Record ETF Inflows: Nearly $1.6 Trillion Poured Into U.S.-Listed Funds Through September

When you wonder why the markets don't sell off, I think you have to take into consideration that the public continues to pump money into exchange-traded funds that let the investor buy a bundle of investments in one trade and make diversification simpler. Nasdaq estimated in July 2026 that ETFs traded more than $234 billion a day, compared with roughly $589 billion in individual company stocks, about 28% of their combined dollar turnover.  

That's an overall US market estimate across trading venues, not a separate NYSE vs. Nasdaq figure.

Investors have kept pouring money into U.S.-listed ETFs this year, adding about $1.57 trillion from January through September, according to FactSet’s monthly reports. Flows were strongest in May, at $199.4 billion, and slowed to $140.9 billion in September. Source: FactSet monthly reports. Figures are rounded. 

Where the money went

Equity ETFs attracted the largest share in most months, though buying shifted as the year progressed. Equity inflows reached $110 billion in January and $114 billion in February, then eased to $70.1 billion in March as markets sold off. Buying picked up again in April and May, when equity funds brought in $133 billion and $130.2 billion, respectively. By August and September, equities accounted for a smaller share of monthly flows as investors added more money to fixed-income and other funds.

Bond ETFs also drew substantial demand. Fixed-income funds took in $53 billion in January and a record $70 billion in February. After slowing in April, bond inflows rose to $60.3 billion in May. In August, fixed income accounted for 33.5% of monthly flows; in September, that share increased to 41.4%. Investors favored short-term Treasury funds in September while also making selective moves into bonds with other maturities.

Investors also moved among sectors and themes. Technology and semiconductor funds attracted interest during parts of the spring and summer, alongside themes such as robotics and AI. But the flow picture wasn't one-way: August saw withdrawals from U.S. technology and semiconductor funds, even as some global technology and robotics funds attracted money. By September, global stocks outside the U.S. drew more equity ETF inflows than U.S. large-cap funds, while ultra-short government bond funds led fixed-income demand.

The pace of inflows has slowed since the spring peak, and the mix has broadened beyond stocks. FactSet's monthly figures add up to approximately $1.57 trillion, while ETFGI reported a separate estimate of $1.56 trillion through September. It's best to keep each provider's figures separate when comparing totals.

This year is beating last year's ETF inflows, but a lot has changed over the last year. I am not going to list all the changes because we all know what they are, but cumulative inflows remain on a record pace even though monthly flows have eased from May’s peak. If the January-through-September average monthly pace continues through October, November, and December, ETFs could attract another $523.9 billion during the fourth quarter, pushing total inflows to approximately $2.10 trillion by year-end 2026.

ETF Liquidity Concerns: What Happens If the Market Crashes?

With approximately $1.57 trillion flowing into U.S.-listed ETFs in the first nine months of 2026, an important question is: What happens to ETF liquidity if the stock market crashes?

The biggest concern isn't necessarily whether ETFs will continue trading, but whether there will be enough liquidity in the underlying securities when investors rush to sell at the same time.

Key liquidity risks include:

• Wider bid/ask spreads as market makers pull back.

• ETF discounts to net asset value (NAV) during heavy selling.

• Forced liquidations and cascading sell orders that can accelerate declines.

• Bond, small-cap, emerging-market, and leveraged ETFs facing greater liquidity stress.

During the March 2020 crash, some bond ETFs traded at significant discounts to reported NAV as underlying bond markets became difficult to price.

Bottom line: Large ETFs like SPY and QQQ generally remain liquid during crashes, but liquidity can deteriorate rapidly elsewhere. The real danger is a rush for the exits when the underlying markets cannot absorb the selling without significant price declines.

INTEL (INTC)

By Tom Incorvia

Points for the Intel chart:

1. Six-month balance. Since late April, INTC has rotated in a wide range from roughly 85 to 140. The market has accepted value there for half a year.

2. Short-term balance at the bottom. Price has built a tight balance around 91 to 95, right on the lower edge of that range. This is the third time the low end has been tested (early August, late August into September, now).

3. Below is thin air. The April rally from the 50s to the 80s was fast with little time spent at any price. If 85 gives way, there is no meaningful structure until the 41 to 51 area, the high-volume balance from January through March.

4. If support holds, the logic of balance says a responsive move back through the range, with a possible test of the opposite extreme in the 130 to 140 zone. The September balance near 116 to 126 is the first stop along the way.

