Where every figure came from
Financial Services Agency, Program for Financial Revival: Revival of the Japanese Economy through Resolving Non-Performing Loans Problems of Major Banks (30 October 2002, provisional translation) (The target and how the FSA set out to reach it): https://www.fsa.go.jp/news/newse/e20021030.pdf. FSA terms: Public Data License 1.0 (CC BY-compatible), source cited; paraphrased. Opening: the FSA set out to normalise the major banks' non-performing loan problem in fiscal 2004 by cutting their non-performing loan ratio to about half, through stricter assessment of assets, stronger capital and better governance, aiming at a financial system people could rely on and that could support structural reform. Section 3(1): for large borrowers classed as needing special attention, major banks were in principle to set provisions loan by loan on a discounted cash flow basis; another round of special inspections would check, ahead of the March 2003 accounts, how accurately banks classified borrowers; the FSA would publish, in aggregate and periodically, the gap between the major banks' own assessments and its inspection results and ask each bank to narrow it, with a business improvement order for a bank that did not. Section 3(2): deferred tax assets were treated as a less solid part of capital, to be evaluated strictly, with an upper limit on what counts as regulatory capital to be examined. Section 3(3): external auditors were to audit strictly the banks' asset assessment, provisioning and write-offs. Section 1(3) heading: the target year is fiscal 2004 (to March 2005). Retrieved 3 October 2026.
Hiroshi Nakaso, Bank for International Settlements, BIS Papers No 6, The financial crisis in Japan during the 1990s: how the Bank of Japan responded and the lessons learnt (October 2001) (The support for the banks before 2002): https://www.bis.org/publ/bppdf/bispap06.pdf. BIS: brief excerpts with the source stated; otherwise paraphrased. Section 1.4.2: the October 1998 laws doubled the public funds available from ¥30 trillion to ¥60 trillion and allowed failing banks to be nationalised. Sections 1.4.1 and 1.5.1: the Long-Term Credit Bank was nationalised on 23 October 1998 and Nippon Credit Bank on 13 December 1998. Section 1.5.2: in March 1999 public capital was put into 15 major banks. Read from https://www.bis.org/publications/paper-6-financial-crisis-japan-during-1990s-how-bank-japan-responded-and-lessons-learnt.pdf. Retrieved 3 October 2026.
Financial Services Agency, The Progress of "Program for Financial Revival" (31 December 2004, provisional translation) (The major banks' bad-loan ratio, 2002 to 2004; the inspections and the published gap): https://www.fsa.go.jp/en/policy/progress.pdf. FSA terms: Public Data License 1.0, source cited. The major banks' NPL ratio, by date: March 2002 8.4%; September 2002 8.1%; March 2003 7.2%; September 2003 6.5%; March 2004 5.2%; September 2004 4.7%. The same table records that the FSA disclosed the gap between the major banks' self-assessments and its inspection results on 8 November 2002, 9 September 2003 and 16 September 2004, and announced special inspection results on 25 April 2003, 14 November 2003, 27 April 2004 and 12 November 2004. The IMF's figures for fiscal 2002 (8½ and 7¼ per cent, PIN 03/112) are the same series rounded differently; the FSA's are used. Retrieved 3 October 2026.
Financial Services Agency, FSA Newsletter, March 2006: Status of Non-Performing Loans (as of End of September 2005) (What the bad-loan ratio measures): https://www.fsa.go.jp/en/newsletter/2006/03b.html. FSA terms: Public Data License 1.0, source cited; paraphrased. The ratio is non-performing loans (loans disclosed under the Financial Reconstruction Law) divided by total credit; the newsletter gives the major banks at 4.7% (September 2004) and 2.9% (March 2005), and calls the halving objective successfully achieved, in the FSA's own words. Register rows F115 to F117 (s3/figures_s3.py), read twice: the cached file and an independent second download, s3/read2/fsanl0603.html, 4 October 2026. Retrieved 4 October 2026.
