Using Margin With Debt Near $1.5 Trillion: Maintenance Calls, Position Sizing and When to Cut Back

October 25, 2026

FINRA's own margin statistics put investor margin debt at roughly $1.5 trillion as of mid-2026, after a 140% climb from the most recent low at the end of 2022. That raw dollar figure is a record. It is also, by itself, close to meaningless — every prior cycle set a new nominal record too, because the market it is borrowed against has also grown. The more useful question is not "is this a record," but "where does this sit next to the handful of times leverage expanded this fast before, and what happened afterward." The judgment call this data supports: the pace of the recent increase is genuinely unusual by historical standards, even though the debt load relative to the market it is backing is not.

This article is for readers deciding what to do with a margin account now: maintenance calls, sizing and when to cut back. For what followed the 2000, 2007 and 2021 margin-debt peaks, see What Happened After the 2000, 2007 and 2021 Leverage Peaks.

Illustration only, not tied to this article's specific figures. Photo: Kidfly182, Wikimedia Commons, licensed under CC BY-SA 4.0.

Which of five margin situations you are reading this from

Which of those two readings should change what you do depends less on the market than on where you personally stand with borrowed money. The table below sorts the five situations a reader of this article is likely to be in, and names the one number that is actually load-bearing for each. The rest of the article works through the rows in order.

If this is your situation	The number that matters for you	What the record says about it
No margin account; index funds or cash only	The aggregate FINRA debit balance, as background only	Other investors being forced to sell moves prices you hold: the three prior peaks were followed by S&P 500 declines of 49%, 57% and 25%
Margin account open, balance at zero	Your broker's maintenance-margin percentage (FINRA floor 25%, brokers commonly 25-30%)	Nothing is triggered at a zero balance; the trigger math below starts applying the day you borrow
Margin used, cushion not checked recently	Your account's current equity ratio against that maintenance level	On a 50%-funded position at a 25% requirement the call lands after a 33.3% decline; at 30% it lands after 28.6%
Margin used and the cushion is thin	The balance you could repay without forced selling through a 25%-57% drawdown	That is the range actually recorded after the 2000, 2007 and 2021 peaks
Deciding whether to use margin at all	Margin debt as a share of S&P 500 market capitalization: 2.3% in Q2 2026	Equal to the 1997-onward average, which is why the dollar record and an ordinary ratio are both true of the same $1.5 trillion

The record dollar figure, and the figure underneath it

Margin debt is money borrowed against securities already held in a brokerage account — a lever that can magnify both gains and losses. FINRA has published monthly debit-balance data since the 1990s, which is why the comparisons below are possible at all. The approximate month-end peaks, rounded to the nearest billion, are:

Peak month	Margin debt level (approx.)	What followed
March 2000	~$279 billion	S&P 500 fell 49% from March 2000 to October 2002
July 2007	~$381 billion	S&P 500 fell 57% from October 2007 to March 2009
October 2021	~$936 billion	S&P 500 fell 25% from January to October 2022
June 2026 (most recent FINRA reading)	$1,502,072 million (≈$1.50 trillion)	Not yet known — this is the open question

June 2026's figure was up 6.1% from May 2026's $1,415,557 million — the third consecutive monthly rise — and up 49.0% year-over-year from June 2025's $1,007,961 million. An August 2026 reading, per a mid-October 2026 commentary citing FINRA's own data, held near the same roughly $1.5 trillion level. Longer term, real (inflation-adjusted) margin debt has risen about 604% since 1997, versus 354% for the inflation-adjusted S&P 500 over the same stretch — leverage has grown faster than the market it is financing, not merely alongside it.

Dollar figures alone are not comparable across 26 years of market growth, so the more apples-to-apples metric is margin debt as a share of S&P 500 market capitalization. In the second quarter of 2026, that ratio stood at 2.3% — which, as of that reading, was exactly equal to the historical average for this ratio going back to 1997. That is the counterintuitive part of this story: the headline dollar figure is an all-time record, but the same debt measured against the market's own (much larger) size is sitting at a long-run average, not an extreme. Both of those statements are true about the same $1.5 trillion at the same time; which one matters more depends on whether you think the risk that matters is the absolute size of leveraged positions that could be forced to unwind, or the debt's size relative to the market's capacity to absorb a shock.