5. Catalyst. Earnings are marked for later this month. A short-term balance sitting at a range extreme right into a report sets up the decision: acceptance below or rejection back into value.

One-line takeaway: Intel is coiled at the bottom of six months of value, and the earnings report likely decides whether it breaks to the 40s or rotates back toward the top.

With 35+ years of trading experience, Tom Incorvia brings a practical, real-world perspective to market analysis. Trained in Market Profile under Peter Steidlmayer, his approach centers on reading market structure through volume and price.

Explore Tom’s course: Volume Profile: Dynamics of Fair Value.

Wednesday’s ES session opened at 7785.00 and spent Globex inside a 36-point range, with a high of 7793.00 and a low of 7757.00. By the cash open, price was back at 7779.25. The overnight trade had held above Tuesday’s 7755.25 settlement, but the opening bell still brought another test lower.

That test reached 7751.00 during the 10:10–10:15 a.m. ET interval. Buyers then rebuilt the advance, carrying the ES to 7812.25 during the 2:15–2:20 p.m. interval. The cash-session range was 61.25 points. It was a two-way session with a solid recovery from the morning low, followed by a pullback from the afternoon high.

The ES was at 7804.00 at 3:50 p.m. and finished the cash session at 7799.00. Wednesday’s MOC flipped from a small sell Reveal to a $2.3 billion buy imbalance at 3:55, then narrowed to $0.2 billion to buy as outstanding orders on both sides declined into the close. The 5 p.m. last price and official settlement were both 7800.25, up 45.00 points, or 0.58%, from Tuesday’s settlement.

NQ settled at 30007.00, gaining 331.75 points, or 1.12%. ES volume totaled 1.62 million. Both contracts finished higher, although the late pullback showed that the afternoon highs still attracted sellers.

Outstanding buying rises, then both sides pair down

The Reveal for the MOC on October 21 was $0.2 billion to sell. The dashboard showed $4.1 billion of outstanding buys against $4.3 billion of sells. Beneath that balance, NYSE leaned $1.3 billion to buy while Nasdaq leaned $1.5 billion to sell. The S&P 500 was close to balanced at $0.1 billion to buy. Dollar direction and symbol participation were telling different stories.

The outstanding sell balance widened to $0.4 billion at 3:52 before switching to $0.4 billion to buy at 3:54. Between 3:54 and 3:55, outstanding buys increased from $4.3 billion to $5.2 billion while outstanding sells declined from $3.9 billion to $2.8 billion. That combination lifted the net buy balance to $2.3 billion. The increase in unmatched buy dollars deserves attention; it does not identify the traders or the gross volume of new orders.

From 3:55 to 4:00, both outstanding sides contracted: buys fell to $1.1 billion and sells to $0.9 billion, leaving $0.2 billion to buy. Such contraction is compatible with normal order pairing into the close. It is not, by itself, evidence of a bearish price move or weakening demand.

The accompanying Reveal snapshot showed GOOGL and NVDA leading buyers, while META led sellers. Financials carried $0.6 billion to buy and information technology $0.7 billion to sell. By the close, 50.2% of symbols favored buying. Historical ranking is not yet established: the dashboard and historical feed require reconciliation before valid same-clock size or day-to-day comparisons can be published.

See how the Market Imbalance Meter tracks buy and sell pressure into the close.

How the NYSE and Nasdaq auctions work
How to read the MIM

Thursday, October 22: Tom Incorvia joins The Pit for Market Profile at 9:15 a.m. ET.

Watch Tom Incorvia’s Market Profile show today at 9:15 a.m. ET.

Don't trade the open alone. Join us live in The Pit for free, follow the conversation, and bring your questions.

Economic Calendar

Earnings Calendar

Polaris Trading Group

Wednesday’s Polaris Trading Group discussion moved through an upside inside gap, a morning downside rotation, a recovery above the central decision area, and an afternoon reversal. David kept the focus on acceptance around 7780 ±5 rather than committing to one direction for the entire session. Trade updates were room reports and educational examples, with both successful target calls and a stopped long in the record.

Opening conditions and the gap

• David described the opening as an upside inside gap and said this type of gap tends to fill at a high rate. That was the morning framework, not a guarantee that every opening push should be faded.

