Strait Risk: Why Japan & AI Could Drive a Global SelloffUSD/JPYOANDA:USDJPYadamnassibHow the yen became the world's funding currency After Japan's bubble collapsed in 1990, the BoJ cut rates to zero and kept them there. In September 2016 they went further with Yield Curve Control (YCC), explicitly capping the 10-year Japanese Government Bonds at 0% while holding TONA1! overnight rates at -0.1%, buying unlimited bonds to enforce it. At its peak the BoJ owned 53% of Japan's entire government bond market. They still held 46% at end-2024 even after tapering reserves. That policy created a simple trade: borrow in yen at near zero, invest in US Treasuries, equities, or anything with a yield. With the Fed at 4.25-5.5% and Japan at zero, the rate differential alone was worth ~500bps annually before any price appreciation. This became what is now known as the Carry Trade. The yen also depreciated 35% between 2021 and 2024, which made every dollar-denominated position even more profitable in yen terms. The trade fed itself. What is actually at stake Japan is the single largest foreign holder of US Treasuries at $1.14 trillion, about 15% of all foreign-held US government debt. That position did not get built on fundamentals. When your country's domestic government bonds yield 0%, public pension funds, state-owned banks, and insurance giants go shopping abroad. The yen was your vehicle; and the US Treasuries were the destination. When the carry trade unwinds, those Treasuries get sold, and that is the direct transmission from a Japanese rate shock to global markets. On July 31, 2024, the BoJ hiked rates to 0.25%. In 4 trading days the Nikkei fell over 18%. The VIX hit 65. USDJPY dropped 11%. JPMorgan estimated 65-75% of global speculative carry positioning closed out within weeks. The hardest hit assets were not EM or commodities, they were US momentum and AI stocks, exactly what the carry trade had been funding. That was a carefully planned, and telegraphed move. Yet, the markets still nearly broke. The Hormuz scenario Japan imports 90% of its oil from the Middle East through the Strait of Hormuz, a corridor just 33km wide that carries roughly one-fifth of global oil supply daily. A closure forces this chain: Oil spikes -> Japan inflation surges -> BoJ forced to hike faster than planned -> JGB yields rise -> yen borrowing costs jump -> carry trade math breaks -> Rapid and instantaneous repatriation The difference from 2024 is there is no guidance, no preparation, no managed exit. It is an external shock. Repatriation is abrupt, margin calls hit simultaneously, and there is no orderly unwind. The second vector: helium Qatar produces 30-36% of global helium, processed at Ras Laffan and shipped through the same Strait. The containers that carry it can only keep helium in liquid form for 35-48 days before it vents and is gone permanently. Helium has no substitute in semiconductor manufacturing. It is critical for wafer cooling and equipment purging. Taiwan sourced 69% of its helium from Gulf states in 2024. South Korea sourced 55% in 2025. These two countries make the majority of the world's advanced chips. A prolonged closure does not just slow chip production. It directly constrains the physical buildout of AI infrastructure. The market treats this as a footnote. It should not. Bottom line Vector 1: Oil shock BoJ forced hikes Carry unwind Global deleveraging Vector 2: Hormuz closure Qatar helium offline Chip supply constrained AI buildout stalls tech multiple compression The 2024 stress test was controlled and still produced an 18% crash and a VIX of 65. A forced version of this, hitting both channels at once, is a different order of magnitude. JPY bullish. Equities bearish. Vol spikes. Tech takes the second hit. Oil ignites. FX and rates transmit. Tech follows. USDJPY hit 160.26 on March 28 and is pressing that level again now. A break above it puts the BoJ in an impossible position, and the pieces this post describes are already in motion.