Cabinet Office, Government of Japan, Annual Report on the Japanese Economy and Public Finance 2003, Summary (October 2003), chapter 1 and chapter 2 (Write-offs in fiscal 2002, and the turn in the business cycle): https://www5.cao.go.jp/keizai3/2003/1024wp-keizai/summary2.pdf. Paraphrased. Chapter 2, Section 2, Rebuilding Finance: in fiscal 2002 the major banks' outstanding non-performing loans fell by ¥6.5 trillion to ¥20.2 trillion; loans to borrowers in danger of bankruptcy or below were actively written off from balance sheets; new bad loans fell; the cost of disposing of bad loans was larger than the banks' core operating profit. Chapter 1, Section 1 (https://www5.cao.go.jp/keizai3/2003/1024wp-keizai/summary1.pdf): the business cycle troughed in January 2002, with a recovery phase after it (the page: the bottom of the downturn, after which output began to recover), and the recovery originated from exports and from positive signs mainly in the business sector, starting with improving corporate profits (the Cabinet Office's account). Retrieved 3 October 2026.
Financial Services Agency, Statement by the Minister for Financial Services, Normalization of the NPLs Problems of Major Banks (25 May 2005, provisional translation); and Financial Statements of Major Banks: March 2005 (25 May 2005) (What the FSA said about its target in May 2005): https://www.fsa.go.jp/en/announce/state/20050525-1e.html. FSA terms: Public Data License 1.0, source cited; brief quotation. The Minister said that, on the major banks' results for fiscal 2004, their aggregate non-performing loan rate "has dropped to 2.9%", and that the goal of halving the March 2002 ratio "has thus been achieved". This is the FSA's own statement on its own target. The statement also describes the bad loans as a heavy drag on the economy since the bubble burst, and says financial administration was moving from emergency responses to a forward-looking phase. The FSA table (https://www.fsa.go.jp/news/newse/e20050525-1.pdf), row 11 major banks total, gives 2.93%. Both read twice: the statement from a second independent download, the table from the Internet Archive capture https://web.archive.org/web/20061018002615id_/http://www.fsa.go.jp:80/news/newse/e20050525-1.pdf (capture 2006-10-18), whose row matches the live file. Retrieved 4 October 2026.
World Bank, World Development Indicators: GDP growth (annual %) (NY.GDP.MKTP.KD.ZG), inflation, GDP deflator (annual %) (NY.GDP.DEFL.KD.ZG) and unemployment, total (% of total labor force, modeled ILO estimate) (SL.UEM.TOTL.ZS), Japan (Growth and unemployment, 2002 to 2006; GDP in money terms): https://data.worldbank.org/indicator/NY.GDP.MKTP.KD.ZG?locations=JP. CC BY 4.0, credit World Bank. Data last updated 13 July 2026, retrieved 3 October 2026 through the World Bank API and downloaded again independently (identical, re-compared 4 October 2026). GDP growth, per cent: 2002 0.08, 2003 1.38, 2004 2.11, 2005 1.85, 2006 1.55, each year from 2003 to 2006 above 1 per cent. Unemployment (https://data.worldbank.org/indicator/SL.UEM.TOTL.ZS?locations=JP) is highest in 2002 of the years 1991 to 2006 and lower in each year from 2003 to 2006. Derived here: "barely grew in 2002"; and GDP in money terms, approximated as (1 + real growth) x (1 + change in the GDP deflator) - 1 from the rows above and the deflator rows (data/wb_NY.GDP.DEFL.KD.ZG.json, read twice, identical): 1997 1.4, 1998 -1.9, 1999 -1.7, 2000 1.7, 2001 -0.7, 2002 -1.3, 2003 -0.4, 2004 0.8, 2005 0.5, 2006 0.6, so it shrank in 5 years after 1997 (1998, 1999, 2001, 2002, 2003) and grew by less than 2 per cent in every year from 1997 to 2006.
