Asia’s Currency Markets Are Better Connected: Are They Better Protected?

Asia's currency markets no longer concentrate financial stress in the same way they did during the 1997 crisis. Risk is distributed more widely, but disturbances also travel through a denser network of relationships. For central banks, this creates a difficult problem: a region can become less dependent on one troubled currency while remaining exposed to shocks spreading across several markets.

This changing pattern is examined in A quarter century after the 1997 Asian crisis: Looking back, where are we now?, by Samet Gunay, Péter Juhász and Helena Naffa of Corvinus University of Budapest's Institute of Finance. Published in Asia and the Global Economy, the study combines three analytical approaches to examine how currencies contribute to regional risk, how concentrated those contributions are and how stress travels between markets.

The findings suggest that Asia's vulnerability has changed form. The Japanese yen remains a prominent transmitter of regional shocks, while several other currencies repeatedly absorb them. For policymakers and development institutions, the implication is that stronger domestic safeguards must be assessed alongside external connections capable of transmitting pressure into otherwise more resilient economies.

Risk Has Dispersed, but the Channels for Contagion Have Expanded

The researchers examine daily exchange-rate data for ten East and Southeast Asian currencies, obtained from LSEG Eikon and spanning January 1994 to October 2025. Their principal comparison places the 1994–2000 crisis-era window alongside October 2018–September 2025. The approach distinguishes each currency's contribution to modelled losses under extreme conditions from the statistical relationships connecting currency markets.

During the earlier period, the study records abrupt concentrations of risk. The Thai baht accounted for 72.14% of the model's systemic-risk measure in the second quarter of 1997, during the buildup to the crisis. Indonesia's rupiah subsequently became a major contributor. These figures represent shares of modelled regional risk, rather than exchange-rate declines or losses in national income.

The recent period presents a broader distribution across several currencies. The study's concentration index consequently remains substantially lower than during the earlier crisis window. Such diversification reduces the dominance of individual currencies within the model, but it cannot establish that the total amount of risk has fallen. A more even distribution can still leave many markets exposed simultaneously.

The network results reinforce that distinction. Under the shorter-horizon specification, significant directional links involving volatility increased from nine in the earlier window to 35 in the recent one. Links involving extreme movements rose from 17 to 39. These are predictive statistical connections, rather than monetary measures of damage, but they indicate that regional interdependence extends across more channels.

Put together, the findings challenge assessments that equate a lower concentration of stress with comprehensive financial security. The practical question becomes whether authorities can identify and manage disturbances moving through several currencies at once. Monitoring only the market with the largest individual risk contribution could leave important transmission routes outside the field of view.

The Yen's Renewed Weight Changes the Surveillance Agenda

The yen provides the clearest example of shifting systemic importance. Its average contribution to the modelled risk measure increased from 7.52% in October 2018–December 2021 to 14.74% in January 2022–September 2025, a rise of approximately 96%. Separately, the network analysis identifies the Japanese currency as a dominant transmitter of regional spillovers across both comparison periods.

The distinction between these findings is important. A currency can account for a moderate share of measured regional risk while still influencing movements elsewhere. Risk contribution describes its place within an aggregate portfolio; transmission describes its connections to other markets. Looking at either measure alone can therefore produce an incomplete picture of where policymakers should direct attention.

The authors interpret monetary-policy divergence, yen depreciation pressures and changing capital flows as plausible explanations for Japan's renewed contribution. Their analysis, however, does not isolate the causal effect of particular policy decisions. The evidence supports closer examination of yen-related developments as part of regional surveillance, rather than a claim that Japanese policy alone determines Asia's financial stability.

Exposure also depends on which currencies receive shocks. The rupiah and South Korean won emerge as persistent receivers of extreme-risk spillovers in both periods, while the Chinese renminbi and Malaysian ringgit become more prominent receivers recently. The resulting policy distinction is consequential: economies absorbing external pressure may require different safeguards from those primarily transmitting it.

National Financial Defences Need a Regional Response

For governments and supervisors, the study supports combining domestic financial assessments with regular monitoring of cross-currency relationships. Reserve adequacy, external debt and banking exposures remain essential, while network indicators could help identify changing sources of pressure. Scenario exercises would become more informative if they considered several possible origins of stress and currencies' changing roles over time.

The authors propose incorporating their measures into ASEAN+3 surveillance and the Chiang Mai Initiative Multilateralisation framework. The institutional connection is practical: CMIM provides regional currency-swap arrangements addressing balance-of-payments and short-term liquidity difficulties, with AMRO supporting operational readiness. Research-based signals could inform consultation and preparedness, although they are not sufficiently validated to determine lending or intervention automatically.

For developing economies, the broader concern extends beyond exchange-rate movements themselves. External currency pressure can complicate import financing, foreign-currency debt servicing and business investment. These are analytical implications rather than outcomes measured by the study. Establishing how financial spillovers affect public budgets, employment and household purchasing power would make the evidence more directly useful for development policy.

Businesses likewise have reason to examine whether apparently diversified currency exposures remain vulnerable to shared shocks. Development agencies could support supervisory capacity, better exposure reporting and access to appropriate hedging instruments. However, stronger liquidity requirements and tighter foreign-currency exposure limits involve trade-offs: protection may come with higher financing costs, particularly for smaller firms with limited alternatives.

Better Warning Signals Still Need Stronger Proof

The study gains analytical depth from combining complementary methods, equal-length comparison windows and rolling estimates that track changing relationships. Its statistical tests also adjust for multiple comparisons, reducing the likelihood of identifying chance connections. These strengths make the framework useful for examining vulnerability, but they do not establish that it can reliably predict the next crisis.

Several limitations affect interpretation. The ten-currency sample excludes economies such as India and Vietnam, while daily observations cannot capture the fastest intraday transmission. Portfolio weights follow normalized exchange-rate paths rather than actual banking or corporate exposures. Shared movements against the dollar could also help explain some apparent regional connections, complicating efforts to distinguish direct spillovers from common external influences.

An additional clarification concerns the treatment of currency movements. Although the paper defines positive returns as depreciation against the dollar, parts of its discussion describe appreciation and depreciation inconsistently. The relationship between that quotation convention and its downside-loss measures needs clarification before those estimates can be interpreted straightforwardly as measures of depreciation-related financial stress.

The proposed concentration thresholds are also descriptive reference points, not empirically established crisis triggers. Further research should test warning performance against historical crises and previously unseen data, including false alarms and missed events. Broader currency coverage and direct measures of financial exposure would help establish whether the framework improves decisions beyond existing surveillance tools.