Hook
Greg Abel, the man now holding Warren Buffett's seat, told the world Berkshire Hathaway remains committed to Japan's five trading houses. The market heard eternal loyalty. I heard something else.
I heard a carry trade with a borrowed maturity date.
Context
Let's map the mechanics, because the headlines skip the plumbing. Berkshire has issued over ¥1.5 trillion in yen-denominated bonds since 2020. The proceeds bought Mitsubishi, Mitsui, Itochu, Sumitomo, and Marubeni. The play is elegant: borrow at near-zero rates in a currency the central bank was actively suppressing, then park that capital in high-dividend Japanese equities trading at a persistent valuation discount.
For years, this was the cleanest trade in global markets. Negative rates plus quantitative easing plus yield curve control created an artificial environment where the cost of capital approached zero. The Bank of Japan essentially subsidized Berkshire's position.
The trap isn't that the strategy was wrong. The trap is that everyone calls it "long-term" while ignoring that its foundation was a policy artifact.
Core
Here's what Abel's reassurance doesn't say: the Bank of Japan ended negative rates in March 2024. The policy rate now sits at 0.5% to 0.75%. Each additional hike compresses Berkshire's spread between yen borrowing costs and dollar-denominated returns.
My 2022 Terra/Luna analysis taught me to watch how macro liquidity drains cascade through micro structures. This is the same disease, different patient. When the Fed tightened, algorithmic stablecoins broke because their yield assumptions inverted. When the BOJ tightens, carry trades break because their cost assumptions invert. The timeline is slower, but the physics are identical.
Run the numbers. Berkshire's yen bonds issued in 2020-2021 carried coupons around 0.2% to 0.5%. Japanese trading houses yield roughly 3-5% in dividends. The spread looked generous. But if the BOJ pushes toward 1% and new issuance requires 1.2% coupons, Berkshire's new capital deployed into Japanese equities loses its margin of safety. The existing book survives. The expansion thesis weakens.
There's a second layer. The five sogo shosha aren't really Japanese companies. They're global commodity and supply chain conglomerates with Japanese addresses. Mitsubishi and Mitsui generate substantial revenue from energy, metals, and food trading across Southeast Asia and Australia. Itochu's non-resource segments dominate its earnings. Marubeni's power and infrastructure assets span emerging markets.
Berkshire isn't betting on Japan's 0.5% potential growth. It's betting on global trade flows routed through Japanese intermediaries.
This is where my 2026 AI-Crypto compute market work connects. When I analyzed whether decentralized GPU networks could disrupt centralized cloud providers, the key variable was the same: who controls the routing layer? In global trade, the sogo shosha control routing. They coordinate suppliers, logistics, financing, and risk absorption across borders. That's a structural position, not a cyclical one.
But structural positions still face cyclical pain. Global trade contracted in 2020, spiked in 2021-2022, and has been grinding sideways since. The shosha's earnings move with the CRB commodity index, not with Japanese GDP. Abel's "reaffirmation" implicitly forecasts stable-to-improving commodity prices. If he's wrong about that, the "long-term commitment" becomes a loss-mitigation exercise.
The market impact is straightforward. Berkshire holds roughly 8-10% of each trading house. Abel's statement eliminates the tail risk of sudden liquidation. That's meaningful for price support. But it doesn't create new demand. This is confirmation, not innovation. Institutions that already positioned for Berkshire's exit need to reassess. Institutions that priced in continued accumulation are already long.
Contrarian
The most dangerous word in Abel's vocabulary is "long-term."
Berkshire's history contradicts the permanence implied. The airline positions built in 2016-2017 with great fanfare were liquidated in April 2020, at the bottom. Buffett called it a mistake, then exited anyway. The "forever holding period" applies to wholly-owned businesses like See's Candies. Public market positions enjoy no such immunity. When the thesis breaks, the position goes.
The break risk isn't BOJ policy alone. It's the combination of BOJ tightening, yen appreciation, and global trade contraction hitting simultaneously. If the yen strengthens toward 130 or below, Berkshire's dollar-denominated returns on its yen assets get crushed. The carry trade swings from positive to negative. Abel's "reaffirmation" today is a statement of current conditions, not a binding contract for future behavior.
There's also a crowding problem. Berkshire's presence attracted imitators. Foreign ownership of Japanese equities rose steadily since 2023, partly tracking Buffett's disclosed positions. If Abel's reassurance triggers still more foreign buying, the valuation gap that made the trade attractive simply closes. Berkshire ends up paying more for the same exposure. That's not conviction. That's momentum chasing, dressed in a suit.
Takeaway
What's the signal beneath the signal? Abel's statement tells us more about Berkshire's internal capital allocation logic than about Japan. It says the firm still believes in value extraction from policy-normalizing economies with cheaper currencies. It says they believe global trade persists despite fragmentation rhetoric. It says they believe the shosha's transformation toward energy transition investments and supply chain reorganization will compound.
The trap isn't Japan. The trap is confusing stability with permanence.
Chaos is just data that hasn't been timestamped yet. The yen carry trade will end when the BOJ's credibility problem becomes more pressing than its growth problem, or when global trade produces a shock that the shosha's diversification can't absorb. Abel's reassurance doesn't date that event. It just extends the position.
Watch the 13F filings. Watch the BOJ's projections. Watch commodity prices. When those three stop aligning with Abel's thesis, "long-term" will start looking a lot like "until the next quarterly review."
The carry trade always forgets it was borrowed. The lender never does.