The stocks the index is quietly buying.
Every dollar that flows into an ASX ETF buys a fixed basket of shares, price-insensitively, on a schedule, whether the stock is cheap or dear. It is the one large buyer in the market with no opinion. We reconstruct that buyer from the bottom up, company by company. In August it added +A$1.24 billion of mechanical demand across ASX equities. The number isn’t the story. Where it tilted is.
A buyer with no opinion
Passive money is the most predictable flow in the market, and the least understood at the single-stock level. When a saver’s contribution lands in a broad-market ETF, the fund does not decide whether to buy Commonwealth Bank. It buys CBA because the index says so, at the index’s weight, at whatever price the screen shows. Multiply that across every ETF holding a name and you get a bid that is large, mechanical and utterly indifferent to valuation. For a positioning desk, indifference is the point: it is demand you can anticipate, because it is a function of flows and weights you can observe, not judgements you cannot.
We measure it in two quantities. The first is the stock of ownership, the dollars every ETF collectively holds in a company, its look-through footprint. The second is the monthly flow, the mechanical demand a company receives this month purely because money moved into the funds that hold it, at those funds’ weights. The first tells you how exposed a name is to the passive complex; the second tells you which way, and how hard, that complex is currently pushing.
mechanical demand = Σfunds netflowfund × weightfund,stock
Built by resolving the look-through holdings of 282 ASX ETFs to their underlying ASX-listed shares, weighted by each fund’s size and this month’s net flow. Dual-listed names are resolved to the correct listing; geared funds excluded.
The bid is broad, so read the cross-section
In a net-inflow month the sign tells you nothing. The tilt is everything.
August was a net-inflow month for Australian ETFs, and here the mechanics assert themselves bluntly: if money is flowing into the funds, and the funds hold the index, then every widely-held name gets bought. Of the 119 ASX companies with more than A$100m of ETF ownership, all 119 recorded positive mechanical demand. The direction of the bid is not information. Knowing that passive bought CBA in a month passive bought everything is worth nothing.
The information is in the cross-section: how hard the bid pressed on each name relative to the ownership it already carried. We measure that as demand intensity, this month’s mechanical demand as a percentage of the stock’s ETF base. A name held mostly by ETFs that drew heavy inflow scores high; a name held by funds that were quiet scores low, whatever its size. The median name ran at 1.39%. The spread around it is the map of where August’s passive money actually leaned.
The passive bid pressed hardest on established, profitable businesses, and ran coolest through high-multiple growth software, a rotation executed without a single discretionary decision.
The top of the chart is a portrait of quality, not of any one sector. It is led by mining-services engineer Monadelphous, appliance-maker Breville and a2 Milk, funds-platform Pinnacle, gold miner Perseus and blue-chip CSL, established, cash-generative businesses that happen to sit in the thematic and quality-tilted ETFs where August’s marginal dollar went. The bottom is just as legible, and more interesting: high-growth software and platforms, Carsales, NextDC, Megaport, Life360, joined by the A-REITs. None are being sold; every one still printed positive demand. They simply sit in funds that drew a thinner slice of the flow, so the passive bid reached them at less than half the market’s pace.
One subtlety separates this from a simple flow screen: the biggest dollar bid and the strongest intensity are not the same name. BHP received the largest mechanical demand in the market, +A$108.9m, yet its intensity was just 1.22%, below the median. BHP is bought heavily because it sits near the top of nearly every index, not because August’s flows favoured it. Read the two together and you separate the stocks the passive complex is structurally obliged to buy from the ones it is currently choosing to chase.
Why the “gold crowding” headline is mostly offshore
A ticker-level screen would have handed you a clean story this month: Newmont, the world’s largest gold miner, as the ASX’s hottest and most-crowded name, around A$690m held across 50 funds at a 3.15% intensity, the perfect “passive is piling into gold” headline. It is also wrong. Newmont is dual-listed, an ASX CDI and a US primary line. Resolve the holdings by listing and the ASX CDI is an ordinary name; the rest is US-listed Newmont, held by global gold-miner ETFs whose flows have nothing to do with the ASX, and correctly booked offshore in Part 2.
Ticker screen (wrong)
Resolved by listing
The genuine domestic gold bid is Perseus Mining at 1.79%, a real tilt but a modest one; the gold complex as a whole ran at 1.60% against a 1.39% median. A ticker is not a security, and a dual-listing resolved carelessly imports an offshore story into your home signal. That distinction is the line between a screen and a signal.
