Important interpretation. A score of 9.2/10 describes accumulated structural pressure under this model. It is not a probability, a market forecast, a recommendation, or a measure of how close a crash is in time. Unless otherwise stated,
P(x) is the inclusive empirical percentile—observations less than or equal to the latest reading divided by all published observations in that same series. For inverse measures, observations greater than or equal to the latest reading are counted. No missing observations are estimated. Components are calculated at full precision and displayed to one decimal. The fixed metric weights are policy judgements, not values fitted to maximise past crash prediction.
clamp[0,10] limits a result to the zero-to-ten range.
Open the machine-readable calculation register.
30% weight
1. Valuation pressure
8.8
In plain English
Measures how much investors are paying for each dollar of sustainable company earnings, viewed through three different valuation lenses.
Why it is a crash indicator
High valuations do not cause crashes or identify their date. They reduce the margin for disappointment: when expectations change, prices may have much further to fall before reaching historically normal relationships.
Current score
Shiller CAPE: 40.576 (2026-09), percentile 98.9%; S&P 500 forward P/E: 20.849 (2026-04), percentile 81.7%; inverse equity-risk premium: 2.796 (2026-04), percentile 79.5%. Score 8.8/10.
Formula
Valuation score = 10 × [ 40% × P(CAPE) + 30% × P(forward P/E) + 30% × P(negative equity-risk premium) ]
Public sources
Limitations and historical exceptions
Valuations can remain extreme for years. The three series begin in different years, and the current observations are published on different schedules. Forward P/E depends on analysts’ earnings forecasts; CAPE relies on ten years of inflation-adjusted earnings; the equity-premium estimate is model-dependent. The 1987 crash was not preceded by dot-com-style valuations, while some expensive markets have corrected gradually.
20% weight
2. System leverage
9.1
In plain English
Measures the borrowing most capable of forcing asset sales: customer margin debt, hedge-fund exposures and leverage inside major life insurers.
Why it is a crash indicator
Leveraged investors cannot always wait for prices to recover. Falling collateral values can trigger margin calls, redemptions and compulsory selling, turning an ordinary decline into a self-reinforcing event.
Current score
FINRA customer margin debt / GDP: 4.46% (2026-08; GDP denominator 2026-Q2), percentile 99.7%; qualifying hedge-fund gross notional exposure / NAV: 8.201 (2026-Q2), percentile 98.1%; top-ten hedge-fund leverage: 19.249 (2026-Q2), percentile 72.2%; large life-insurer assets / equity: 11.940 (2025-Q4), percentile 100.0%. Score 9.1/10.
Formula
Margin-debt ratio = FINRA customer margin debt ÷ nominal GDP
Aggregate hedge leverage = qualifying hedge-fund gross notional exposure ÷ qualifying hedge-fund NAV
Leverage score = 10 × [ 25% × P(margin-debt ratio) + 30% × P(aggregate hedge leverage) + 30% × P(top-ten hedge-fund leverage) + 15% × P(life-insurer assets ÷ equity) ]
Public sources
Limitations and historical exceptions
This version deliberately measures market-sensitive leverage rather than household mortgage stress. The series begin in 1997, 2013 and 2001, so it cannot be used as an identical numerical backtest for 1929, 1973 or 1987. Gross notional exposure can overstate economically offsetting positions, while netting can understate liquidity and counterparty risk. Form PF data are aggregated and reported with a lag. Leverage was central in 1929 and 2008, but was not the sole cause of every historical crash.
15% weight
3. Profit–labour divergence
10.0
In plain English
Measures whether corporate profits and financial wealth are rising much faster than workers’ share of economic output.
Why it is a crash indicator
Household consumption ultimately supports company revenue. If profits and asset prices detach from household purchasing power, growth may become increasingly dependent on borrowing, asset appreciation and continued confidence.
Current score
after-tax corporate profits / GDP: 11.91% (2026-Q2), percentile 100.0%; inverse nonfarm-business labour-share index: 93.446 (2026-Q2), percentile 100.0%. Score 10.0/10.
Formula
Profit–labour score = 10 × [ 50% × P(after-tax corporate profits ÷ GDP) + 50% × P(negative labour share) ]
Public sources
Limitations and historical exceptions
International profits can inflate domestic profit measures. Labour-share data are revised and do not capture every form of employee compensation. This is a long-term vulnerability rather than a precise clock, and was less decisive before 1987 than before 1929 or 2008.
8% weight
4. Household equity saturation
10.0
In plain English
Measures how much household financial wealth is already invested in shares.
Why it is a crash indicator
When available investment capital is already heavily committed, less new money remains to support further gains. A decline can also create a larger negative wealth effect and reduce spending.
Current score
Household equity / financial assets: 48.23% (2026-Q2), percentile 100.0%. Score 10.0/10.
Formula
Household equity saturation = household directly and indirectly held equities ÷ total household financial assets
Score = 10 × P(household equity saturation)
Public source
Limitations and historical exceptions
The measure rises partly because share prices have risen, so it overlaps with valuation. Ownership is concentrated among wealthier households. It was extreme before 2000 and elevated before 2008, but not unusually high before 1973 or 1987.
10% weight
5. Financial-market crowding
10.0
In plain English
Compares total equity value with both the money supply and underlying economic output.
Why it is a crash indicator
When financial claims grow much faster than money or productive output, unusually favourable assumptions are needed to sustain those claims. During a rush for liquidity, there can be far more assets for sale than cash willing to buy them.
Current score
domestic equity value / M2: 4.726 (2026-Q2), percentile 100.0%; domestic equity value / GDP: 3.355 (2026-Q2), percentile 100.0%. Score 10.0/10.
