The noise fades, but the pattern remembers
Goldman Sachs was bullish on three Asian currencies—Korean won, Taiwanese dollar, Malaysian ringgit. All three are down against the U.S. dollar in 2026. Taiwan’s won the loser’s podium, falling 3.05%. Ringgit dropped 2.8%. Won shed 1.9%. The investment bank’s “AI-driven trade surplus” narrative? Evaporated under the weight of a dollar that refuses to bend.
This is not a story of a broken model. It’s a story of a hidden variable—the dollar’s gravity—that Goldman’s spreadsheet forgot to code.
We didn’t just watch the chart, we lived it
I’ve been tracking macro undercurrents since the 2017 Telegram days. Back then, I sprinted through ICO mania, spotting vulnerabilities in minting functions minutes before the crowd. Now, the battlefield is different: central bank policy, capital flows, and the quiet tyranny of the Fed.
Goldman’s thesis was simple: chip exports boom, current account surpluses swell, currencies strengthen. Korea’s surplus was set to double to ~$300B (13.9% of GDP). Taiwan’s hit 25% of GDP. Malaysia rode on AI-linked FDI. But the market didn’t obey. Why? Because the dollar rose nearly 3% year-to-date, crushing every Asian contender—except one.
Renminbi appreciated 3.32% against the dollar in 2026, the only Asian currency to do so. Goldman kept its USD/CNY 6.50 forecast. But this isn’t magic. It’s policy intervention—China’s central bank using reserves, offshore bills, and capital controls to hold the line.
Core: The AI/Energy Split That Dollar Obscured
Scratch beneath the surface. The delta between the weakest “AI currency” (Taiwan dollar, -3.05%) and the strongest “energy currency” (Philippine peso, -4.48%) is only 1.43 percentage points. That’s the alpha Goldman was talking about. The problem? It’s buried under the dollar’s weight.
Goldman’s framework splits Asia into two worlds: - AI exporters: Korea, Taiwan, Malaysia — strong current accounts, semiconductor-driven growth. - Energy importers: Thailand, Indonesia, Philippines — oil-sensitive, weak current accounts, limited policy space.
In 2026, the spread between them exists, but it’s narrow. The dollar’s strength is the common denominator that drags all boats lower. Goldman’s call was a ‘relative value’ trade—long AI currencies, short energy currencies—not an absolute one. The market misread it as a bullish absolute call. Classic mismatch between sell-side storytelling and buy-side execution.
Where Goldman’s model cracked
Three blind spots: 1. Dollar dominance: The model assumed U.S. rate cuts and a weaker dollar. Instead, Fed held rates high due to sticky inflation. Dollar index climbed. 2. AI investment cycle timing: Goldman priced in immediate export acceleration. But 2026 data shows chip orders flat-lining as hyperscalers (Microsoft, Google, Meta) pace their capex. The cycle hasn’t broken, but it’s stretching. 3. Capital flow dynamics: The bank noted that “waning foreign equity outflows” paved the way for won strength. But those outflows turned to outflows again as global risk appetite faded into dollar assets.
Contrarian: The real trade isn’t currencies—it’s volatility
Trust the code, verify the art, ignore the hype
Here’s what the headline misses: Goldman’s framework is not wrong, it’s incomplete. The AI/energy split is real and will amplify when the dollar cycle turns. But until then, the smart money isn’t buying Asian currencies outright. It’s selling options on the spread.
Imagine a basket: long KRW/THB (won vs. baht), short USD/Asia. This pair captures the AI-energy divergence while hedging dollar risk. That’s where the alpha hides—not in absolute FX bets, but in structured relative value trades.
Meanwhile, the renminbi remains the outlier. Its appreciation is a policy artifact, not a market signal. Don’t extrapolate it to other Asian currencies. China’s capital controls create a “controlled float” that no other Asian central bank can replicate.
Takeaway: Watch the triggers, not the headlines
The next move depends on three signals: - U.S. tech capex guidance (next earnings season): If Microsoft or Alphabet slash AI spending, the AI currency alpha evaporates. - Oil prices: A Brent spike above $120 torpedoes Thai baht and Indonesian rupiah, widening the divergence. - Fed pivot: A rate cut could reverse the dollar’s strength, freeing AI currencies to rally.
Goldman’s 2026 bet lost to the dollar’s silent scream. But the pattern remembers. When the Fed blinks, the won, ringgit, and Taiwan dollar will roar. Until then, trade the spread. Not the story.
Signal: The divergence is alive, just compressed. Noise: Goldman was “wrong.” In markets, being early is the same as being wrong—until it’s not.