The maintenance number that decides when the choice stops being yours

The mechanism behind "forced selling" is worth making concrete, because it explains why margin debt matters even when the aggregate ratio looks ordinary. Under Regulation T, a US investor can typically borrow up to 50% of a stock purchase's value — buying $20,000 of stock with $10,000 of their own cash and a $10,000 margin loan. That starting position has $10,000 of equity against $20,000 of securities, a 50% equity ratio. Most brokers require a maintenance margin of 25% to 30% of the account's current value (FINRA sets a 25% floor; individual brokers commonly set it higher). Using a 25% maintenance requirement, a margin call triggers once the account's equity ratio falls to 25% — which, holding the $10,000 loan fixed, happens when the stock's value has fallen to $13,333, a 33.3% decline from the original $20,000. At a broker's stricter 30% maintenance level, the trigger comes sooner: at $14,286 in stock value, only a 28.6% decline. Once that line is crossed, the broker can sell the position without waiting for the investor's approval, and does not have to sell at a price the investor would have chosen. That mechanical trigger — not investor sentiment — is what converts a price decline into forced selling, and it is the specific risk that a larger aggregate margin-debt balance scales up across the market at the same time.

The pace reading, and how small its sample is

Rather than just comparing single-month peaks, it is worth looking at how fast margin debt climbed in the run-up to the 2000 and 2007 peaks, because 2026 has already produced a comparably fast move. From March to May 2026, margin debt rose 15.9% over those two months combined, one of the fastest two-month increases on record. For comparison:

January 2000: margin debt rose 19.3% in two months. Over the following 24 months, the S&P 500 returned −18.9%.
February 2000: a 17.5% two-month rise. The following 24 months returned −19.0%.
May 2007: a 20.3% two-month rise. The following 24 months returned −39.9%.
June 2007: a 19.4% two-month rise. The following 24 months returned −38.8%.

Every one of those four historical spikes was followed by a negative 24-month return for the S&P 500 — in two cases, severely negative. That is a small sample (four data points from two market cycles), and a sharp rise in margin debt does not, on its own, cause a decline; it is more plausible that both the borrowing and the eventual decline share a common cause, which is that leverage tends to expand fastest exactly when confidence — and therefore risk — is already at its highest. Treating four observations as a reliable forecasting rule would overstate what this data can tell you.

If you read the dollar record as stretched: the case, and the case against

The case against reading too much into the dollar record: nominal margin debt has no reason to stay flat while the economy, corporate earnings, and the market's total value all grow over nearly three decades. A 2.3% margin-debt-to-market-cap ratio sitting at its historical average, rather than at a 2000-style or 2007-style extreme, is a real piece of evidence that this cycle's leverage is not obviously more stretched than normal. The case for taking the record seriously anyway: the market-cap ratio is itself backward-looking and partly mechanical — if valuations are elevated for reasons unrelated to leverage, the ratio can look "normal" even while the dollar amount that would need to be sold in a margin-call cascade is objectively larger than in any previous cycle. A separate, related figure worth flagging: investors' aggregate credit balance — the cash-and-credit side of brokerage accounts, as distinct from margin debt itself — fell to a record low of negative $1.06 trillion in the same mid-2026 reading, meaning that in aggregate, investors currently owe their brokers more in net terms than they hold in cash. That is a leverage signal in its own right, independent of the margin-debt dollar figure above.

If you hold margin now: sizing against the range that was actually recorded

None of this is a timing signal, and nothing here should be read as an instruction to buy, sell, or adjust a specific portfolio — this is market history, not individualized investment advice, and a licensed financial professional is the right source for decisions about your own account. What the historical record does support is a narrower, practical point: margin debt is inherently forced-selling risk, because a broker can issue a margin call and liquidate positions at the worst possible moment, regardless of whether you personally believe the broader market is at a historically normal or extreme leverage level. Investors who use margin at all are generally better served by sizing it to a level they could repay without forced selling in a 25%-57% drawdown — the range actually observed after the three historical peaks above — rather than by trying to time an exit based on this month's aggregate FINRA reading. Checking your own account's maintenance-margin percentage and current equity cushion against the worked example above is a more actionable exercise than guessing whether the market-wide total is "too high."

What would move you to a different row

Two things would move a reader from one row of that table to another, and neither is a market forecast. The first is a change in your own balance or your broker's maintenance requirement, which is visible in your account today and is the only input in the table you control. The second is the market-cap ratio leaving the 2.3% long-run average it currently sits on, which would show up in FINRA's monthly release and in the commentary that tracks it. Neither tells you what the next 24 months hold; both tell you which row you are reading from.

Sources and disclaimer

General investment education only, not personalized investment, tax, or legal advice. Figures are as of the dates noted above and sourced to FINRA's official margin statistics; consult a licensed financial professional about your own account.

Start Here: every analysis by topic · About · Privacy Policy

Search This Blog

US Stock Insights