# Global Wealth Flow Monitor — 19 October 2026

Foreign demand for long-term U.S. securities recovered sharply in August, but short-term dollar positions moved in the opposite direction. The U.S. Treasury recorded **$279.2 billion in adjusted net long-term securities inflows**, after **-$27.9 billion in July**. Foreign investors bought **$300.7 billion of U.S. long-term securities**, while U.S. residents purchased **$21.6 billion of foreign securities**. Foreign Treasury-bill holdings fell **$47.8 billion** and banks' own net dollar liabilities to foreign residents fell **$41.6 billion**. The result is a change in maturity composition, not a uniform movement of global wealth.

Newer fund data show a similar division. In the week ended 7 October, U.S. long-term funds and ETFs received **$12.35 billion combined**, with **$24.53 billion into bonds** but **$18.36 billion leaving domestic-equity funds**. U.S. money-market fund assets subsequently declined **$40.46 billion** in the week ended 14 October, remaining at **$7.923 trillion**. The populations and reporting periods differ: these figures cannot be added to the Treasury balance.

**Information cutoff: 19 October 2026, 00:20 Brasília time.** Treasury TIC refers to August, released 16 October; ICI long-term funds to 7 October, released 14 October; money-market funds to 14 October, released 15 October.

## Long-term inflows recovered while short-term dollar liabilities fell

The Treasury's 16 October release shows a **$307.1 billion** swing in adjusted net long-term securities transactions from July to August. Foreign purchases of U.S. long-term securities included **$257.8 billion by private investors** and **$42.9 billion by foreign official institutions**. Those two components are already part of the $300.7 billion foreign purchase total. The adjusted $279.2 billion balance also accounts for U.S. purchases abroad and technical adjustments; it is not an additional pool of money.

| Treasury TIC measure | July | August | Interpretation |
|---|---:|---:|---|
| Adjusted net long-term transactions | -$27.9bn | +$279.2bn | Cross-border transaction flow |
| Foreign net purchases of U.S. long-term securities | +$40.6bn | +$300.7bn | Foreign portfolio demand |
| U.S. purchases of foreign securities | +$68.5bn | +$21.6bn | Outward acquisitions, positive purchase convention |
| Foreign change in Treasury-bill holdings | +$39.0bn | -$47.8bn | Short-term security position |
| Bank net dollar liabilities to foreign residents | +$50.1bn | -$41.6bn | Banking liability, not securities flow |
| Total net TIC inflow | +$87.4bn | +$199.8bn | Includes long/short securities and banking |

```chart
type: line
title: Adjusted U.S. net long-term securities transactions
unit: US$ billion; monthly flows, not market-value changes
2026-05 | 231.2
2026-06 | 175.5
2026-07 | -27.9
2026-08 | 279.2
```

The series is volatile, not a monotonic recovery. The twelve-month net long-term total through August was **$1.425 trillion**, versus **$1.221 trillion** in the previous twelve months. Both are flows, not outstanding holdings. Foreign private purchases of Treasury notes and bonds reached **$90.9 billion** in August and official purchases **$18.1 billion**; these are subsets of the long-term total. Treasury warns that custodial reporting cannot reliably identify ultimate beneficial owners by country. TIC also excludes foreign direct investment, so it does not establish newly commissioned U.S. production capacity. [U.S. Treasury, 16 October](https://home.treasury.gov/news/press-releases/sb0662).

## Bonds and world-equity funds attracted subscriptions despite domestic-equity redemptions

ICI's 14 October release revised the combined flow for the week ended 30 September to **-$2.299 billion**, compared with **-$1.969 billion** in the earlier vintage. The week ended 7 October then registered **+$12.351 billion**. Traditional long-term mutual funds lost **$35.87 billion** while ETFs issued **$48.22 billion** net. These are different investment wrappers; the figures do not prove that the same investors switched directly between them.

```chart
type: line
title: Combined U.S. long-term mutual-fund and ETF flows
unit: US$ billion; ICI revised weekly estimates
2026-09-09 | 1.629
2026-09-16 | -9.453
2026-09-23 | 20.037
2026-09-30 | -2.299
2026-10-07 | 12.351
```

Domestic-equity funds lost **$18.355 billion**, world-equity funds gained **$6.949 billion**, and bond funds gained **$24.527 billion**. Within bonds, taxable funds received **$25.433 billion** and municipal funds lost **$0.906 billion**. The world-equity category does not identify the final countries receiving investment. Nor does the bond total reveal duration, credit quality or the identity of the investors selling domestic equities. The observed change is selective fund demand, not a complete map of global savings. [ICI, 14 October](https://www.ici.org/research/stats/combined_flows).

