The system

How the measurement works

Three questions, answered from three separate inputs: what people say online, what ETFs file, and what ETFs trade at. They meet on one narrative and are never added together. First the machine, then the five questions issuers ask about it.

The machine

Attention runs along the top, ownership through the middle, performance along the bottom. They share a narrative and nothing else.

Swipe the chart sideways to see all of it

Six channels· Search· News· Video· Forums· Reference pages· Investor boardsETF documents· Prospectus objective· Named secondary· Holdings· AssetsMarket prices· Weekly closes· Distributions· The S&P 500· Fees, as tradedAttentionOwnershipPerformance1Score0 to 100 per channel2Averageinto one level3Placeamong all 224Classifyby the filing5Weighassets × exposure6Shareof the capital7Measureweekly total return8Removethe market’s share9Comparewith its narrativethree streams, never mergedThenarrativeshown side by side onThe boardall 22, every FridayThe wireevery weekdayA pilotweekly and monthly
See it working: The board ranks all 22 every Friday. The wire applies it to the news every weekday.

Nine steps

1
Score each channel

Every week each of the six channels gives each narrative a score from 0 to 100: how much of that channel’s attention the story drew.

2
Average into one level

The six scores are averaged with equal weights. No channel can carry a narrative on its own, and a missing channel is shown as missing, never filled in.

3
Place against the other 21

The level is ranked against the week’s other narratives. The top 15% are HOT, the bottom 15% QUIET. It says where a story sits this week, not where it is going.

4
Classify by the filed objective

Each ETF’s narrative comes from its prospectus, not its name. A second narrative counts only where the filing names it.

5
Weigh assets by exposure

An ETF’s assets are multiplied by how much of it belongs to the narrative. An ETF can carry two narratives; its money is split between them.

6
Share of the capital

Each ETF’s share of all the money carrying the narrative, and how much of it the five largest ETFs hold.

7
Measure weekly returns

Each ETF’s total return at its traded price, Friday to Friday, over up to five years. Distributions are added back; fees are already in the price.

8
Take the market out

The part of each week’s return that the S&P 500 explains is removed, for the ETF and for its narrative alike. What is left is what each did on its own.

9
Compare with its narrative

What is left is compared with the narrative’s other ETFs, never with the ETF itself. A fit of 0.70 or more reads Yes, 0.40 to 0.70 Partly, below 0.40 No.

Five questions, answered

What an issuer asks first, each in two sentences and one diagram.

Why not just use the ETF’s name?

Names are marketing; filings are not. IUSB is called Core Universal, but its filing names investment grade and high yield, so it carries the rate regime first and the credit signal second.

The filingwhat it must holdThe namewhat it is called×primarysecondthe split follows the filing,never the label

Why three answers, and never one score?

How loud a story is, who holds it and whether an ETF delivers it are different questions. One score would hide the cases that matter most: loud and thinly held, or widely held and not delivered.

How loudattentionWho holds itownershipDelivers itperformanceone score×shown side by side, never added

Why take the market out?

Most ETFs rise and fall with the market, so almost any stock ETF moves with almost any story. Taking the market out shows whether an ETF still follows its own.

Moves with its storythe first test: the market does much of the workWith the market taken outthe harder test: what is left is the ETF’s own

Is any of this a forecast?

No. A level is one week’s reading, like a thermometer, and returns and fit describe the past. Nothing on the site says a narrative or an ETF is rising, falling or about to move.

This weekone reading×Next weeknot predicted

Can an ETF pay for placement?

No. ETFs are listed by their measured share of the narrative’s capital. There is no editorial slot and nothing to buy.

ETF AETF BETF CETF DETF Eorder = measured share. Nothing else.

The words, defined

NarrativeOne of 22 ratified stories the market trades. The list changes only by ratification.
LevelHow loud a narrative is this week: six channel scores averaged, shown on a fixed 25 to 60 scale.
BandWhere the level sits against the other 21 this week. HOT is the top 15%, QUIET the bottom 15%.
Capital Share of NarrativeAn ETF’s assets times its exposure to the narrative, as a share of all the capital carrying it.
Top fiveHow much of that capital the five largest holders have.
Market price total returnWhat an ETF returned at its traded price, with distributions added back. Fees are already in the price.
FitHow closely an ETF’s weekly returns, with the market taken out, move with its narrative’s other ETFs. Read from 0 (unrelated) to 1 (in step).
DeliversYes at a fit of 0.70 or more, Partly from 0.40 to 0.70, No below 0.40.

What it cannot see

The limits, stated where the method is.

A margin of errorA fit measured over one year carries about 0.14 either way; over five years, about 0.06. Ranks closer than that are shown as ties.
Today’s map, applied to the pastAn ETF in a narrative today is treated as having been in it throughout, weighted by today’s assets.
Survivors onlyETFs that closed are not in the price history, so early years show only the ETFs still trading.

Run it on your own ETFs, or read it every weekday

The pilot runs all three on the narratives your ETFs carry and turns them into ideas for your marketing, distribution and product teams, every week. The wire lands every weekday. Free. No card.