# Global Wealth Flow Monitor — 2 November 2026

The latest international banking release changes the interpretation of global capital movement more than a single week of securities-market prices. The Bank for International Settlements (BIS) reports US$40.7 trillion of cross-border bank credit at end-June 2026, after an exchange-rate- and break-adjusted quarterly expansion of US$877 billion. Euro-denominated credit accounted for US$424 billion of that increase, against US$383 billion denominated in dollars. These are credit positions and transactions, not a transfer of productive assets from one country to another. Meanwhile, US fund statistics published on 28 October show a return to positive combined long-term flows after an extraordinary institutional withdrawal in the preceding week. The contrast between banking balance sheets, investment wrappers and real investment is the central finding of this edition. [1][2]

**Information cutoff:** 2 November 2026, 05:25 America/Sao_Paulo. The most recent ICI combined fund observation ends 21 October; money-market assets end 28 October; BIS bank-credit observations end June; TIC international securities transactions cover August; COFER reserves and US external wealth positions cover Q2. No November market close, live price or unpublished monthly observation is assumed.

## The measurement problem: four distinct forms of change

An **asset stock** is a position at a stated valuation date. A **transaction flow** records new lending, repayment, purchases, sales or net fund subscriptions during a period. **Repricing and foreign exchange** change the translated value of a stock without necessarily moving an underlying security. A **change of control** requires evidence of voting rights, legal title, beneficial ownership, concessions, operating rights or effective corporate governance changing hands. These categories can interact but cannot be added as though they were the same quantity.

The comparison window must also match the statistical instrument. The ICI weekly series allows a one-week comparison of estimated fund flows, and its money-market series allows comparisons of weekly asset stocks. The BIS provides a quarterly change adjusted for exchange-rate movements and reporting breaks, plus a year-on-year growth rate. IMF reserve shares permit a Q1–Q2 comparison. August TIC cannot be relabeled an October flow. Neither the available releases nor the reporting definitions support a consistent one-month, three-month and one-year panel across all assets; missing cells are explicitly left unestimated.

| Evidence family | Latest observed period | Measurement | Defensible comparison | Main blind spot |
|---|---|---|---|---|
| BIS cross-border bank credit | End-June 2026 | US$40.7tn stock; +US$877bn Q2 adjusted credit | Quarter and +11% year on year | Final use and beneficial owner |
| ICI long-term funds and ETFs | Week to 21 Oct 2026 | +US$12.324bn estimated net flow | Prior weeks, same revised series | End-investor destination |
| ICI money-market funds | Week to 28 Oct 2026 | US$7,901.28bn assets | 21 and 14 Oct stocks | Cash account counterparties |
| US Treasury TIC | August 2026 | +US$199.8bn total net inflow | Monthly TIC when available | Custody versus final ownership |
| IMF COFER | 2026 Q2 | US$13.22tn official FX reserves | 2026 Q1 | Confidential country allocation |
| BEA external position | End-June 2026 | -US$22.42tn US net IIP | End-March 2026 | Valuation versus transactions |
| UNCTAD investment | H1 / first nine months 2026 | FDI flows / M&A deal value | Specifically stated 2025 bases | Projects announced versus built |
| Gold-backed ETFs | September 2026 | 4,256 tonnes; US$574bn AUM | August holdings and value | Official bullion reserves excluded |

The ledger makes an important asymmetry visible: the freshest financial observations are not the most comprehensive, and the broadest international accounts are not the most current. A coherent analysis must preserve both properties rather than select whichever number is largest.

## Bank credit: euro growth is material, but not a dollar exit

The BIS release of 29 October distinguishes **cross-border claims** from **cross-border credit**. Total adjusted claims rose US$1.3 trillion in Q2, whereas the narrower credit concept—loans plus banks' debt-security holdings—rose US$877 billion. Within credit, loans increased US$560 billion and holdings of debt securities US$317 billion. Counting the US$1.3 trillion claims increase on top of the US$877 billion credit increase would double-count overlapping balance-sheet categories. [1]

