ETF Assets and Manager Revenue Move Apart
We counted all 317 players on the ETF map. Of the 146 that issue ETFs, none uses revenue as its headline figure; the 16 players that do all sit elsewhere on the chain, at index firms, market makers and brokerage apps.
ETF Assets and Manager Revenue Move Apart
about 4 min
The number ETF stories use most is assets under management. So when assets swing sharply, does company revenue move with them? This post follows that single question through the ETF & Asset Management Industry Map, which covers 317 players along the whole chain of managers, index firms, exchanges, brokerage apps and data vendors, with research as of October 15, 2026.
This time we did not pick standout cases. We counted the whole map by place on the chain, headline figure and status, and three things came out. Of the 146 issuers that make ETFs, 0 use revenue as their headline figure and 94 use AUM. The 16 players whose headline figure is revenue all sit elsewhere on the chain: index firms, market makers, brokerage apps and the like. And of the 25 players that disappeared from the map, 18 did not shut down but were acquired or had their names absorbed.
Assets swung hard, revenue didn't swing as much
- In China, assets shrank while large managers' revenue grew. Mainland ETF assets fell from RMB 6.00T (end-2025) to RMB 4.93T (2026-09), −17.9% (✔). The map says redemptions by state money such as Central Huijin drove this. 510300, once the largest mainland ETF, went from RMB 422.26B (2025-12) to RMB 94.87B (2026-06) (✔). Yet ChinaAMC's and E Fund's company-wide revenue rose +31–34% year on year in 1H 2026, and the map says the shock hit Huatai-PineBridge (AUM −28.9%). The reason it gives is the 0.15% fee. That reads as state money having generated little revenue for its size.
- In Japan, even the largest holder leaving barely dents fees. Japanese ETF assets are ¥136.2T (2026-09). The 83.5% for trust banks (mostly BOJ) and 2.6% for retail come from a 2025-07 TSE survey covering ¥93.8T, so the date and base differ (△). The BOJ began selling in 2026-01, but at ¥330B/yr at book value, which the map says would take over 100 years to finish. By the map's calculation, that reduces Nomura's fee revenue by only ~¥260M/yr (<0.8%).
- With thin fees, size doesn't convert easily into money. Korea's TIGER US S&P 500 charges 0.0068%, and US index equity ETFs carry an asset-weighted average fee of 0.14% (ICI ✔). The map estimates the annual fee pool for the whole $24.03T of ETFs worldwide at ~$35B–$45B/yr (includes assumptions △).
Count the whole map and issuer revenue is barely written down
- The 146 issuers are measured by assets. Of the 146 large, mid-size, thematic, active, leveraged, crypto and AI managers, 94 use AUM as their headline figure and 0 use revenue. 145 carry a management-fee tag, yet the "revenue / money" field is unverified for 137 and the profit field for 143.
- The ownership structure explains part of it. Of the 146 issuers, 80 are affiliates of financial groups or conglomerates, and 13 are listed companies. An affiliate's ETF income easily disappears into group results, so what an ETF business earns shows up only at a few firms, like BlackRock, that disclose by product line.
- Revenue is visible elsewhere on the chain. The 16 players with revenue as their headline figure are 6 index firms, 5 market makers, 2 brokerage apps, 2 data vendors and 1 platform. The money of those standing on the road ETFs travel is easier to see than the money of those who make them.
When fees hit bottom, what holds up revenue?
Following one company, BlackRock, the world's #1, makes the answer clearer.
- ETFs' share grew, but the map says ETFs aren't what held up the fee rate. ETFs' share of base fees rose from 40.5% (2023) to 45.1% (9M 2026) (✔). The firm's base fee rate rebounded from 14.9 bps (2024 trough) to 15.4 bps (calculated). The map points to private markets: 2% of AUM generating 11% of base fees, ~79 bps (✔).
- A headline product isn't necessarily big money. IBIT, at $65.74B in net assets (2026-10-14 ✔), is the world's largest crypto ETF. But multiplied by its 0.25% fee, that is ~$164M a year, ~0.7% of BlackRock's 2025 revenue (calculated). Disclosed digital assets base fees were $39M in 3Q26, −36% (✔).
- Tech's share is flat, and securities lending is a supplement. Technology services revenue is $588M a quarter (+14% ✔), but its share stayed roughly flat, 8.3% to 7.8%. Securities lending is $253M a quarter (✔), which the map calls a de facto hidden fee.
Elsewhere on the chain, money lands differently
- Index firms charge in proportion to assets and keep a lot. All 21 index firms carry an index-licensing tag, and 13 also sell data subscriptions. Of S&P DJI's $534M revenue in 2Q 2026, asset-linked fees were $348M (+22%), with an adjusted operating margin of 71.5% (✔).