Formula
Liquidity ratio = total domestic equity value ÷ quarter-end M2
Economic ratio = total domestic equity value ÷ annualised quarterly GDP
Crowding score = 10 × [ 50% × P(liquidity ratio) + 50% × P(economic ratio) ]
Public sources
Limitations and historical exceptions
M2 is incomplete as a definition of market liquidity and its definition changes. Market-cap/GDP partly overlaps with valuation. The ratios are combined rather than counted separately. They were powerful before 2000 and 2008, but not consistently extreme before 1973 or 1987.
7% weight
6. Market concentration
9.4
In plain English
Measures how far the broad sharemarket has moved away from equal participation and toward dependence on a small number of very large companies.
Why it is a crash indicator
Disappointment in one or two dominant companies can affect the whole index. Passive funds, active funds and retirement accounts can also end up crowded into the same positions.
Current score
Retained approved Feb–Jun 2026 S&P observations; source access/structured inputs unavailable. These inputs were not refreshed this month.
Formula
N = number of index constituents; W1 = largest-company weight; W10 = top-ten weight; HHI = sum of squared decimal constituent weights; effective breadth = 1 ÷ HHI
Largest-company dominance = 10 × [1 − (1 ÷ N) ÷ W1]
Top-ten dominance = 10 × [1 − (10 ÷ N) ÷ W10]
Effective-breadth loss = 10 × [1 − (1 ÷ HHI) ÷ N]
Concentration score = equal-weighted mean of the three component scores
Public sources
Limitations and historical exceptions
This is a transparent cross-sectional concentration score, not a historical crash percentile. The HHI and index-characteristic observations have different publication dates, and class shares mean the index has slightly more securities than companies. Concentration can reflect genuine profitability and is not inherently unhealthy. It was important before 1973 and 2000 but was not a universal feature of 1987 or 2008.
7% weight
7. Speculative bubble anatomy
7.3
In plain English
Looks across public US industry portfolios for exceptional gains, accelerating prices and rapidly increasing volatility.
Why it is a crash indicator
Research covering US industries since 1926 found that large two-year industry run-ups materially increased the subsequent probability of a major industry collapse. Acceleration, volatility, turnover and issuance helped distinguish healthy growth from fragile bubbles.
Current score
Hardw leads the public industry screen (2026-08): 24-month market-relative return 76.4%; 12-month relative-return acceleration 45.8 percentage points; volatility ratio 1.86×. Score 7.3/10.
Formula
Market benchmark = Fama–French monthly market return, calculated as Mkt−RF + RF
Market-relative return = [product of (1 + industry return) ÷ product of (1 + market return)] − 1
For each of the 49 public industry portfolios:
Run-up = min(10, 10 × positive 24-month market-relative return)
Acceleration = min(10, 10 × positive [latest 12-month market-relative return − preceding 12-month market-relative return])
Volatility ratio = latest 63-trading-day realised volatility ÷ median of all rolling 63-day volatilities in the latest 756 trading days
Volatility = clamp[0,10](5 + 5 × [volatility ratio − 1])
Industry score = 50% × Run-up + 25% × Acceleration + 25% × Volatility
Metric score = highest industry score
Public sources
Limitations and historical exceptions
This is a disclosed adaptation of the published research, not the authors’ fitted model. Public turnover and issuance are not yet included. Selecting the maximum of 49 industries can produce false positives. An industry collapse need not become a system-wide crash, and the French files may be revised after publication.
3% weight
8. Geopolitical shock risk
7.9
In plain English
Measures the intensity of wars, military threats, terrorism and severe international tension reported by major newspapers.
Why it is a crash indicator
Geopolitical shocks can disrupt energy, trade, inflation, investment and confidence in markets that are already fragile. It is therefore an external shock multiplier, not a stand-alone timing signal.
Current score
recent GPR index since 1985: 117.919 (2026-08), percentile 77.6%; historical GPR index since 1900: 118.203 (2026-08), percentile 80.7%. Score 7.9/10.
Formula
Geopolitical score = 10 × [ 50% × P(recent GPR index since 1985) + 50% × P(historical GPR index since 1900) ]
Public sources
Limitations and historical exceptions
The index depends on news coverage and language. It was not elevated before several market crashes, including 1929, 1973 and 2000. It receives only a 3% weight because it measures external shock risk rather than accumulated financial imbalance.
Unscored context
Black swans
How this section should be read
The six scenarios on the Meridian are illustrations, not forecasts. No probability, score or timing claim is attached to any of them. Their purpose is to show how an external event could interact with elevated valuation, leverage, concentration and liquidity pressure.
Public context sources
Qualifications
The 400–800% range is the result of illustrative modelling shown in Chaos Awaits, pages 32–33; it is not an observed historical return, a promise, or a forecast. A severe event need not produce a market crash, and a crash need not require a dramatic external event. Public authorities identify these channels as risks requiring preparedness, not as certain outcomes. Scenario descriptions must not be read as statements that an event is likely, imminent or inevitable.
Current headline
Weighted structural pressure
9.2
30% × 8.8 Valuation pressure
20% × 9.1 System leverage
15% × 10.0 Profit–labour divergence
8% × 10.0 Household equity saturation
10% × 10.0 Financial-market crowding
7% × 9.4 Market concentration
7% × 7.3 Speculative bubble anatomy
3% × 7.9 Geopolitical shock risk
Headline calculated from full-precision metric scores = 9.1623 = 9.2/10 rounded
For comparison, the one-decimal display values produce 9.1660.
Sandwich Meridian summary
Accumulated structural pressure: 9.2/10 — Extreme. Financial markets are displaying historically exceptional valuation, market-sensitive leverage, concentration and crowding. This does not establish when a correction will begin, but it indicates that the potential consequences of a significant shock may be unusually severe.