The currency composition is more revealing than a headline about de-dollarisation. Euro-denominated credit increased US$424 billion, dollar-denominated credit US$383 billion and other denominations accounted for the remaining approximately US$70 billion of the US$877 billion total. Both euro and dollar cross-border credit segments nevertheless grew 11% year on year. The euro's share of foreign-currency credit to nonbank borrowers has risen since 2022 even at constant exchange rates, according to the BIS global liquidity indicators. Euro international debt securities grew 13% year on year, compared with 6.7% for dollar-denominated equivalents. The evidence supports a relative shift in the **marginal currency mix**, not disappearance of dollar financing. [1]

```chart
type: bar
title: BIS adjusted cross-border bank credit expansion by currency, Q2 2026
unit: US$ billion; components of the same quarterly total
Euro | 424
US dollar | 383
Other currencies (residual) | 70
```

The residual is calculated as 877 minus 424 minus 383; it is not a published country allocation. Banks can lend in euros to borrowers outside the euro area while still holding dollar assets elsewhere. Borrowers may select euro financing because of interest-rate differentials, trade invoicing, natural hedges or investor demand; the aggregate BIS data cannot identify which motive dominates. A euro loan is also a liability for the borrower and a claim for the lender, not newly created national net wealth.

The emerging-market regional picture is uneven. BIS credit to emerging and developing economies expanded US$105 billion in Q2 and grew 8.9% year on year. Emerging Europe grew 19%, Africa and the Middle East 17%, Latin America and the Caribbean 9.3%, while emerging Asia contracted 2.9% year on year despite a positive US$25 billion quarterly movement. These are credit growth rates, not national GDP, capital formation or household welfare indicators. Currency mismatch, refinancing needs and the distribution of borrowers matter more than a regional ranking. [1]

## The fund-flow reversal: what the extraordinary October week did not prove

ICI's revised combined estimate for the week ended 14 October is **-US$471.677 billion**. For the following week, ending 21 October, it is **+US$12.324 billion**. Earlier comparable observations were +US$12.667 billion for 7 October and -US$2.301 billion for 30 September. The sharp reversal means that the negative reading cannot be projected mechanically into a persistent global liquidation. It does not prove that all prior redemptions were reversed by the same accounts. [2]

```chart
type: bar
title: ICI combined long-term mutual-fund flows and ETF issuance
unit: US$ billion; weekly net estimates, revised series
2026-09-30 | -2.301
2026-10-07 | 12.667
2026-10-14 | -471.677
2026-10-21 | 12.324
```

The ICI attributes the exceptional domestic-equity mutual-fund outflows of 14 October primarily to activity in **collective investment trusts**, not to a wave of individual-investor redemptions. This methodological statement is positive evidence against an indiscriminate retail-flight narrative, but it does not reveal the ultimate receiving account, the asset baskets sold or the continuity of beneficial exposure. A transfer between institutional wrappers can change reported fund flows without changing the ultimate economic owner; a genuine sale can also occur. Both mechanisms remain possible until custody and portfolio data discriminate between them. [2]

For 21 October, the combined +US$12.324 billion consists of estimated mutual-fund outflows of US$33.50 billion offset by US$45.82 billion of ETF net issuance. Equity funds contributed +US$13.892 billion, hybrid funds -US$2.088 billion, bond funds -US$0.166 billion and commodity products +US$0.686 billion. Domestic equity contributed +US$7.090 billion and world equity +US$6.802 billion. ETF issuance is already included in the combined total and must not be added again. [2]

| Combined ICI category | 14 Oct, US$bn | 21 Oct, US$bn | Interpretation |
|---|---:|---:|---|
| Domestic equity | -479.390 | 7.090 | Institutional wrapper anomaly, then positive estimate |
| World equity | 2.972 | 6.802 | Positive vehicle flow; geography of ultimate purchase unknown |
| Bond | 7.923 | -0.166 | Small negative after positive prior week |
| Commodity | -0.850 | 0.686 | Mixed ETF activity, not sovereign gold buying |
| Combined total | -471.677 | 12.324 | Revised net estimate, not portfolio performance |

Fund inflows are neither returns nor profits. The categories are based on prospectus objectives and reporting populations, not a full ledger of cross-border counterparties. The new observation is a normalization of an estimated vehicle-flow statistic, not proof that risk tolerance, asset valuations or corporate control recovered.