The base load: where the dollars actually go
Intensity finds the tilt. Raw dollars find the weight, and the weight is astonishingly concentrated.
Strip out intensity and rank by dollars, and the passive complex shows its true shape: a mega-cap machine. The eight largest mechanical bids (BHP, the four major banks, CSL, Macquarie and Wesfarmers) absorbed 41% of the entire month’s A$1.24bn. Widen to twenty names and you have captured 61%. Three hundred-odd resolved companies share the passive bid, but two dozen of them take almost two-thirds of it.
| Stock | Company | Mech. demand | ETF-held | In funds |
|---|---|---|---|---|
| BHP | BHP Group | +A$108.9m | A$8.89bn | 43 |
| CBA | Commonwealth Bank | +A$104.5m | A$7.15bn | 37 |
| NAB | National Australia Bank | +A$65.1m | A$4.28bn | 62 |
| WBC | Westpac | +A$51.5m | A$3.61bn | 40 |
| ANZ | ANZ Group | +A$49.7m | A$3.44bn | 39 |
| CSL | CSL | +A$46.1m | A$2.69bn | 42 |
| MQG | Macquarie Group | +A$41.8m | A$2.68bn | 40 |
| WES | Wesfarmers | +A$39.5m | A$2.20bn | 36 |
The sector concentration underneath is starker still. The four major banks plus Macquarie drew +A$313m between them, a quarter of the market’s entire passive bid funnelled into five financials, at an intensity (1.48%) just above the median. Iron ore (BHP, Rio Tinto, Fortescue) took a further +A$152m, but at just 1.25% intensity: heavy in dollars, light in tilt, the definition of base load. The contrast with gold’s 1.60% is the point of separating the two lenses. “Miners” is not one trade to the passive bid; it is a base-load iron-ore position and a modestly-chased gold position sitting in the same sector bucket.
Crowding: where a reversal would bite
The bid that lifts a name price-insensitively will, one day, leave it the same way.
Everything above runs in reverse. The property that makes passive demand tradeable on the way in, that it does not care about price, is exactly what makes it dangerous on the way out. When flows turn, the funds sell the index at the index’s weight, into whatever bid exists, and the names they all hold in common take the pressure together. A stock owned across many ETFs is not diversified by that breadth; it is correlated by it. Breadth, the number of distinct funds holding a name, is the cleanest crowding gauge we have, and unlike ownership-as-%-of-float it needs no shares-outstanding data.
| Stock | In funds | ETF-held | Intensity |
|---|---|---|---|
| NAB | 62 | A$4.28bn | 1.52% |
| TLS | 45 | A$1.91bn | 1.59% |
| BHP | 43 | A$8.89bn | 1.22% |
| SUN | 43 | A$750m | 1.41% |
| CSL | 42 | A$2.69bn | 1.71% |
| FMG | 42 | A$1.03bn | 1.39% |
NAB is the most broadly-owned name in the complex, held across 62 of the 282 funds, the base-load position most exposed to a broad, indiscriminate outflow. The pattern here is the mirror of the intensity chart: the widest-held names are the banks, Telstra, BHP and CSL, the index’s ballast. They are not where the tilt is, but they are where a flow reversal would be felt first and hardest, because everyone owns them through everything.
Method & limits, up front
- It’s a floor. True ETF demand is at least what we measure. Look-through covers 92% of passive FUM and 84% of the market; the uncovered part is commodity/bond funds with no equities to resolve, and non-disclosing active managers.
- Mechanical, not total. This isolates the price-insensitive passive bid by construction. It excludes discretionary and active buying, and geared funds, whose leverage would overstate weights. That is the feature, not a gap.
- Names are resolved to a listing, not a ticker. Each line is matched to an ASX-listed equity by ticker and corroborating name, and dual-listed companies are attributed by the holding fund’s mandate, so a global fund’s US line is not booked as the ASX CDI (see the gold mirage, above). Bonds, cash, foreign and collision lines are excluded.
- Period alignment. Holdings weights are the current look-through; flows and FUM are the latest ASX disclosure (August). A labelled approximation until holdings history accumulates and every input is point-in-time.
- Dollars, not % of float. A shares-outstanding feed would express demand as a share of each company’s register, a clean upgrade we can add, not a prerequisite.
The same engine, pointed at the global look-through. Via their ETFs, Australians now hold more Nvidia than Commonwealth Bank, and a bigger book offshore than at home.