The bank says AI-linked exporters such as South Korea, Taiwan and Malaysia should outperform energy-importing peers, while the yuan remains a separate constructive call and Singapore stays neutral.
Goldman Sachs says an artificial intelligence investment boom is reshaping Asian foreign-exchange markets, favoring currencies tied to semiconductor exports over those of economies more exposed to imported energy costs. The bank remains most constructive on the South Korean won, the Taiwan dollar and the Malaysian ringgit, while also keeping a separate positive view on the Chinese yuan. Goldman forecasts South Korea's current account surplus will nearly double to about $300 billion this year, or 13.9% of GDP, and said reduced foreign equity outflows are allowing that external strength to support the won more clearly. It also expects the Taiwan dollar to benefit from surging semiconductor exports and a widening trade surplus, with Taiwan's current account surplus seen at 25% of GDP even if interest rates remain unchanged. Malaysia is the third currency tied to Goldman's AI theme, supported by resilient AI-led growth, strong exports and sustained foreign direct investment. By contrast, the bank sees weaker prospects for the Thai baht, Indonesian rupiah and Philippine peso because of exposure to oil prices, lower real rates, or policy and governance concerns. Market performance this year still shows broad dollar strength, with the dollar index up nearly 3% in 2026. The won, Taiwan dollar and ringgit have all weakened against the greenback, but by less than several energy-importing peers. China is the exception, with the yuan up 3.32% this year; Goldman kept its 12-month USD/CNY forecast at 6.50, citing undervaluation and Beijing's push to internationalize the currency. The bank is also constructive on the Indian rupee for reasons not tied to AI, while remaining neutral on the Singapore dollar after policy was left unchanged.