Market margin maintenance ratio
How much cushion the whole market's leveraged holders have before their brokers start selling them out. The market-wide version of this ratio stopped being published as an official statistic and largely disappeared from the public record; the daily files behind it did not, so the series is rebuilt here and runs unbroken from 2004 to today.
- Most places either dropped this line or kept a few recent years. This is the full daily history, over twenty years of it.
- When the market-wide ratio approaches the 140 percent band, forced selling from margin calls starts amplifying declines rather than following them.
- The reconstruction method and every approximation in it are written out below. Nothing here is a black box you have to take on trust.
Try it
The chart opens on the last three years. Drag the zoom bar under it to pull the window back to 2004, or use the buttons. The dashed lines mark the traditional 140 percent warning band and a 156 percent recent reference level.
How it is built
Four steps, each of which you could redo yourself from the same public files. The interesting one is step two, because the piece the formula needs is the piece nobody publishes.
Value the collateral at the daily close
Margin balances are published every day as share counts per ticker. Each balance is valued at that day's close and summed across the market, which gives the market value of everything currently pledged against margin loans.
Rebuild the loan amount, because it is not published
The official daily files carry balances in shares, never the money borrowed against them. The loan side is rebuilt with a weighted average cost method: each ticker's average margin purchase cost is rolled forward day by day as new margin buying comes in, then multiplied by the margin ratio to get the amount lent.
Divide one by the other
Market maintenance ratio = collateral value divided by loan amount. A reading of 160 percent means the pledged shares are worth 1.6 times what was borrowed against them. The closer the market gets to the 130 to 140 percent band, the more accounts sit near a forced sale.
State the approximations instead of hiding them
Known limits, in the open: listed common shares only, the margin ratio is held at its normal level rather than tracked per ticker, and the effect of ex-dividend dates on carried cost is not modelled. Levels and turning points are meaningful; the second decimal place is not.
Terms
Four words carry most of this page:
Where the line is
This is a reconstruction, not a republication. The collateral side is straightforward — published balances at published closes — but the loan side has to be rebuilt, because share counts are public and the money borrowed against them is not. The weighted average cost method used here is a reasonable way to do that and it is still an approximation: a market where margin positions turn over faster than average will carry a slightly stale cost basis, and ex-dividend adjustments to that basis are not modelled at all.
What follows from that is a limit on how the number should be read. The level and the turning points survive the approximation; a half-point move between two adjacent days does not. It is also a market aggregate, so it says nothing about the distribution underneath — a market at a comfortable 160 percent can still hold a tail of accounts sitting at 135. This page publishes a temperature reading with its error bars described, not a count of accounts at risk, and no threshold on it was picked by checking which one would have called past declines best.
FAQ
Why is this series hard to find anywhere else?
The market-wide maintenance ratio stopped being published as an official statistic, and most data vendors dropped it or kept only a few recent years. The underlying daily files it was built from are still public, so the series is rebuilt here from those files and runs from 2004 to the present without a gap.
Is this the official number?
No, and it is not presented as one. It is a reconstruction from public daily data using a documented method, with the approximations listed on this page. Treat it as an estimate whose level and turning points are informative, not as a published statistic.
What does a reading of 160 percent actually mean?
That the shares pledged across the whole market are worth about 1.6 times the money borrowed against them. Higher means more cushion before margin calls start; lower means a smaller fall is enough to trigger them. Individual accounts vary widely around the market figure, which is why this is a temperature reading rather than a count of accounts at risk.
Why 140 percent and 156 percent on the chart?
140 percent is the traditional warning band, close enough to the call level that a broad slice of accounts is under pressure. 156 percent is drawn as a recent reference level so the current reading has something to sit against. Neither line is a signal and neither was chosen by testing which threshold would have looked best in hindsight.
Do I need a key or a paid tier?
No. This is a single market-level series, free on the API with no signup, full history included. The endpoint shown above the chart takes a range parameter of 1y, 3y, 5y, 10y or all.
Run it yourself
The free tier needs no signup — call this feature's API straight away with the demo key. Unlock as-of history and full ticker coverage on a paid tier.