IMF, Monetary and Financial Statistics: Japan, share price index, monthly average, 2010=100 (JPN.EQTS.PA_IX.M) (Share prices from their 2003 low): https://api.imf.org/external/sdmx/2.1/data/IMF.STA,MFS_FMP/JPN.EQTS.PA_IX.M. IMF data, reused with credit: IMF, Monetary and Financial Statistics. The index reaches its lowest point between 1990 and 2007 in April 2003 (89.4) and stands at 181.9 in December 2005. Derived here: about twice the April 2003 level. Retrieved 3 October 2026 through the IMF data API and read twice (independent second download, re-compared 4 October 2026).
Bank of Japan, Change in the Guideline for Money Market Operations (9 March 2006) (The end of quantitative easing): https://www.boj.or.jp/en/mopo/mpmdeci/mpr_2006/k060309.htm. Bank of Japan text, paraphrased with credit (the Bank's terms allow copying with the Bank credited, not commercial copying). The Bank moved its operating target from the outstanding balance of current accounts at the Bank back to the uncollateralised overnight call rate, to stay at effectively zero; current-account balances were to come down over a few months towards the level of required reserves. It said year-on-year changes in the consumer price index had turned positive and were expected to stay positive, and judged that the conditions of its 2001 commitment (consumer prices excluding fresh food stably at zero or above year on year) were fulfilled. It described the economy as continuing to recover steadily, with business fixed investment rising on high corporate profits and private consumption solid. Retrieved 3 October 2026.
Bank of Japan, (Reference) Unconventional Monetary Policy Measures from the Late 1990s (The dates of quantitative easing and of the later easing): https://www.boj.or.jp/en/mopo/outline/bpreview/ref.htm. Bank of Japan text, paraphrased with credit. The Bank lists its Quantitative Easing Policy as March 2001 to March 2006, and Quantitative and Qualitative Monetary Easing (QQE) as April 2013 to March 2024. Retrieved 3 October 2026.
IMF, World Economic Outlook database, April 2026: general government gross debt, Japan, per cent of GDP (GGXWDG_NGDP) (General government gross debt, 1990 to 2006; the chart (1992 to 2006)): https://api.imf.org/external/sdmx/2.1/data/IMF.RES,WEO/JPN.GGXWDG_NGDP.A. IMF data, reused with credit: IMF, World Economic Outlook database, April 2026 (WEO 9.0.0, published 14 April 2026), read through the IMF SDMX API on 3 October 2026 and downloaded again independently (identical, compared 4 October 2026). Rows in the lane media pack s3/m8_debt_rows.json. General government is central government, local government and social security funds; gross debt, unconsolidated. Per cent of GDP: 1990 54.8 (stated in the text, not charted); charted: 1992 57.9, 1993 63.2, 1994 73.3, 1995 80.7, 1996 85.3, 1997 91.3, 1998 101.6, 1999 113.5, 2000 118.5, 2001 126.8, 2002 134.6, 2003 140.2, 2004 148.8, 2005 153.4, 2006 152.1. Rounded on the page to 55 (1990, stated in the text) and 153 (2005); the chart plots 1992 to 2006. Derived here: the 2005 value is 2.8 times the 1990 value, "nearly trebled"; the ratio is highest on the chart in 2005. Older IMF vintages give different levels (PIN 03/112 put end-2002 gross debt higher); this module uses the April 2026 data and names them.
IMF, Public Information Notice 03/112: IMF Concludes 2003 Article IV Consultation with Japan (September 2003) (Deficits and the rising debt): http://www.imf.org/external/np/sec/pn/2003/pn03112.htm. IMF copyright; paraphrased. Background: fiscal policy was expansionary in fiscal 2002; before adjustment for the economic cycle the deficit rose to 7½ per cent of GDP, mainly because revenue fell; continued deficits, the IMF wrote, had fed rapidly growing government debt. Read from the Internet Archive capture of the original file, https://web.archive.org/web/20041215212340id_/http://www.imf.org/external/np/sec/pn/2003/pn03112.htm (capture 2004-12-15). Retrieved 3 October 2026.
All wording is our own. Charts are drawn from the data named under them.