## Cash vehicles: the latest stock fell despite the previous week's increase

ICI reported US$7,901.28 billion in US money-market fund assets for 28 October, down US$43.29 billion from US$7,944.57 billion on 21 October. Government funds fell US$37.23 billion, prime funds US$3.07 billion and tax-exempt funds US$2.99 billion. Institutional share classes accounted for US$33.66 billion of the total decline, while retail classes fell US$9.63 billion. The series explicitly treats weekly asset changes as a proxy primarily driven by flows, subject to revisions and composition effects. [3]

```chart
type: line
title: US money-market fund assets at weekly reporting dates
unit: US$ billion; assets, not securities transactions
2026-10-14 | 7923.08
2026-10-21 | 7944.57
2026-10-28 | 7901.28
```

This stock series has a different observation week from the latest combined long-term fund flows. Therefore, it cannot identify where the 14 October collective-trust withdrawal landed. A decline in government money-market funds could reflect treasury cash management, taxes, institutional liquidity requirements or portfolio transfers. None of those motives is established by the aggregate. Likewise, the US$7.90 trillion stock is not a sovereign reserve, a bank deposit aggregate or uncommitted cash ready to buy factories.

The distinction matters for second-order effects. Money-market vehicles provide demand for Treasury bills, repurchase agreements and other short-duration instruments, subject to their mandates. If subscriptions weaken while Treasury issuance or dealer financing needs rise, short-term market rates could face pressure. But the evidence needed is matched changes in bill supply, repo pricing, holdings and fund portfolios—not merely one weekly fall in fund assets.

## International securities: August inflows and the valuation of external wealth

The US Treasury's release of 16 October reports **US$199.8 billion** of total net TIC inflows in August, including US$178.5 billion private and US$21.3 billion official inflows. Foreign residents purchased US$300.7 billion of long-term US securities, while US residents bought US$21.6 billion of foreign long-term securities. After the Treasury's specified adjustments, estimated net long-term securities transactions were US$279.2 billion. These are different TIC aggregates, not additive components of a global wealth balance. September TIC was scheduled for 18 November and is not available at this cutoff. [4]

A securities transaction changes a portfolio claim, but it need not confer operational control of the issuer. A foreign fund's purchase of a US Treasury creates a debt claim on the sovereign; it does not make the buyer an owner of American infrastructure. Custodial residence may also differ from ultimate beneficial ownership. The distinction becomes especially consequential when policy commentary treats Treasury purchases as proof that a particular government or family has acquired domestic productive assets.

The Bureau of Economic Analysis reports a US net international investment position of **-US$22.42 trillion** at end-June, against -US$21.27 trillion at end-March. US external assets were US$46.97 trillion and liabilities US$69.39 trillion. During Q2, asset values increased US$3.72 trillion, including US$3.03 trillion of price changes and US$663.3 billion of transactions. Liabilities increased US$4.87 trillion, including US$3.95 trillion of price changes and US$978.9 billion of transactions. Much of the quarter's position change therefore came from valuation rather than new cross-border purchases. [5]

A more negative net investment position can coexist with inflows and rising foreign-held US asset prices. It does not, by itself, quantify wealth transferred from American households to foreign governments. Balance-sheet accounting is indispensable, but it is not a substitute for identifying owners, valuation conventions and legal claims.

## Reserve currencies and gold: reallocation is not a uniform exit

The IMF's 30 September COFER release puts official foreign-exchange reserves at **US$13.22 trillion in Q2 2026**, up from US$13.10 trillion in Q1. The dollar share declined from a revised 57.18% to 56.70%, while the euro share rose from 20.04% to 20.60%. Euro claims increased by nearly US$100 billion, which the IMF attributes largely to active purchasing. The renminbi share rose from 1.98% to 2.11%, and the yen share fell from 5.43% to 4.96%, with active selling, bond-price changes and depreciation all relevant to the latter. [6]

```chart
type: bar
title: Selected shares of official foreign-exchange reserves
unit: percent of COFER global reserves; quarters shown separately
US dollar 2026Q1 | 57.18
US dollar 2026Q2 | 56.70
Euro 2026Q1 | 20.04
Euro 2026Q2 | 20.60
```

A falling dollar share is not equivalent to a dollar reserve sale of the same percentage of US$13.22 trillion: the denominator and other currency values also move. The IMF's revised Q1 series is used consistently here; an older release's unrevised Q1 estimate must not be silently substituted. COFER is a global currency aggregate, not a public country-by-country register of reserve managers. Nor does it include official gold in the foreign-exchange share denominator.

