Methodology

Institutional managers above $100 million in US equity assets file a Form 13F with the SEC each quarter, listing every reportable position. We hold the full history of those filings and compare each quarter to the one before it. Where a manager files an amended report, the superseding version is used.

Every dollar figure in these calculations is rebuilt as shares × quarter-end closing price rather than taken from the value a filer reports. Reporting conventions vary between filers, and deriving value from share counts and market prices keeps every manager on the same basis.

Turnover follows the standard definition used for registered funds: the smaller of total purchases and total sales, divided by average portfolio value. Taking the smaller side is what makes managers of different sizes comparable — a manager taking in new money buys far more than it sells, and that inflow is not trading.

Each quarter is measured against the previous one and the result is annualised. Positions opened from nothing count as purchases in full; positions closed out count as sales in full.

Bands are fixed thresholds, not rankings. A manager moves between them when its own trading changes, not when the managers around it do.

For every security, we compare the weight it holds in a manager’s portfolio against the weight it holds in the total market, and add up the smaller of the two. A portfolio holding the whole market in proportion scores near 100%. A portfolio of a dozen conviction positions scores near zero.

This is the same construction as active share, measured against the market as a whole rather than a named index, since a 13F filing does not say what a manager benchmarks itself to.

Overlap on its own cannot separate an index manager from a large bank. Both hold thousands of names in roughly market proportion. What distinguishes them is that one barely trades and the other trades constantly — so orientation reads both measures at once.

• Index / Passive — broad holdings, very little trading.
• Diversified active — broad holdings that genuinely turn over.
• Concentrated long-term — a focused book, held patiently.
• Concentrated trader — a focused book that moves.

Every security is ranked against the whole universe on three characteristics: how cheap it is (price to book, price to earnings, price to sales, price/earnings to growth), how fast it is growing (reported earnings and revenue growth, plus the growth implied by the gap between trailing and forward earnings multiples), and how much income it pays. A manager’s score on each is the weighted average across everything it holds.

Those scores are then expressed as a tilt against the market — the manager’s score minus the market’s own. This matters more than it sounds. Measured absolutely, every broadly diversified manager looks alike, because a diversified portfolio converges on the market by construction. Measured as a tilt, the question becomes the useful one: relative to the market, does this manager lean cheap or expensive, fast-growing or slow?

A tilt of +10 means this manager’s holdings sit ten percentage points further up the cheapness rankings than the market average. Zero means indistinguishable from the market.

Figures are average tilts in percentage points, live from the current quarter.

• Deep Value and Value — holdings materially cheaper than the market, and growing more slowly.
• GARP — cheaper than the market and growing faster. Growth at a reasonable price means both at once, so it is tested on both measures rather than as a midpoint.
• Core — no meaningful tilt in either direction. The portfolio looks like the market.
• Growth and Aggressive Growth — faster growing and more expensive than the market, by degree.
• Yield — a decisive tilt toward income-paying holdings.
• Momentum — see below.
• Index — tracks the market by holding its constituents. ETF — allocates primarily through funds rather than individual securities.

Momentum is measured through trading rather than holdings. For each quarter we compare the prior-year returns of what a manager bought against the prior-year returns of what it sold. Buying winners and selling laggards produces a positive reading.

Measuring holdings instead would mostly restate growth — owning shares that have risen and owning growth companies are largely the same portfolio. What a manager does with new money is a different signal.

The reading is averaged across several quarters, because a single quarter of purchases landing in risen shares is chance rather than strategy. It is also withheld from portfolios that track the market, which add to whatever has risen simply because index weights drift, and from portfolios that trade too little for the question to mean anything. Only the most decisive tenth of managers carry the label.

• Holdings — SEC Form 13F, filed quarterly, taken from EDGAR directly.
• Prices — quarter-end closing prices and trailing returns for each security.
• Company fundamentals — valuation, growth and income measures across 5,736 securities, refreshed on a rolling basis.

Classifications are recalculated weekly, and again after each quarter’s filings are processed. Because 13F reporting runs 45 days behind quarter end, the most recent filings describe positions as they stood at the close of the previous quarter.