Money-market fund assets fell from **$7,963.55 billion** on 7 October to **$7,923.08 billion** on 14 October. Institutional funds accounted for **$39.38 billion** of the decline. The earlier week's sharp increase followed by a partial reversal undermines any interpretation based on a single weekly change. Money-market assets are a stock, mainly influenced by net subscriptions, and cannot be treated as the world's available cash. [ICI, 15 October](https://www.ici.org/research/stats/mmf).

## Client allocations through major managers are not proprietary capital

BlackRock reported **$205 billion of net client inflows in Q3**, **$869 billion over twelve months**, and **$15.5 trillion of assets under management** on 30 September. The quarterly flow shows client demand for the manager's products; AUM is a stock that also reflects market prices. Neither establishes beneficial ownership of those assets by BlackRock. The firm identified demand across ETFs, active equity, private markets and cash. Its client flows can overlap with ICI fund statistics and must not be added to them. [BlackRock, 14 October](https://www.blackrock.com/corporate/newsroom/media/press-releases/blackrock-reports-third-quarter-2026).

The IMF's October financial-stability assessment identifies concentrated AI exposures, leverage, sovereign-yield shocks and operational dependencies as possible channels through which asset repricing can amplify financing stress. These are vulnerabilities, not evidence of a crisis this week. Published house views are distinct from observed client flows or proprietary positions. [IMF, October GFSR](https://www.imf.org/en/Publications/GFSR/Issues/2026/10/13/global-financial-stability-report-october-2026).

## Physical gold holdings rose while their dollar valuation fell

September remains the latest full-month global gold-ETF observation. The World Gold Council reports **$10 billion of inflows**, **67 tonnes added** and **4,256 tonnes held**, while ETF assets under management declined **7% month on month to $574 billion** as gold prices fell. Q3 inflows of **$31 billion** include September; the periods are not additive. Rising quantities and falling asset values can coexist. Gold ETFs are private portfolios, not monetary-authority reserves; the IMF's COFER covers official foreign-exchange assets and excludes gold. [WGC, 7 October](https://www.gold.org/goldhub/data/global-gold-backed-etf-holdings-and-flows).

```flow
Foreign long-term securities purchases rise → August TIC inflow rebounds → long-term funding demand improves; ultimate owners remain uncertain
Foreign Treasury-bill holdings and bank dollar liabilities fall → short-term dollar exposure contracts → maturity composition diverges
Domestic-equity fund withdrawals + world-equity and bond subscriptions → selective fund demand → country destinations remain unidentified
Gold ETF subscriptions + falling gold price → tonnes held rise while AUM falls → quantity differs from valuation
```

```map
title: Geographic scope and attribution limits
United States | TIC reports August foreign long-term securities purchases | Custodian residence does not establish beneficial ownership
World-equity funds | Positive subscriptions contrast with domestic-equity withdrawals | Country destinations are not identified
Europe and North America | Both contributed to September gold ETF inflows | Listing location is not investor residence
Global reserve managers | COFER measures official FX composition | Different population and period from private flows
```

## The next releases will test whether the change persists

| Verification | Release or period | Test |
|---|---|---|
| ICI long-term and money-market funds | October weeks | Persistence of bond/world-equity demand |
| Treasury TIC | September, scheduled 18 November | Whether August's inflow continues |
| BIS banking statistics | Next release scheduled 30 October | Cross-border bank credit versus securities flows |
| IMF COFER | Next quarterly update | Currency shares after valuation effects |
| WGC gold ETF data | October | Subscriptions, physical holdings and AUM separately |
| Corporate and public investment disclosures | Ongoing | Actual capital investment, commissioned facilities and ownership |

A lasting redistribution of productive capacity would require verified new investment projects spending, operational factories, infrastructure utilisation, asset ownership or enforceable control rights. Announced data-centre investment and sovereign-fund commitments are not commissioned capacity. The central alternative to the present interpretation is a renewed sovereign-yield shock that causes simultaneous fund redemptions, tighter credit and weaker primary issuance. Conversely, persistent foreign long-term purchases and orderly credit markets would strengthen the selective-demand interpretation. These are testable conditions, not investment recommendations.


