create ESG-related institutional selling pressure. The financial exposure is manageable (estimated 5-8% of natural gas segment profit) β but the reputational risk with ESG-mandated investors is real and ongoing.
Lawson integration costs front-load the pain. Analysts model Β₯40-50 billion in integration costs over FY2026-FY2028. This is a timing headwind, not a structural problem. The data flywheel benefits come later.
Valuation: Where Is the Mispricing?
At 10x forward P/E, Mitsubishi trades at a modest premium to Mitsui (9x) and Marubeni (8.8x). Against global industrials, consumer companies, or US conglomerates β all trading 18-25x β the discount is stark.
The earnings yield at 10x is approximately 10%. Against Japan's risk-free rate of 0.5%, that represents a 950 basis point spread. Even against global corporate bond yields of ~5%, the 500bp spread is historically generous for a company of Mitsubishi's quality and earnings consistency.
Price-to-book is 1.6x for a business generating 19% ROE. That implies close to zero growth expectations embedded in the price β a clear mismatch with the buyback-driven EPS trajectory and Lawson optionality.
The Call
BUY. Mitsubishi is the most defensible, highest-quality entry into the sogo shosha category. For investors who want the Berkshire-endorsed Japanese trading company thesis with maximum diversification, highest ROE in the group, and the largest buyback program, Mitsubishi is the anchor.
- Entry range: Β₯3,100β3,400
- 12-month target: Β₯4,000 (10.5x FY2026 estimated EPS)
- Risk: LOW-MEDIUM (most diversified of the five)
- Watch level: Below Β₯2,800 signals sustained earnings deterioration
Full analysis with charts and peer comparison at averin.com
This is not financial advice. Positions may change. Do your own due diligence.
Tags: Japan stocks, Sogo Shosha, Mitsubishi, Berkshire Hathaway, Value Investing, Japanese equities, TSE 8058