Private gold ETFs provide an instructive counterexample. The World Gold Council reported September inflows of approximately US$10 billion and an increase of 67 tonnes to 4,256 tonnes held. Yet aggregate ETF assets under management fell about 7% to US$574 billion because the gold price declined. Physical quantity, net subscriptions and marked-to-market value moved in different directions. The September observation does not establish October central-bank purchases, sovereign reserve policy or a physical bullion transfer to any named country. [7]

## Direct investment, mergers and the difference between a deal and a factory

UNCTAD's 25 October Global Investment Trends Monitor reports that FDI increased 37% in the first half of 2026 relative to the average half-year level of 2025, driven largely by developed economies and financial-centre fluctuations. Cross-border M&A value rose 57% to US$494 billion over the first nine months, while deal counts fell 23%. new investment projects project announcements fell 15%, and international project-finance deals fell 11%. These comparisons have explicitly different periods and denominators and should not be combined into a single growth rate. [8]

The divergence matters for control. A completed acquisition can transfer corporate voting power, even if no new plant is constructed. A new investment projects announcement can describe intended capacity without transferring an existing enterprise or creating operational output. A large acquisition can raise the monetary value of FDI-related transactions while the number of new productive projects falls. UNCTAD also notes that AI-related sectors represented more than 40% of announced US new investment projects investment value; this is **announced value**, not verified installed computing capacity or electricity supply. [8]

The World Bank Group's fiscal-2026 statement records US$112 billion of private capital mobilization, up from US$35 billion in FY2022, but mobilization is a development-finance metric and must not be added to UNCTAD global FDI as a separate nonoverlapping pool. Its DRIVE announcement in October describes a Singapore-linked regional infrastructure credit fund **targeting US$750 million**, with IFC investing **up to US$50 million** alongside Japanese partners. The fund target is not capital already raised; the IFC ceiling is not a verified disbursement; and neither is a commissioned power project. [9][10]

## Energy, minerals and chips: financing only becomes power after commissioning

Capital can move quickly across securities while electrical infrastructure takes years to connect. The IEA's Electricity 2026 grid assessment estimates more than 2,500 GW of renewable, storage and large-load projects stalled in global grid queues, with annual grid investment needing to rise approximately 50% from around US$400 billion by 2030. These are planning estimates and project queues, not completed capacity. The IEA also identifies permitting, transformers, skilled labour and grid component prices as constraints that financing alone cannot remove. [11]

The EIA's October Short-Term Energy Outlook recorded a September average Brent price of US$114 per barrel amid disruptions to Middle Eastern infrastructure and shipping. This is an observed September price and an agency analysis of the disruption, not a November spot quote. Energy-importing firms face higher working-capital needs and possible currency pressure when fuel costs rise; exporters may receive higher gross revenue without necessarily gaining fiscal room after subsidies, transport costs and contractual obligations. The transmission differs across sovereigns, utilities and households. [12]

The USGS Mineral Commodity Summaries 2026 provide 2025 production and reserves context for more than 90 minerals. A resource estimate is not a reserve available at today's prices, and a reserve is not refined output. Rare-earth mining, separation, magnet manufacturing, industrial electricity and export permissions are distinct control points. Claims of a geopolitical transfer require contracts, equity records, licences or verified production—not the gross in-ground value of ore. [13]

For semiconductors, an advanced memory-chip supply agreement or a data-centre capital expenditure plan can indicate demand and bargaining leverage. It does not establish delivered wafers, qualified yields, installed accelerators or commercially available compute. The critical chain is financing → permits and grid connection → manufacturing and logistics → commissioning → operating output → cash earnings. Its weakest physical link can determine the pace at which a financial commitment becomes productive capacity.

```flow
Capital subscription or credit approval → legal financing close → equipment and grid procurement → construction and commissioning → delivered output → operating cash flow
Higher sovereign yields → collateral repricing and margin needs → dealer and nonbank deleveraging → tighter project credit → delayed productive investment
```