## The August portfolio split matters more than the headline inflow

The TIC release permits a more precise reading of what foreigners acquired. In August, net **private** purchases of U.S. Treasury notes and bonds were $90.9 billion, private purchases of corporate bonds $36.8 billion, and private purchases of U.S. equities $131.9 billion. Private agency-bond transactions were negative $1.7 billion. These securities categories belong to the private-investor subtotal, not to four independent additions to the headline adjusted inflow. Official transactions followed another pattern: foreign official institutions bought $18.1 billion of Treasuries and $22.0 billion of equities, alongside $4.8 billion in corporate bonds and negative $2.0 billion in agency bonds. The divergence is important because reserve managers, private funds and intermediaries have different mandates, currency needs and liquidity horizons.

```chart
type: bar
title: August 2026 private foreign net purchases of U.S. securities
unit: US$ billion; Treasury TIC, net transaction flow
Treasury notes/bonds | 90.9
Agency bonds | -1.7
Corporate bonds | 36.8
Equities | 131.9
```

Equities were the largest of these four private categories in August, but a single month's composition cannot prove that surplus economies are permanently reallocating their savings from government bonds into technology shares. Equity purchases may reflect broad-market exposure, custody chains, tactical reallocations or other sectors. Neither transaction data nor a fund-manager earnings release identifies the underlying beneficial owner or the ultimate economic purpose of each dollar. The official/private distinction also matters for political interpretations: an official institution can alter reserve and portfolio allocations without changing the currency denomination of every underlying financial claim. A country-level claim requires country and instrument decomposition rather than extrapolation from the aggregate. The source is the [Treasury's August TIC table](https://home.treasury.gov/news/press-releases/sb0662/).

## A large change in asset values is not the same as a flow of savings

The U.S. Bureau of Economic Analysis reported a net international investment position of **negative $22.42 trillion** at the end of the second quarter of 2026, compared with negative $21.27 trillion at the end of the first. U.S. external financial assets stood at **$46.97 trillion** and liabilities to nonresidents at **$69.39 trillion**. The underlying quarter matters: these are balance-sheet positions as of June, not October fund transactions or August TIC purchases. A deterioration in the net position cannot simply be relabeled as a new overseas financing flow.

The BEA attributed the second-quarter **$3.72 trillion increase in U.S. external assets** primarily to about **$3.03 trillion of price changes** and **$663.3 billion of financial transactions**. External liabilities rose **$4.87 trillion**, driven mainly by roughly **$3.95 trillion in prices** and **$978.9 billion in transactions**. The broad aggregates include other changes and rounding; they are not a precise two-term identity. In the same release, the current-account deficit widened to **$246.0 billion**, from a revised **$212.6 billion** in the first quarter. That macroeconomic balance is related to saving and investment but should not be treated as numerically identical to monthly TIC, weekly fund flows or the change in net external wealth. Differences in coverage, accounting, valuation and timing are substantial. [BEA, 24 September 2026](https://www.bea.gov/index.php/news/2026/us-international-transactions-and-investment-position-2nd-quarter-2026).

These distinctions govern any claim that the United States is attracting productive resources. Purchasing an existing listed share can finance a seller rather than a new factory. A bond issued to fund construction may support future capacity, but only if proceeds are deployed and the project reaches operation. Price appreciation increases the current market value of the holder's claim without itself moving a corresponding sum across a border. A merger can transfer control without equivalent new investment projects investment. A complete map of economic power therefore requires matching securities transactions with direct-investment statistics, ownership rights, industrial activity and capital formation.

## ETF issuance does not reveal the final location of capital

The weekly ICI table can be decomposed more fully. For the week ended 7 October, ETFs recorded an estimated **$21.175 billion** in equity net issuance, with **$12.542 billion** in domestic-equity products and **$8.633 billion** in world-equity products. ETF bond products took in **$25.412 billion**, including **$20.783 billion** in taxable bonds and **$4.629 billion** in municipals. Total equity-and-bond flows are not the same as all-wrapper flows because hybrid and other categories also contribute. These are fund-share creation estimates, not trades by every financial investor or verified portfolio commitments by the ETF sponsor.

The combined domestic-equity category nevertheless showed a withdrawal of $18.355 billion. That coexistence means traditional mutual-fund redemptions outweighed positive domestic-equity ETF issuance in the combined population; it does not identify an identical person selling one vehicle to buy the other. Similarly, a positive world-equity fund category can hold both developed- and emerging-market assets, U.S.-listed international companies or changing country weights. The institutional accounting unit is the U.S. fund product, not the capital account of each country receiving exposure. The [ICI separate ETF release](https://www.iciglobal.org/research/stats/etf_flows) and the [combined release](https://www.iciglobal.org/research/stats/combined_flows) are useful together precisely because one decomposes the fund wrapper and the other combines wrappers; adding their totals would double count.

