Of the five major Japanese trading companies that Warren Buffett's Berkshire Hathaway holds, Marubeni Corporation attracts the least attention outside specialist circles. That relative obscurity is also why it trades at the lowest valuation: sub-9x forward earnings with a 3.5% dividend yield. At ¥2,350 per share in May 2026, it is the cheapest of the Big Five — and it is cheap for reasons that are both real and potentially misunderstood by markets focused on short-term tariff headlines rather than the underlying business structure.
What Marubeni Does — Two Very Different Businesses Under One Roof
Marubeni has an unusual structure for a sogo shosha. Its two dominant business divisions are agribusiness and power/infrastructure — and they carry fundamentally different risk profiles.
The agribusiness division is anchored by the Gavilon grain business acquired in 2012 for $3.6 billion. Gavilon is one of the largest grain originators in North America — operating elevator storage, origination, and export facilities across the US Midwest and Gulf, moving corn, soybeans, and wheat primarily to Japanese food manufacturers and Asian buyers. This business is directly in the crossfire of US tariff policy. The 24% US tariff on Japanese goods — currently paused through July 2026 — creates real uncertainty for grain export economics.
The power and infrastructure division tells a completely different story. Marubeni operates as one of Asia's largest independent power producers (IPPs), with operational projects spanning 35 countries and combined generating capacity exceeding 10 gigawatts. These include solar, wind, LNG-fired thermal, and conventional plants across Saudi Arabia, the Philippines, Taiwan, Australia, and Southeast Asia. Power project revenues derive from 20-30 year concession agreements with government or near-government counterparties. They are dollar-denominated. They are structurally insulated from US-Japan trade policy. And the pipeline of new projects for FY2026-FY2031 exceeds 15 gigawatts.
FY2025 Results: Solid Foundation Masked by One Risk Factor
Marubeni posted ¥550 billion ($3.7B) in net profit for FY2025. The power segment grew 12% year-over-year as new IPP projects came online across Asia and the Middle East. The agribusiness segment saw margin compression as grain spreads tightened and anticipation of US trade friction began weighing on forward contract volumes. Dividend per share increased to ¥95. Return on equity: 14%.
The ROE of 14% is respectable — below Itochu (18%) and Mitsubishi (19%), but above the historical sogo shosha average of 8-10%. It reflects the blend of low-volatility IPP revenues with higher-volatility commodity trading.
The Bull Case: Two Structural Arguments
First, power infrastructure growth is completely independent of US tariff outcomes. The energy transition across Asia, the Middle East, and Africa is creating sustained demand for IPPs with project finance expertise, offtake relationships, and government partnerships. Marubeni's 35-country footprint and 15GW+ new pipeline are driven by electrification of developing economies — a multi-decade structural trend that no trade negotiation can derail. As new projects come online through FY2026-FY2028, this division's earnings contribution will grow regardless of grain market dynamics.
Second, tariff risk appears priced in — and may be overstated. The 24% US tariff is paused through July 2026. Japan-US trade talks are ongoing. Most analysts in Tokyo and Washington estimate 60-70% probability of a deal before the pause expires — because both governments face strong domestic economic incentives to reach an agreement. If a deal is struck, the tariff discount embedded in Marubeni's valuation versus peers could partially or fully reverse, producing 15-20% outperformance.
The Bear Case: Binary Outcome Risk
If US-Japan trade negotiations fail, consequences for Marubeni's agribusiness division are direct and material. Gavilon's grain export volumes would face meaningful reduction under sustained 24% tariffs. Analysts model a 15-20% reduction in agribusiness profit in a no-deal scenario — which would reduce total net profit by approximately 8-10%, pushing earnings toward ¥490-500 billion.
Execution risk in frontier market power projects. Marubeni's IPP expansion into developing markets introduces country risk, construction execution risk, and offtaker credit quality risk that the developed-market portfolio does not carry. Projects in countries with currency instability or weaker rule-of-law create tail risks that are difficult to model precisely.
Valuation: Sub-9x Creates Asymmetric Risk/Reward
At 8.8x forward earnings, Marubeni is the cheapest of the five on every standard metric: P/E, price-to-book (1.4x), and EV/EBITDA. The discount versus Mitsui (9x) and Mitsubishi (10x) is almost entirely explained by the grain tariff overhang — which is a binary risk that either resolves (trade deal) or escalates (permanent tariffs).
The key observation our analysts note: at sub-9x earnings with a growing trade-policy-independent power infrastructure business, the downside scenario (no deal) appears largely priced. The upside scenario (deal) appears not priced. That asymmetry is what creates the investment opportunity — and why Marubeni is the only speculative entry of the five.
Verdict: Speculative Buy — With a Defined Thesis
SPECULATIVE BUY — for investors with explicit conviction on the US-Japan trade deal closing before July 2026.
The binary nature of the tariff risk requires investors to have a view. Those who believe the deal closes have a clear catalyst for 15-20% outperformance versus peers. Those who believe tariffs become permanent should avoid. This is not a stock for passive holders — it requires monitoring trade negotiation headlines.
Entry range: ¥2,200–2,400 Target if trade deal closes: ¥2,800–3,000 (12 months) Risk if no deal and permanent tariffs: ¥1,800–2,000 (exit position) Risk level: HIGH (binary trade policy outcome) Income return while waiting: 3.5% annual dividend