HSBC’s chief economist warns of parallels to 1997 Asian crisis

The last time rising US borrowing costs, a plummeting Japanese yen, and infectious tech optimism converged in Asia, the result was a financial crisis that toppled governments and wiped out decades of economic progress. HSBC’s chief Asia economist Frederick Neumann thinks the ingredients are lining up again.

In a note dated August 31, Neumann drew pointed comparisons between the current macro environment and the conditions that preceded the 1997 Asian financial crisis. Three factors, in particular, caught his attention: sharply rising US Treasury yields, persistent yen weakness, and investor exuberance around a dominant technology narrative. In the 1990s, that narrative was the internet boom. Today, it’s artificial intelligence.

The parallels are hard to ignore

Start with yields. In the early-to-mid 1990s, US Treasury rates surged from roughly 5% in October 1993 to nearly 8% by November 1994, sending shockwaves through emerging markets that depended on dollar-denominated capital. The current trajectory looks uncomfortably familiar. The 10-year US Treasury yield has climbed to approximately 4.79%, an increase of about 80 basis points since February 2026 alone. The broader arc is even more dramatic: yields have risen from around 0.5% in August 2020 to their current level, a move that has fundamentally repriced the cost of dollar funding worldwide.

Then there’s the yen. Before the 1997 crisis, the Japanese currency depreciated roughly 55%, sliding from 80 to 130 against the dollar. That cheapened Japanese exports and put competitive pressure on the rest of Asia, contributing to the imbalances that eventually cracked.

And finally, the tech enthusiasm. In the mid-1990s, the internet revolution drove capital into Asian manufacturing hubs that supplied the hardware underpinning the boom. Today, the AI investment cycle is doing the same thing, channeling enormous demand into semiconductor and electronics exporters like South Korea, Japan, and Singapore.

Why Neumann says this time is different (and actually means it)

The core distinction: in the 1990s, most Asian economies were net capital importers. They relied on foreign savings to fund domestic investment, ran current account deficits, and maintained fragile financial systems with limited regulatory oversight. When dollar funding costs spiked, the capital flows reversed, currencies collapsed, and the dominoes fell from Thailand to Indonesia to South Korea.

Today’s Asian economies are, by and large, net capital exporters. They run current account surpluses, hold substantial foreign exchange reserves, and operate under regulatory frameworks that were rebuilt specifically in response to the 1997 meltdown. The structural vulnerability that made the original crisis so contagious, an overreliance on short-term foreign capital, has been largely engineered out of the system.

The new vulnerability: an AI demand shock

Rather than financial-system fragility, Neumann identifies the real threat as a potential downturn in US demand for AI hardware. South Korea, Japan, and Singapore have built significant export exposure to the AI supply chain, particularly in semiconductors and advanced electronics. If US appetite for AI infrastructure cools, the transmission mechanism into Asian economies would be direct and painful.

The US Treasury itself appears to be watching the yield situation carefully. Plans are in motion to double liquidity-support buybacks for longer-dated debt, scaling operations from $2 billion to at least $4 billion per operation starting September 9.

For investors with exposure to Asian markets, Neumann’s framework suggests the new playbook requires monitoring US tech capex trends, semiconductor order books, and the durability of AI-related demand. A cooling in that demand wouldn’t trigger a 1997-style meltdown, but it could deliver a meaningful earnings shock to Asia’s most advanced economies at precisely the moment when rising borrowing costs are already squeezing margins.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.