## Geography: which routes are actually observed?

```map
title: Regional attribution and transmission, with observation limits
United States | August net TIC inflow US$199.8bn | Security transactions; ultimate beneficial owners not fully identified
Euro area and external borrowers | Q2 euro-denominated bank credit +US$424bn | Currency of loan does not identify lender nationality or productive destination
Emerging Europe / Africa and Middle East | BIS annual credit growth 19% / 17% | Regional bank credit, not sovereign net wealth
South and Southeast Asia | DRIVE infrastructure credit fund targets US$750m | Proposed fund size; not financed or commissioned capacity
```

| Origin / actor | Destination / exposure | Mechanism | Observed evidence | Limitation and possible second-order effect |
|---|---|---|---|---|
| BIS-reporting banks | Cross-border borrowers | Loans and debt-security holdings | +US$877bn Q2 adjusted credit [1] | No final project use; refinancing and currency mismatch may amplify stress |
| Nonresident investors | US securities | Portfolio purchases | +US$199.8bn August net TIC [4] | Custody obscures owners; demand can affect yields, not control of factories |
| Fund shareholders / ETF participants | US long-term fund wrappers | Redemptions and net issuance | +US$12.324bn week to 21 Oct [2] | No matched investor identities; market liquidity response uncertain |
| Reserve managers | Currency claims | Reserve purchases and repricing | Euro share 20.60% Q2 [6] | Confidential country detail; exchange-rate transmission depends on liabilities |
| Acquiring companies | Target companies | M&A and voting rights | US$494bn cross-border M&A value, nine months [8] | Deal completion and concentration require transaction-level verification |
| Development-finance institutions | Asian infrastructure borrowers | Credit fund / risk sharing | US$750m target, IFC up to US$50m [10] | No proven deployment; cost of energy may fall only after commissioning |

This matrix is not a closed global sources-and-uses account. The observations overlap in institutions, cover different periods and sometimes measure different sides of the same transaction. An international loan may fund a security purchase; a bank may hold the bond issued by a company; a fund may own shares in that bank. Summing every recorded position would multiply one underlying exposure.

## Competing explanations and falsifiable tests

**Scenario A — diversification at the margin, not replacement of the dollar.** Euro credit and euro reserve shares continue rising while dollar credit still expands. This is consistent with borrowers and reserve managers broadening currency choices under relative cost and risk considerations. The interpretation weakens if subsequent constant-FX BIS data reverse the euro-share increase or if reserve gains prove almost entirely valuation effects. Its second-order implication is greater sensitivity to euro financing conditions among borrowers outside the currency union, not necessarily a loss of US market depth.

**Scenario B — an institutional accounting shock rather than mass liquidation.** The ICI collective-trust effect is transitory and subsequent combined weekly flows remain closer to ordinary magnitudes. This is consistent with the latest +US$12.324 billion observation, but the same end-investor exposures have not been tracked. The hypothesis weakens if later revisions reveal persistent asset sales across unrelated fund categories, matched custody outflows and falling holdings. A genuine liquidation would affect market depth and collateral more directly than a wrapper migration.

**Scenario C — headline capital strength with weak project conversion.** UNCTAD's larger M&A values coexist with fewer deals and new investment projects announcements, while grid queues and physical constraints delay productive deployment. This explanation strengthens if project cancellations, connection delays and low disbursement rates persist despite large financing announcements. It weakens if completed investment, grid connections and measurable industrial output accelerate broadly, including outside high-income hubs. A narrow set of large transactions should not be mistaken for economy-wide productive renewal.

