FinObservatory

Crisis atlas / CHN / 1994–1999

China, 1994–1999

One crisis episode: distinct crisis years merged across gaps of up to two non-crisis years. Types, source agreement, and macro context are drawn from the union of five primary chronologies.

CurrencyBanking

Crisis years in this episode

Every distinct (year, type) event inside the 1994–1999 window, with the chronologies flagging it and the count of agreeing sources. Start-year sources (GMD, JST) mark only the first year, so agreement falls in the continuation years.

  1. 1994
    Currencyrr1 source
  2. 1997
    Bankingrr1 source
  3. 1998
    Bankinggmd, laeven_valencia, rr3 sources
  4. 1999
    Bankingrr1 source

Source: Global Macro Database 2026_06 (Müller, Xu, Lehbib & Chen 2025) | Jordà-Schularick-Taylor Macrohistory R6 | Laeven & Valencia (2020, 2026) | Reinhart-Rogoff via HBS BFFS | ECB/ESRB Financial Crises Database Methodology

Laeven-Valencia banking-crisis detail

Episode-level fiscal cost and output loss from the systemic banking crises database. Fiscal cost is the gross outlay of banking-sector rescues; output loss is the cumulative deviation of real GDP from its pre-crisis trend, both as a percent of GDP.

Banking episode 1998
18.0%
Fiscal cost, % of GDP
19.4%
Output loss, % of GDP
1998
Episode end year

Source: Laeven & Valencia (2026), IMF WP/26/94, Systemic Banking Crises Database 1970-2025 Methodology

Macro context, 19872006

Seven years either side of the episode. Each panel has its own scale; the crisis window is shaded. Only indicators with data for China in this window are shown. Bank credit growth (year over year in total loans) is available for the 18 Jordà-Schularick-Taylor advanced economies.

Inflation (% y/y)
-100102030198719942006
Real GDP growth (% y/y)
051015198719942006
Policy rate (%)
051015198719942006
Govt debt (% of GDP)
0102030198719942006
Unemployment (%)
024198719942006
Current account (% of GDP)
-50510198719942006

Source: Global Macro Database 2026_06 (Müller, Xu, Lehbib & Chen 2025) | Jordà-Schularick-Taylor Macrohistory R6 (credit) Real GDP growth is derived as the year-over-year change in real GDP; current account and government debt are percent of GDP; policy rate is the central-bank rate; bank credit growth is the year-over-year change in JST total loans. Methodology

How each source dates this episode

The chronologies disagree on start years and durations. This table is generated from the event rows, not curated: it lists, per crisis type, which source flags which years inside the window.

Currency
  • Reinhart-Rogoff1994
Banking
  • Reinhart-Rogoff1997, 1998, 1999
  • Global Macro Database1998
  • Laeven-Valencia1998

Source: Global Macro Database 2026_06 (Müller, Xu, Lehbib & Chen 2025) | Jordà-Schularick-Taylor Macrohistory R6 | Laeven & Valencia (2020, 2026) | Reinhart-Rogoff via HBS BFFS | ECB/ESRB Financial Crises Database Methodology

Documented policy responses

3 interventions from the Metrick-Schmelzing banking-crisis database whose recorded year falls inside this episode window (1994–1999). Each row is one documented government or central-bank action, tagged with the database’s own intervention categories and short code, under the crisis code it assigns. Matching is exact: same country, intervention year inside the window. The database’s global-crisis rows (no country) are never matched here.

CHN-1992China
  1. Oct-1996Capital injectionsAHCI
CHN-1999China
  1. Oct-1998RestructuringRES
  2. 1999-2000Asset managementBBAM

Source: Metrick-Schmelzing Banking-Crisis Interventions Database (Yale Program on Financial Stability) Cite as: Metrick, Andrew, and Paul Schmelzing, "Banking-Crisis Interventions Across Time and Space," working paper, 2024 (dataset consulted 2026-07-10). License: Creative Commons attribution. Methodology

Policy-response case studies (external)

For qualitative accounts of how authorities intervened in systemic crises, see the Yale Program on Financial Stability New Bagehot Project, a library of financial-crisis intervention case studies. It is a separate qualitative resource; no specific case is asserted to match this episode.