💸 モジュール5.4：債務サイクル — ミンスキー & ラインハート

Hyman Minsky (1919–1996) was an American post-Keynesian economist at Washington University in St. Louis. His life's work — the Financial Instability Hypothesis (FIH) — was largely ignored during his lifetime and only widely received after the global financial crisis of 2008. The central idea: stability breeds instability. Extended calm periods gradually tempt market participants into ever riskier financing — until the system collapses.

3つの資金調達フェーズ

Minsky distinguishes three qualitatively different financing structures that can apply in this order to any market participant (household, firm, state):

• 1. Hedge financing (green, broad at bottom): Cash flow from the asset or business exceeds interest + principal. The borrower can service the loan from current income — entirely independent of refinancing or asset value. Classic example: a landlord whose rental income (after tax + costs) safely covers the monthly instalment.
• 2. Speculative financing (amber, middle): Cash flow covers only the interest, not the principal. The borrower must refinance at maturity. Works as long as refinancing remains available — becomes vulnerable as soon as banks tighten or interest rates rise. The subprime home buyer of 2005 with an ARM (adjustable rate mortgage) who planned "we'll refinance cheaper in 2 years".
• 3. Ponzi financing (red, apex): Cash flow does not even cover the interest. The position lives entirely on asset price appreciation: you sell later at a higher price or take out a second mortgage on the higher book value. If the price rally pauses even briefly, the entire chain collapses. NINJA loans (No Income, No Job, no Assets) and no-doc loans of 2006/07 were pure Ponzi structures.

2007年のサブプライム住宅ローン：典型的なポンジ

The US subprime market 2005–2007 illustrates Minsky's end phase exemplarily. Home buyers with weak credit ratings took out loans whose monthly payments exceeded their income. The logic only worked as long as house prices rose 10–15% every year — then one could refinance after 12 months or sell at a profit. When prices first stagnated in summer 2006 and then fell, the Ponzi layer collapsed in seconds. The chain reaction reached the speculative layer in 2007–08 (banks could no longer refinance in the interbank market) — and in 2008 finally also parts of the hedge layer. Minsky had described exactly this mechanism in 1986 in Stabilizing an Unstable Economy.

Practical significance for traders: the question is not "how high is the debt?", but "what proportion falls on each of the three layers?". An economy with 300% debt/GDP but 80% hedge share is more robust than one with 150% debt/GDP and 30% Ponzi share.

Carmen Reinhart (Harvard Kennedy School, formerly chief economist of the World Bank) and Kenneth Rogoff (Harvard, former IMF chief economist) published in 2009 the book This Time Is Different — Eight Centuries of Financial Folly. The work became the standard reference of crisis research and has since adorned virtually every CV of a central banker or finance minister.

データ基盤：8世紀、66カ国

The book quantitatively analyzes 800 years of financial crises across 66 countries — from Florentine banking crises of the 14th century to the subprime crisis of 2008. The database covers sovereign defaults, banking crises, inflation episodes, currency devaluations and debt restructurings. Reinhart and Rogoff document for the first time empirically that these events are not random, but follow recurring patterns.

主要な知見

• Sequence: Debt accumulation → banking crisis → sovereign crisis (state default) → high inflation or devaluation. The chain almost always runs in this order.
• Banking crisis follow-on costs: On average, public debt doubles in the three years after a banking crisis — through bailouts, recession-induced revenue shortfalls and fiscal support packages.
• "Bunching": Crises occur in clusters. A global wave is followed by 10–20 years of relative calm — until experience fades and the next wave rolls in.
• Equity markets recover faster than employment: Equities need on average 3.4 years to return to pre-crisis highs, unemployment 4.8 years.

「今回は違う」シンドローム

The book title is bitter irony. Before every crisis the current generation believes the old rules no longer apply — because of better monetary policy, smarter risk models, new technologies or an exceptionally strong growth model. Reinhart and Rogoff show: exactly this argument has been advanced before every single pre-crisis period. Before 1929 it was "the Federal Reserve has abolished bank panics", before 1997 it was "the Asian Tigers have found a superior capitalist model mix", before 2007 it was "subprime risks are diversified, CDOs resolve concentration risk".

書籍に掲載された4つの歴史的事例

• Spain in the 17th century: Six sovereign defaults in 100 years (1607, 1627, 1647, 1652, 1660, 1666) — Spain financed Habsburg wars through ever-new Genoese loans, until no investor would subscribe.
• Argentina in the 19th/20th century: Eight sovereign defaults since independence — the serial champion. Argentina was among the 10 richest countries in the world in 1900; today an emerging market with chronic hyperinflation.
• Asian crisis 1997: Thailand, Indonesia, South Korea — triggered by carry trades and short-term USD loans against local assets. Classic Reinhart/Rogoff pattern: boom in capital inflows → banking stress → currency crisis.
• USA 2008 (Global Financial Crisis): First fully global industrialized-country crisis since the 1930s. Confirmed Reinhart/Rogoff's prediction from the book (pre-publication manuscript already circulating in 2008): subprime triggers banking crisis → bailouts double public debt → years of weak growth.

