Key Takeaways

  • High margin debt is a vulnerability amplifier. It says how far losses could spread if stress arrived, not whether stress is coming.
  • The panel reads leverage against its own history, then asks whether credit, volatility and breadth independently show stress. The two answers are never blended into one score.
  • Changes use the prior calendar month. A missing month leaves a gap and a stated reason, never a guess or a zero.
  • Margin data are monthly and delayed. The panel shows how old the latest figure is against its own cadence.
  • The default data are an illustrative sample with fictional dates. Live feeds are not connected: paste the monthly figures from the primary source to analyse real values.

Why Is Leverage a Vulnerability Amplifier and Not a Direction Signal?

Margin debt is borrowed money held against securities. When prices fall, lenders can require more collateral, and holders who cannot supply it may have to reduce positions. That mechanism can make a decline larger or faster than it would otherwise be. It does not cause the decline, and it says nothing about when a decline might begin.

Aggregate margin balances have stayed elevated through long expansions. A high reading is therefore a description of fragility: the system could transmit a shock more forcefully. Whether a shock arrives is a separate question, which is why the panel asks credit, volatility and breadth for their own independent states instead of folding them into the leverage reading. The full background is in the guide on margin debt and leverage.

Leverage Context Panel

Illustrative sample loaded. The values and dates below are fictional and are not current or historical market data. Live feeds are not connected. Paste monthly figures from FINRA margin statistics to analyse real values.

Margin statistics

One row per month: date (YYYY-MM-DD), debit balance, free credit in cash accounts, free credit in margin accounts, then optionally equity market value and its date. Separate values with commas or tabs, without thousands separators. Leave a cell empty for a missing value: it stays missing and is never read as zero.

Window and dates

A percentile needs at least 36 monthly observations in the window, so the 1-year window shows none, and a 3-year window shows none if even one month is missing.

Observations dated after this are ignored, so no percentile uses a later value.

When blank, publication is assumed 21 days after month end and flagged as an assumption.

Confirmation states

Each family has its own state, observation date and cadence. Take them from the credit, volatility and breadth sources you trust. They are judged for freshness separately and never change the leverage reading.

Confirmation inputs: state, observation date and cadence for credit, volatility and breadth.
FamilyStateObserved onCadence

What this does not mean

  • Leverage is not a timer for market declines. High aggregate leverage can persist through long expansions without any stress following.
  • Aggregate data do not describe an individual investor. A balance across all margin accounts says nothing about any one account.
  • Margin statistics are delayed. They are published after each month ends and are not a real-time reading.
  • Financing conditions and portfolio structures differ across market participants, so the same aggregate balance carries different meaning for different holders.

How Does the Fragility Confirmation Matrix Work?

The matrix crosses the leverage state with the combined credit and volatility state. Leverage is elevated at or above the 80th percentile, low at or below the 20th, and typical between. Credit and volatility are each supplied as calm or deteriorating, and the pair counts as deteriorating when either one is. A stale or unavailable family is excluded and reported as incomplete, never read as calm.

Leverage state by combined credit and volatility state
LeverageCredit and volatility calmCredit or volatility deteriorating
Elevatedelevated vulnerability, unconfirmed stresselevated vulnerability with stress confirmation
Typicaltypical leverage, no stress confirmationmarket stress without exceptional leverage signal
Lowlimited leverage signalmarket stress with a limited leverage signal

Breadth appears as its own row beside credit and volatility so you can see whether participation agrees, but it does not move the matrix. Learn how each family reads in the guides on credit spreads and risk appetite and VIX term structure.

How the Panel Computes Each Measure

  1. Changes. The 1-month, 3-month and 12-month changes compare the latest figure with the observation exactly that many calendar months earlier. If that month is missing the change is not computed and the reason is shown. The previous row is never substituted. The 3-month change is reported as cumulative by default, or annualized if you choose, and both are returned.
  2. Percentile. The panel ranks the latest value within the chosen window using only observations dated on or before the as of date. It needs at least 36 monthly observations, and it ranks the nominal balance, the 12-month change and, when available, the ratio separately.
  3. Ratio. Margin debit divided by equity market value, shown only when the market value carries the same date as the debit balance. A different date disables the ratio. Nothing is interpolated, because interpolation creates false precision.
  4. Free credit. Free credit in cash accounts and in margin accounts is shown beside debit with its own 12-month change. It is never netted into debit.
  5. Freshness. The latest margin figure is judged against a monthly cadence, and each confirmation family against its own cadence. A missing release is listed as awaited, never treated as zero.
  6. Leverage state. Based on the ratio percentile when a ratio exists, otherwise the 12-month change percentile, otherwise the nominal percentile with a stated caution.

The leverage family, its two components and its weight follow the published sentiment composite framework, and the timing vocabulary follows the guide on data latency and vintages.

What This Tool Does Not Do

  • It fetches nothing. There is no live connection to FINRA, FRED, Cboe or any other source. Values come from what you paste or from the labelled illustrative sample.
  • It does not forecast. Every label describes the evidence supplied, not an outcome.
  • It is an aggregate view. It cannot describe any single account, firm or holder.
  • It does not read prose as a series. Financial-system leverage commentary, such as the Federal Reserve Financial Stability Report, is narrative and is not converted into a numeric series here.
  • Its thresholds are conventions. The 80th and 20th percentile cut-offs are published here and applied mechanically. They are not statistical tests.
  • Nothing here is investment advice. It is a reading aid for primary data. See risk management for position-level thinking.

Frequently Asked Questions

What does margin debt measure?

Margin debt is the aggregate debit balance in customers' securities margin accounts at member firms, which is money borrowed against securities to hold positions. FINRA reports it monthly alongside free credit balances in cash accounts and in margin accounts. It is an aggregate across all participants, so it describes how much borrowing exists in the system, not what any one investor holds or owes.

Does high margin debt mean a market decline is coming?

No. High aggregate leverage describes how much a shock could be amplified, not when or whether a shock arrives. Margin balances have stayed high for long stretches of expansion without stress following. The panel therefore pairs the leverage reading with independent credit, volatility and breadth states and labels the result as vulnerability with or without confirmation.

Why does the panel use percentiles and a ratio instead of the raw balance?

Nominal margin balances trend upward with market size and inflation, so a raw balance tends to look high late in any long expansion. The panel shows the balance as reported, then the 12-month change, the percentile within a chosen lookback window, and a ratio to equity market value when a denominator with the same date exists. Those scale-aware views are kept separate from the nominal value.

Why is free credit not subtracted from margin debt?

No published methodology supports a net figure, and free credit sits in two different kinds of account. The panel shows free credit in cash accounts and free credit in margin accounts beside the debit balance and never nets them. Netting would imply that idle cash offsets borrowing for the same holders, which aggregate data cannot establish.

How old is margin statistics data?

Margin statistics are monthly and are published after the month ends, so the latest figure always describes a date already in the past. The panel judges freshness against the monthly cadence using the publication date you enter, or a flagged assumed lag when you leave it blank. A release that has not yet appeared is listed as awaited and is never treated as a zero balance.

What does elevated vulnerability, unconfirmed stress mean?

It means aggregate leverage is high relative to its own selected history while the credit and volatility states supplied are calm. Leverage could amplify an adverse move if one occurred, but no cross-market stress is visible in the inputs. It is a description of the current evidence, not a forecast and not a recommendation.

References

These are the primary sources for the series the panel describes. The page is educational, and its sample data are fictional and labelled as such.

Jurisdiction: United States. Last reviewed by the Swoopr Editorial Team in October 2026. This page is educational and is not personalized investment advice.

Related Reading