Hook
The year is 2026. Goldman Sachs, the oracle of Wall Street, declares its love for three Asian currencies: the Korean won, the Taiwanese dollar, and the Malaysian ringgit. The thesis is clean, almost poetic—artificial intelligence investment will supercharge exports, fatten current account surpluses, and push these currencies higher. The market responds with a collective shrug. All three are down against the US dollar. The won? Down. The ringgit? Down. The Taiwanese dollar? Down the most—3.05% as the greenback rose nearly 3%.
Where liquidity hides, narrative finds its voice. And right now, the only voice that matters is the one coming from the Federal Reserve.
Context
Goldman’s analysis was not without logic. They framed Asia’s macro landscape through a bifurcation lens: countries riding the AI wave (South Korea, Taiwan, Malaysia) versus those hobbled by energy imports (Thailand, Indonesia, the Philippines). The former, they argued, would see their currencies strengthen as semiconductor exports surged and capital inflows followed. Korea’s current account surplus was forecast to nearly double to $300 billion, while Taiwan’s surplus-to-GDP ratio hit a staggering 25%. The ringgit drew support from steady foreign direct investment, supposedly tied to the China-plus-one manufacturing shift.
But the market is a cruel teacher. The dollar index climbed, and every Asian currency except the Chinese yuan took a hit. Even the “AI champions” could not escape the gravitational pull of Fed policy. The only outlier—the yuan—rose 3.32%, a move that smells more of heavy intervention than free-market strength.
This is not just a story about foreign exchange. It is a parable for the crypto markets we live in. Chasing ghosts in the algorithmic machine of fundamentals, we often forget that the macro liquidity cycle is the only god that matters.
Core
Let me dismantle the Goldman thesis piece by piece, because the failure of this model holds profound implications for how we analyze crypto assets.
First, the “trade surplus equals currency strength” equation is a relic of a simpler world. In the pre-2008 golden age, a current account surplus would indeed support a currency, as foreign buyers needed that country’s money to pay for its goods. But today, capital flows dwarf trade flows. The global FX market trades over $7.5 trillion per day; trade makes up a tiny fraction. What matters is global portfolio allocation, and when the dollar offers real yields that are positive and attractive, capital flows out of emerging markets regardless of their trade books. The AI export story is like a DeFi protocol with a brilliant tokenomics model—when the overall liquidity pool is shrinking, even the best yield-bearing asset gets drained.
Second, the AI investment cycle is itself hostage to dollar liquidity. The tech giants (Meta, Google, Microsoft) that burn cash on GPU clusters finance those capex through debt and equity markets. If the Fed keeps rates high, the cost of capital rises, and CFOs start asking whether $50 billion in AI infrastructure is worth it. The first sign of a slowdown in AI capital expenditure will send the Korean won and Taiwanese dollar into a tailspin far worse than any energy-importing currency. The concentration of these economies in a single sector (semiconductors) is a double-edged sword. When the sword falls, the same leverage that boosted the surplus amplifies the crash.
Based on my experience during the DeFi Summer of 2020, I saw this exact pattern play out in crypto. Yield farmers piled into protocols like SushiSwap and Yearn, chasing APYs that seemed justified by “protocol revenue.” But when the overall liquidity of the crypto market tightened—triggered by a sudden plunge in Bitcoin—those yields evaporated, and the native tokens crashed faster than the market average. The real driver wasn’t the protocol’s fundamentals; it was the macro liquidity tide. The same principle applies here: a current account surplus is just a protocol’s revenue. The dollar liquidity cycle is the total market capital.
Third, the capital flow dynamics are more nuanced than Goldman admitted. They argued that reduced foreign equity outflows would allow the current account surplus to shine. But they missed the fact that during risk-off episodes, global investors repatriate capital to the United States not just from equities, but from bonds, real estate, and currency reserves. The entire balance of payments shifts. Even if equity outflows slow, fixed-income outflows can accelerate. The net effect is that the surplus never gets a chance to bid up the currency because the cap exodus is larger. This is analogous to a stablecoin peg—you can have a huge reserve backing, but if everyone redeems at once, the peg breaks.