The Bank for International Settlements (BIS) in Basel — informally the "central bank of central banks" — has published the 信用対GDP格差 quarterly since 2010 for around 45 countries. The indicator is part of the Basel III framework and steers the countercyclical capital buffers of large banks.

この指標が測定するもの

The gap is the deviation of the current credit-to-GDP ratio from its long-term trend. The trend is calculated using the Hodrick-Prescott filter with a high smoothing parameter (lambda = 400,000), so that only very slow structural shifts are captured. If the current ratio is 5 percentage points above the trend, the gap is +5%.

Why the trend, not the level? Japan traditionally has high credit/GDP ratios, emerging markets low ones — absolute figures are not comparable across countries. The deviation from the country-specific trend is, however, informative: it measures how quickly debt is currently building up relative to a country's own history.

早期警戒閾値：+10%

BIS research (Drehmann/Borio 2011, Detken et al. 2014) shows empirically: a gap above +10 percentage points indicates a banking or financial crisis probability of around 67% within the following 3 years. Below +2% the probability is about 5%. The threshold is one of the most reliable early warning signals crisis research knows — but it delivers no precise timing: between threshold crossing and an actual crisis there can be 1 to 5 years.

2026年の現状（例示）

The following values are illustrative and may slightly differ depending on the BIS quarterly update. The table shows ten relevant countries with current gap levels:

Freely available: The current BIS credit gap data is published quarterly at bis.org/statistics/c_gaps.htm. Update frequency: ~10 weeks after end of quarter.

When debt becomes unsustainable, there are only five ways out: growth (rare), austerity (politically tearing), default (reputational damage), inflation (hidden default) or reset (radical reordering). History shows: inflation or reset solutions dominate once the mountain of debt reaches a critical size. Four paradigmatic cases:

1. ドイツ1923年 — ハイパーインフレによる債務リセット

トリガー： Versailles reparations of 132 billion gold marks, plus wartime debt from WW1, Ruhr occupation by France/Belgium in 1923 (coal deliveries as payment in kind failed).
メカニズム： The Reichsbank printed paper marks at a rapid pace to make reparation payments — from 320 marks/USD in January 1922 the exchange rate rose to 4,200,000,000,000 (4.2 trillion) marks/USD in November 1923. The volume of all domestic debt in marks was thereby effectively devalued to practically zero in real terms.
トレーダーへの教訓： When the debt burden in the domestic currency is extreme, the money-printing route remains politically always open. Those who held debt in dollars or real assets (houses, equities, gold) in 1923 survived; those who held mark-denominated bonds or life insurance policies lost everything.

2. アメリカ1971年 — ニクソン・ショックとブレトンウッズ体制のリセット

トリガー： Vietnam War + Great Society programs produced US deficits incompatible with the USD gold peg (35 USD/ounce, fixed since Bretton Woods 1944). France and the UK were increasingly redeeming their USD reserves for gold.
メカニズム： On 15 August 1971 Nixon announced the unilateral suspension of gold convertibility. The USD lost roughly 60% of its gold value between 1971 and 1980; gold rose from 35 USD/ounce to 850 USD/ounce (1980). In effect a global debt restructuring in favor of the USA.
トレーダーへの教訓： The current fiat money system is only 50 years old — a short span historically. Reset risk remains latently present, especially if the USA over-extends its privileges.

3. ロシア1998年 — ルーブル切り下げとGKOデフォルト

トリガー： Asian crisis 1997 + oil price collapse + overvalued ruble + speculative capital outflows. Russia had borrowed short-term in rubles (GKO bonds, yields 50–150%) and in USD.
メカニズム： On 17 August 1998 the Russian government declared a GKO default and ruble devaluation: the exchange rate fell within weeks from 6 to 24 rubles/USD (−75%). Russia halted foreign payments for 90 days. Long-Term Capital Management (LTCM, US hedge fund with Nobel Prize laureates) collapsed as a knock-on effect — the Fed had to organize a rescue.
トレーダーへの教訓： Emerging market bonds with double-digit yields are never "free money". Carry trades work — until they lose 50% of book value within days. Always validate correlation assumptions in crisis scenarios with a crash stress test.

4. 2008年世界金融危機 — 現代の公的救済モデル

トリガー： Subprime mortgages (Ponzi layer in Minsky's sense) → Lehman bankruptcy September 2008 → global interbank market frozen.
メカニズム： The US Treasury launched TARP (Troubled Asset Relief Program, $700 bn), the Federal Reserve expanded its balance sheet from ~$900 bn (pre-crisis) to $4.5 trillion (QE1, QE2, QE3). Comparable programs in the EU (LTRO, later OMT, then APP), UK (APF), Japan (QQE). Debt was not restructured, but nationalized and partially transferred to central-bank balance sheets.
トレーダーへの教訓： In the modern system central banks monetize crises rather than allow defaults. Consequence: mountains of debt continue to grow, inflation becomes the main adjustment variable (instead of defaults). Real asset classes — gold, real estate, equities of high-quality value companies — benefit long-term. Bond investors lose silently.