In a geographically integrated portfolio, an investor may buy a U.S.-domiciled ETF that owns non-U.S. securities. The cash movement through the fund and the location of the operating company's productive assets can be different from the fund's legal domicile and the brokerage intermediary. This distinction becomes decisive when assessing whether emerging-market companies gained access to financing, whether foreign exchange was converted, and whether future earnings claims changed hands. Current weekly fund data alone cannot answer those questions.

## Liquidity transmission runs through institutions, not a single global cash pool

The ICI's **$7.923 trillion** in U.S. money-market fund assets on 14 October is a stock of fund shares. It is not a measure of aggregate uninvested capital that could instantaneously enter equities, and it is not directly equivalent to bank deposits, central-bank reserves or Treasury-bill supply. Fund subscriptions and redemptions are also sensitive to institutional cash management, settlement conventions, yields and the calendar. A $40.46 billion weekly decline may represent an allocation change without implying a collapse in the underlying supply of market financing. The comparison with TIC is useful for competing liquidity hypotheses, but there is no accounting bridge that makes the numbers mechanically additive.

The October IMF Global Financial Stability Report offers a transmission framework rather than proof of current distress. It highlights vulnerability to an AI-related correction if future investment and profitability disappoint, to a jump in sovereign yields amid fiscal and inflation concerns, and to cyber disruptions in increasingly connected financial infrastructure. It also points to leverage and liquidity mismatches in parts of nonbank finance, including contingent liquidity needs. These factors may cause asset-price declines to interact with margin demands, fund outflows and the sale of more liquid holdings. Such a mechanism becomes relevant if the tightening is measured through spreads, failed issuance, financing costs or redemption terms; a risk description alone is not evidence of contagion.

That logic helps interpret the coexistence of record asset-management stocks and substantial client inflows. A large manager's reported AUM can rise from market appreciation even while some clients withdraw; equally, positive client flows need not signal that the firm has taken a proprietary long position. The BlackRock figures are evidence about demand for its services and products, not a census of all institutional exposures or a forecast of future profits on those positions. The correct comparison is between disclosed portfolio decisions, net flows, and independent financing or price data, with each period labeled.

## Gold, reserve currencies and productive resources answer different ownership questions

The World Gold Council's September observation illustrates a fundamental decomposition. **67 additional tonnes** measure physical quantity held by covered funds; **$10 billion of inflows** is an estimated transaction measure; **$574 billion of AUM** is a dollar-valued stock exposed to the metal's price. A decline in AUM despite additional tonnes does not imply that September's subscriptions were fictitious. Conversely, neither higher tonnes nor a higher dollar stock would prove a government increased its bullion reserves. These funds represent a specific private investment channel, while national monetary authorities report reserve assets under different accounting rules.

The geographical label of an ETF listing is also incomplete as a beneficiary map. A fund listed in London can have international shareholders and custody chains, and bullion held in a particular vault does not necessarily identify the residence of its ultimate beneficial owner. The Council reports Europe and North America as leading sources of September subscriptions at the fund-listing level. For claims about national wealth or strategic autonomy, one would also need official reserve reports, private custodial ownership information where available, trade and physical processing data. None of those observations is supplied by this weekly ETF series alone.

Foreign-exchange reserves tracked by the IMF's COFER are a different unit of analysis again: currency composition of reported official FX claims. Gold is not part of those FX holdings, and changes in reported currency shares can reflect exchange-rate valuation as well as active purchases and sales. A rising dollar weight therefore would not, by itself, establish that monetary authorities executed a proportionate net purchase of dollar securities in the same quarter. Assessments of monetary diversification should combine official disclosures, FX-adjusted reserve figures when available, securities flows and central-bank balance-sheet methods.