**Scenario D — financing stress spreads through collateral.** Higher sovereign yields, energy prices or a concentrated technology-valuation correction trigger margin calls and financing strains. The IMF's October stability assessment identifies sovereign bonds, AI valuations and cyber incidents as potential triggers, with leverage and liquidity mismatches as amplifiers. This is a conditional risk mechanism, not an observed November crisis. Its falsifier is sustained functioning of repo and sovereign markets with stable financing spreads and no forced-selling evidence despite the hypothesized shocks. [14]

| Next observable test | Required release or record | What it could resolve |
|---|---|---|
| Fund normalization | ICI revised weekly flows, ETFs and CIT explanation | Wrapper shift versus persistent liquidation |
| International portfolio continuity | September TIC, scheduled 18 November | August inflow persistence; not October CIT destination |
| Currency mix | Next BIS constant-FX banking and liquidity series | Euro-share persistence without translation bias |
| Reserve allocation | Next IMF COFER vintage | Active reallocations versus currency and bond-price changes |
| Real investment | UNCTAD completed deals, grid-connection and project-finance records | Announcements versus installed capacity and control |
| Liquidity stress | Repo, sovereign yields and financing terms | Whether leverage actually transmits price shocks |

## Assessment at the cutoff

The strongest new evidence is the BIS currency and regional composition of cross-border credit, combined with the ICI normalization after the exceptional collective-trust week. Both show changes in the allocation and intermediation of capital; neither establishes a global handover of productive ownership. The more persistent risk is a mismatch between financial commitments and the slow conversion of those commitments into electricity, equipment, industrial output and durable earnings. Reserve shares and fund flows can change within a quarter or a week; power stations, transmission corridors and manufacturing capacity cannot.

Confidence is **high** in the cited definitions and published aggregates, **medium** in the interpretation of relative currency diversification and institutional normalization, and **low** in any claim about unidentified beneficial owners, unobserved transfers or future project delivery. The evidence does not support a single global wealth-flow number. It supports a set of separate, dated and falsifiable accounts.

## Primary sources and release dates

[1] BIS, *International banking statistics and global liquidity indicators at end-June 2026*, 29 October 2026: https://www.bis.org/publications/202610-commentary-ibs-gli

[2] Investment Company Institute, *Combined Estimated Long-Term Flows and ETF Net Issuance*, 28 October 2026: https://www.ici.org/research/stats/combined_flows

[3] Investment Company Institute, *Money Market Fund Assets*, 29 October 2026: https://www.ici.org/research/stats/mmf

[4] US Treasury, *Treasury International Capital Data for August 2026*, 16 October 2026: https://home.treasury.gov/news/press-releases/sb0662/

[5] US Bureau of Economic Analysis, *US International Transactions and Investment Position, Q2 2026*, 24 September 2026: https://www.bea.gov/news/2026/us-international-transactions-and-investment-position-2nd-quarter-2026

[6] IMF, *COFER Data Brief, Q2 2026*, 30 September 2026: https://data.imf.org/news/imf%20data%20brief%20september%2030

[7] World Gold Council, *Gold ETF Flows: September 2026*, 7 October 2026: https://www.gold.org/goldhub/data/global-gold-backed-etf-holdings-and-flows

[8] UNCTAD, *Global Investment Trends Monitor No. 51*, 25 October 2026: https://unctad.org/publication/global-investment-trends-monitor-no-51

[9] World Bank Group, *Mobilizes Record Private Capital for Developing Countries*, 17 September 2026: https://www.worldbank.org/en/news/press-release/2026/09/17/world-bank-group-mobilizes-record-private-capital-for-developing-countries-driving

[10] World Bank Group, *World Bank Group and Japan Advance Partnership on Energy Resilience in Asia*, 15 October 2026: https://www.worldbank.org/en/news/press-release/2026/10/15/world-bank-group-and-japan-advance-partnership-on-energy-resilience-in-asia

[11] International Energy Agency, *Electricity 2026: Grids*, 2026: https://www.iea.org/reports/electricity-2026/grids

[12] US Energy Information Administration, *Short-Term Energy Outlook: Global oil markets*, 6 October 2026: https://www.eia.gov/outlooks/steo/marketreview/crude.php

[13] US Geological Survey, *Mineral Commodity Summaries 2026*, revised May 2026: https://www.usgs.gov/publications/mineral-commodity-summaries-2026

[14] IMF, *Global Financial Stability Report: Resilience, Risk, and Transformation*, October 2026: https://www.imf.org/en/publications/gfsr/issues/2026/10/13/global-financial-stability-report-october-2026