- Market makers earn on spreads, not fees. All 13 market makers carry a market-making/spread tag. The map records Jane Street's net trading revenue at ~$39.6B (△). Keeping ETF prices pinned to net asset value is a large business in itself.
- Brokerage apps earn on orders and cash interest. Of 38 brokers and apps, 32 carry a brokerage-commission/PFOF tag and 19 an interest-income tag, while 11 carry distribution fees. Trade Republic has 10M+ customers and €150B+ in assets (2025-09 ✔), yet FY2023/24 net profit was €34.81M (✔). By the map's calculation PFOF was 65–97% of net profit, and after the EU ban (2026-07-01) it shifted to €1/€2 order fees (△). The value of a single order seems to matter more to the business than the assets entrusted.
- Even a manager that went global earns a higher margin at home. Mirae Asset's standalone domestic operating margin of 46% tops the Global X holding company's 35% (✔). The claim that most net profit comes from Global X is also ✘: Global X equity-method income is 17% of net profit.
Newcomers go to distribution and active, not to scale
- The 77 emerging players sit in different places. Of the 77 that entered after 2016, brokers and apps are the largest group at 17, thematic boutiques and active/income managers 11 each, advisers and robo-advisers 9 and crypto ETFs 8. All 26 large issuers are legacy players. Places where scale is the moat are hard to enter, so new money goes to distribution and to products that can defend a fee.
- Active ETFs are how managers defend fees. Active ETFs hold $2.72T, 11.3% of the world total, yet took 34% ($663.6B) of Jan–Aug net inflows (✔). The map describes them as a way for managers to defend their fee rates.
- Small managers ride on someone else's platform. Providers worldwide have grown to 1,044, and the map says most list through white-label platforms such as Tidal and HANetf. The industry is splitting into "three giants plus many small players on white-label platforms." The top three hold 58.7% of the world market (✔).
The players that disappeared mostly lost their names
- Of 25, 18 were acquired or absorbed and 7 closed. 22 of 240 legacy players and 3 of 77 emerging players are marked as disappeared. But many legacy entries are rename cases added as a control group, so don't read this as a failure rate.
- Brands go before companies. Amundi bought Lyxor for ~€825M and folded in the brand. In Korea, early brands such as KBSTAR, ARIRANG, KINDEX and KOSEF were renamed or absorbed. Indices too: Lehman's bond indices passed through Barclays to become Bloomberg's, which the map records as indices surviving while only their owners change.
- What closes are small products. 73 US leveraged and inverse ETFs were liquidated in Jan–Jul 2026, and Hashdex, the smallest spot bitcoin ETF, stopped trading on 2026-08-17. According to the map, small products often lose money because fixed costs exceed fees.
Five questions to ask about an ETF company
- How much of the AUM is net inflow and how much is market gains? Of this year's $4.2T gain, roughly half came from net inflows.
- What is the average fee, and which way is it moving? The same assets at 0.0068% and at 0.75% are different businesses.
- Does it disclose ETF revenue separately? Of 146 issuers, 0 use revenue as their headline figure.
- What money sits outside the fee? Look for other lines that hold up the fee rate, such as securities lending, private markets or technology.
- Where on the chain does it sit? In the same ETF boom, index firms earn in proportion to assets, market makers to trading and brokerage apps to orders.
How to read the numbers
- AUM moves with the market. The #1 also depends on the measure. Globally it's iShares 27.3% vs. Vanguard 21.6% (ETFGI ✔), inside the US 29.7% vs. 28.7% (end-2025), while Vanguard leads in net inflows.
- Values we counted. Player counts are our own counts of the map data by primary place on the chain. "Issuers" groups large, mid-size, thematic, active, leveraged, crypto and AI managers by this post's definition. A tag marks that a revenue stream exists, not its share.
- What not to compare. The +31–34% for ChinaAMC and E Fund is company-wide revenue growth for 1H 2026, so don't read it as ETF-only revenue. Mirae Asset's 46% is also a standalone company-wide figure, not ETFs alone. Jane Street's ~$39.6B (△) is total net trading revenue and can't sit beside the fee-pool estimate.
- Two numbers for one thing. Mainland ETF assets appear both as RMB 4.93T from exchange data (✔) and RMB 4.92T via Wind (△). IBIT net assets also vary by date.
- Verdicts and tiers. ✔ = filing, primary source or official statistic, △ = press or estimate, ✘ = not the case. Of the 317, 68 have a ✔ headline figure, 248 a △ and 1 a ✘. Tiers run S $1B+ · A $100M+ · B $10M+ · C $1M+ and mix AUM with revenue, so filter by "Metric Type" as well. The 134 entries rated "?" have no verified headline figure in public filings.
AUM shows the size of the industry, but whose revenue it becomes only shows up once you also look at fees and place on the chain. And for most companies that make ETFs, that revenue is still not visible on its own.