## From financial transactions to control: an evidence map

| Observation and level | What is established | What remains unproved |
|---|---|---|
| August Treasury TIC, securities | Foreign investors purchased U.S. long-term claims; private and official segments can be separated | Ultimate owner, destination of issuer proceeds, enduring strategy |
| October ICI funds, portfolio wrapper | Net subscription or redemption by reported investment category | Direct ownership shift by country, identical investors switching products |
| June BEA international investment position | Marked-to-market cross-border asset and liability stocks | A cross-border payment equal to the stock change |
| September WGC gold-backed funds | Additional covered gold tonnage and subscriptions | Official monetary gold purchase or change of sovereign control |
| BlackRock Q3 earnings | Net client subscriptions and reported managed assets | Manager-owned portfolio, market-wide asset-allocation consensus |
| Corporate capital expenditure | A company's reported spending when independently verified | New operating capacity before commissioning and utilization |

A useful origin-to-destination matrix requires five separate entries: the economic owner before the transaction; the intermediary and legal jurisdiction; the issuer or receiving entity; the form of claim or control right; and the period, currency and evidentiary confidence. Treasury custody data inform some of these cells but not all. International fund statistics add a portfolio-product layer; project data add actual equipment and infrastructure. Missing links should remain explicit rather than being completed from national stereotypes or assumed cross-border routes.

Two economically similar flows can also have different second-order consequences. Foreign acquisition of listed bank equity changes a claim on profits and governance only to the extent of voting rights; the same sum used to subscribe newly issued capital could strengthen the bank's regulatory capital and lending capacity. Purchases of existing sovereign bonds change the holder and potentially the price of state financing, while participation in a new auction can more directly affect debt issuance. Sovereign funds can be investors, development sponsors or controlling owners. The publication therefore separates the direction of security demand from the slower and harder question of productive control.

## Three testable paths for the next month

The **persistence scenario** requires another positive TIC long-term month accompanied by sustained bond and world-equity fund subscriptions, without material deterioration in spreads or primary issuance. That would strengthen the interpretation of selectively durable demand for financial claims. It would not yet establish new productive capacity: evidence of commissioned assets, utilization and enforceable ownership would still be needed. Even a second strong month may reverse as custodial and institutional cash management changes.

The **reversal scenario** begins with a sovereign-yield shock, deteriorating financing conditions or a large correction in concentrated equity sectors. If leveraged intermediaries must post collateral and funds face redemptions, they may liquidate their most marketable assets first. Under that mechanism, previously strong securities inflows can coexist with declining prices and volatile money-market fund assets. Observable markers would include widening credit spreads, strained auction demand, unusually large synchronized redemptions and reduced corporate issuance. The IMF describes this class of vulnerabilities, but the present source set does not show that all these events are occurring.

The **geographic reallocation scenario** would require independently measured fund-country destinations, cross-border FDI and new operational capacity outside the established financial centers, alongside persistent changes in reserve or institutional portfolios. A country with rising inward securities flows might still have little new factory investment, and an investment destination with strong industrial orders might not register a comparable U.S.-reported mutual-fund flow. This is why the next period must be evaluated through distinct datasets rather than a single headline number. The report's principal conclusion remains narrower and observable: long-duration U.S. securities demand rebounded in August, while shorter-term and fund categories did not move uniformly.

## Sources

1. [U.S. Treasury, TIC August 2026](https://home.treasury.gov/news/press-releases/sb0662), 16 October 2026.
2. [ICI, Combined Long-Term Flows and ETF Issuance](https://www.ici.org/research/stats/combined_flows), 14 October 2026.
3. [ICI, Money Market Fund Assets](https://www.ici.org/research/stats/mmf), 15 October 2026.
4. [BlackRock, Q3 2026 results](https://www.blackrock.com/corporate/newsroom/media/press-releases/blackrock-reports-third-quarter-2026), 14 October 2026.
5. [IMF, Global Financial Stability Report](https://www.imf.org/en/Publications/GFSR/Issues/2026/10/13/global-financial-stability-report-october-2026), 13 October 2026.
6. [World Gold Council, September 2026 gold ETF flows](https://www.gold.org/goldhub/data/global-gold-backed-etf-holdings-and-flows), 7 October 2026.
7. [BIS international banking statistics](https://data.bis.org/topics/LBS), next release scheduled 30 October 2026.
8. [BEA, Q2 2026 international accounts](https://www.bea.gov/index.php/news/2026/us-international-transactions-and-investment-position-2nd-quarter-2026), 24 September 2026.
9. [ICI, ETF net issuance](https://www.iciglobal.org/research/stats/etf_flows), 13 October 2026.
10. [IMF COFER](https://data.imf.org/en/datasets/IMF.STA%3ACOFER), Q2 2026.

**Data cutoff:** 2026-10-19 00:20 BRT.