• The first A10 long update reported the initial target filled and the stop adjusted. A subsequent message reported three targets filled, establishing that the opening sequence included a long opportunity before the downside script became the focus.

The bear script below 7780

• The published bearish condition was acceptance below 7780 ±5, with objectives at 7765, 7755, and 7745. Keeping the acceptance requirement attached to those numbers matters: the roadmap was conditional on market behavior around the line in the sand.

• David then reported that 7780 had held precisely and that the 7755 objective was fulfilled. The record supports that target update; it does not establish that the deeper 7745 objective traded or that every participant captured the move.

Execution and working runners

• A discount long was explicitly reported stopped out. Around the same period, David reported three targets filled on an A10 short, with a runner still working. The stopped trade belongs beside the favorable updates when reviewing the day.

• Discussion emphasized the contribution runners can make and questioned excessive micromanagement. Those comments reflect the room’s execution discussion, not a published comparison of strategy returns or evidence that one stop-management choice suits every account.

Pullbacks across time frames

• Participants asked how to identify a pullback and how to handle price approaching a higher-time-frame moving-average resistance while finding lower-time-frame support. The questions highlighted the difficulty of reading conflicting references during a developing move.

• A follow-up asked whether moving averages across time frames must compress before a trend develops. These were learning questions in the written record; the transcript does not supply a complete written rule or a verified answer that can be presented as a new trading system.

The bullish recovery roadmap

• David’s bullish condition was acceptance above the same 7780 ±5 area, with objectives at 7795, 7805 and 7815. The first objective was reported fulfilled late in the morning, and 7805 was reported fulfilled around lunchtime.

• Later commentary described price holding above the prior high as bullish and identified 7815 as the next objective. An afternoon message referenced a 7812 high, so the review should keep 7815 as the stated target rather than silently treating it as achieved.

Afternoon reversal and headlines

• Shortly after 3 p.m., David described an afternoon reversal after the 7812 high and said the D-Level reversal was playing out. He also posted an A4 short from the Auto Trade Assistant, accompanied by a chart showing the decline and subsequent rebound.

• A geopolitical headline entered the discussion, followed by a posted correction narrowing its scope to Saudi Arabia. The sequence illustrates the risk of reacting to an initial alert before its details settle; it does not independently establish why prices moved.

Research and preparation

• The room discussed testing across different market regimes, along with day-of-week and time-of-day variation. Participants also asked about collecting data for a future educational discussion. These were requests for stronger preparation, not completed research results.

• David’s early reminder was that an entry should follow fulfilled conditions. The later two-sided roadmap gave that principle a concrete reference point: one decision area, separate directional scripts, and objectives evaluated as the session unfolded.

Lessons from the session

• Wednesday’s record argues for separating a setup, a target update and a confirmed personal execution. Favorable room reports do not erase the stopped long, and an unfinished objective should remain unfinished in the recap.

• The closing message ended the day’s commentary. The useful takeaway is the process: define the condition, track the response, manage the position, and reassess when price or information changes.

Learn more about Polaris Trading Group

DTG Room Preview

The DTG group sees a defensive backdrop for Thursday, with energy prices, geopolitical headlines, and technology sentiment competing for control of the opening trade. Its preview favors watching how ES responds at the nearby trendlines rather than assuming Wednesday’s recovery will carry straight through.

Market backdrop and catalysts

• The group is watching crude oil, Treasury yields, and semiconductor participation as the main cross-market influences. Its overnight commentary describes large-trader flow as significant but mixed, without a clear directional bias.

• Initial jobless claims are due at 8:30 a.m. ET and the Conference Board Leading Economic Index at 10 a.m. ET. Earnings reactions and further geopolitical headlines could interrupt an otherwise technical opening setup.

ES levels and conditional paths

• Immediate trendline resistance is 7783–7786. A sustained break above that area would improve the group’s case for a move toward 7900 over the next few days.

• Additional resistance zones are 7888–7893, 7945–7950 and 8028–8033. The group’s support zones are 7702–7707 and 7410–7415.

• Without a recovery through nearby resistance, the group expects a possible sideways grind between roughly 7700 and 7800. A clear failure of the 7700 area would weaken that scenario and raise the risk of another decline.

• The group identifies the 50-day moving average near 7771.50 as a reference after price slipped back through it overnight. Treat these as the group’s dated chart levels, distinct from the BTS calculations above.