The Chinese yuan stands as the only counterexample, and it proves the rule. The yuan appreciated against the dollar in 2026 because the People’s Bank of China intervened directly: setting daily fixings below market rates, draining offshore yuan liquidity with bills, and likely using reserves to buy yuan. This is not a market-driven strength; it is a managed one. For crypto investors, think of it as a central bank issuing a fully-backed stablecoin and actively maintaining the peg. It works until it doesn’t—reserves are finite. But in the short term, it creates the illusion of decoupling.
The illusion of control in a fluid world is perhaps the most dangerous mindset for any investor. Goldman believed they could predict currency movements by modeling trade flows. Crypto believers often think they can predict token prices by modeling on-chain activity. Both forget that the ocean is bigger than any boat. When the dollar tide goes out, all boats—regardless of their engine power—fall with the waterline.
Let me zoom further into the “relative value” nuance. In a typical bear market, the question is not which currency goes up; it’s which goes down the least. Goldman’s recommendations were actually a call for a relative-value trade: long the AI currencies, short the energy importer currencies. And indeed, the worst AI currency (TWD at -3.05%) outperformed the best energy importer (PHP at -4.48%). The spread was about 1.4 percentage points. That is positive alpha, but it is drowned out by the absolute loss. In crypto terms, this is like saying “Bitcoin fell 20% but my altcoin only fell 15%.” A relative win is still an absolute pain. The only way to profit is to hedge the dollar exposure—for example, by shorting the dollar index simultaneously. Most retail crypto traders do not hedge. They go long a narrative and get wrecked by macro.
Contrarian
The common narrative among emerging-market bulls is that Asia is decoupling from the US economic cycle, led by China’s rebalancing and the AI revolution. This is the crypto equivalent of “Bitcoin is digital gold that decouples from equities.” Both narratives have been tested and found wanting. In 2022, Bitcoin correlated with the Nasdaq to the tune of +0.8. In 2026, Asian currencies correlated with the DXY to a similar degree. The decoupling myth persists because it is emotionally satisfying—no one wants to believe they are just passengers on a Fed-driven ship. But the data is clear: systemic factors dominate.
What is the contrarian take then? That the Goldman thesis was actually right about the relative ranking, but the market was not pricing the absolute level correctly. If the dollar weakens—say the Fed cuts rates or a recession triggers risk-on flows—then the AI currencies will soar not just relatively, but absolutely. The current pain is a buying opportunity, but only for those who can stomach macro volatility. Similarly, in crypto, the best time to accumulate high-beta plays is when the dollar cycle is peaking and about to reverse. Reading the silence between the blockchain blocks means ignoring the noise of daily price moves and focusing on the macro liquidity signals: real yields, dollar index, Fed balance sheet.
Another contrarian angle: the energy-importing currencies (Thai baht, Indonesian rupiah, Philippine peso) may actually be the better risk/reward if oil prices collapse. Goldman did not model the scenario of a global recession that kills oil demand. If Brent crude drops below $70, those countries’ terms of trade improve dramatically. Their central banks can cut rates, and their currencies can rally. This is like shorting a DeFi token that benefits from high gas fees when you anticipate a L2 scaling solution that will crush fees. The macro switch can flip faster than anyone expects.
Takeaway
So what does all this mean for the crypto investor sitting in Bangkok in mid-2026?
Stop looking at the trading volumes of your favorite altcoin and start looking at the US dollar index. The single most powerful leading indicator for crypto liquidity is the DXY—when the dollar weakens, liquidity flows into risk assets, including crypto. When it strengthens, even the best project tokens bleed. The AI currency story is just a mirror of the crypto cycle: a strong fundamental narrative that gets crushed by macro gravity.
Volatility is just information wearing a mask. The mask here is the AI export boom. The information is that the Fed still controls the global liquidity valve. As long as that valve is tight, no amount of semiconductor surplus will save the won, the ringgit, or your crypto portfolio. The cycle will turn when the dollar cycle turns—not before.