Why cloning a 13F portfolio is harder than it looks

The structural gaps between what a manager did and what a 13F reader can reconstruct.

"Coattail investing" — replicating famous managers' filings — is an old idea, and this page is not advice for or against it. It is an inventory of the gaps between a manager's actual portfolio operation and what a filing reader can reconstruct, because those gaps are facts about the data.

The gaps

Timing. You learn about positions a median of ~75 days after quarter-mid. The prices available to you are not the prices the manager saw, and for volatile names the difference is the whole game.

Selection. You see U.S. reportable longs. You do not see the hedge, the credit position, the foreign leg, or the cash that made the position sizable but survivable. Copying the visible half of a hedged trade reproduces the risk without the hedge.

Exits. The manager exits when they exit; the copier learns up to 45 days later. Bad exits compound worse than good entries — the copier owns every blow-up for at least one extra reporting cycle.

Sizing. A 4% position at a fund with locked-up capital and a decade horizon is a different instrument than 4% of a household portfolio that may need liquidity next year.

What the research use looks like

Aggregated filings are better at describing the market than at generating trades: ownership breadth, crowding, sector rotation, the divergence between what concentrated active managers and index flows are doing. That is the layer this site is built to expose. Where the site shows models based on public-filing replication, they are labeled historical research with their assumptions stated, and never presented as achievable or recommended returns.