Every quarter, a few thousand investment managers tell the SEC what they own. Within hours, a cottage industry of newsletters, screeners and copycat funds goes to work on those filings. The pitch is simple. Find the best managers, buy what they bought, and collect returns you could never earn on your own.
The idea is old enough to have been studied carefully. This article walks through what the research actually found, where the strategy holds up, and where it quietly falls apart. It is not investment advice. It is a reading of the evidence.
What a 13F filing is, in one paragraph
Form 13F is a quarterly report that institutional investment managers with at least $100 million in qualifying assets must file with the SEC. It lists long positions in a defined set of securities, mainly US-listed stocks, certain options and convertible bonds. The official description of the form and who must file is on the SEC's Form 13F page. The filing is due within 45 days after quarter end, and that lag is the whole story of this strategy. We have covered the deadline mechanics in detail in 13F deadlines and the 45-day lag, so this piece assumes you know how the calendar works.
Why anyone thought copying would work
The academic case for cloning rests on two findings that predate the copycat industry.
First, there is evidence that some institutional managers have skill. Not most of them, and not reliably over short windows, but a subset of managers show persistent stock-picking ability in their largest and highest-conviction positions.
Second, prices do not fully adjust the moment a filing becomes public. The market is efficient at absorbing earnings and macro news. It is slower at absorbing the information that a specific manager with a strong record took a new position. If that is true, a public filing still carries usable information even weeks after the trades happened.
Put those two together and you get a hypothesis. Copy the high-conviction positions of skilled managers as soon as the filing hits EDGAR, and you capture a meaningful slice of their edge for free.
What the studies actually found
Several academic papers have tested this directly using the full history of 13F filings on EDGAR. The broad shape of the results is consistent across them, and it is more nuanced than either the promoters or the skeptics tend to admit.
Copying works better than most people expect. Portfolios built from the disclosed holdings of top-performing funds, rebalanced only when new filings arrive, have in several studies produced returns that are close to the returns of the underlying funds before fees. The reason is that most hedge fund alpha, where it exists, comes from positions held for many months or years rather than from trades executed and unwound inside a single quarter. If a manager holds a stock for two years, buying it 45 days late costs you a small fraction of the total move.
The gap is real and it is concentrated. The same research finds that the copycat portfolio does underperform the actual fund in a predictable way. The shortfall is largest in positions that the manager trades quickly, in small and illiquid stocks where the manager's own buying moved the price, and around event-driven trades such as merger arbitrage where the opportunity has a short shelf life. In those cases the 45-day lag is fatal. By the time you can see the position, the catalyst has already played out.
Fees matter more than lag. A recurring conclusion is that the net return advantage of a clone over the original fund comes largely from not paying the fund's fees. A hedge fund that charges two percent of assets and twenty percent of profits has to beat its own clone by a wide margin every year just to break even for its investors. Many do not. This is the strongest practical argument for the copying approach, and it has nothing to do with stock-picking skill.
Selection is the hard part. Studies that cloned the holdings of a broad group of funds found little or no edge. The results depend on picking managers whose past performance reflects skill rather than luck, and on sticking with them through periods when they lag. This is the same problem every fund-of-funds investor faces, and the research gives no reason to think it is easier when you do it yourself.
What 13F filings do not show you
The strategy has structural blind spots that no amount of clever analysis can fix, because the information is simply not in the filing.
Short positions are not reported. A manager who is long one airline and short another is running a relative value trade. You see only the long side. Copying it gives you naked exposure the manager never had. For long-short equity funds this is not a corner case. It is the core of the strategy.
Most derivatives are not reported. Put options on listed stocks appear, but swaps, futures and many structured positions do not. A fund that hedged its entire equity book with index futures looks fully exposed on paper.
Non-US and private holdings are not reported. The 13F covers a specific list of securities. Foreign stocks held directly, bonds, currencies and private investments are outside it.
Position sizes are stale and sometimes misleading. The filing shows the share count on the last day of the quarter. A manager who bought in October and sold in December shows nothing at all. A manager who was selling into the quarter end shows a position they no longer believe in.
Confidential treatment exists. Managers can request that certain positions be withheld from public view while they build them. The SEC describes this process in its 13F frequently asked questions. When the request is granted, the position appears months later in an amended filing, long after any informational value has decayed.
The honest arithmetic of the lag
A simple way to think about the 45-day delay is to ask what fraction of a typical holding period it represents.
If a manager's average holding period is three years, a 45-day lag means you miss roughly four percent of the holding period. For a manager who turns the whole book over every six months, you miss a quarter of it, and the quarter you miss is the one where the manager had the freshest view.
This is why the research consistently points toward low-turnover, concentrated, long-biased managers as the only sensible cloning targets. Activist investors and long-term value managers fit this profile. Quantitative funds, event-driven funds and anything with high turnover do not. For the latter group, the filing is a historical document, not a signal.
There is a second, less obvious cost. When many copycats act on the same filing at the same time, they move the price together. The first few hundred buyers after a filing may get a reasonable entry. The late ones pay a premium created by the early ones. Any edge that exists in public filings is shared among everyone reading them, and it shrinks as more people read.
Free ways to do this yourself
None of this requires a paid product. The raw filings are free and the tools to read them have improved a great deal.
The SEC's EDGAR full-text search and the company filing browser let you pull any manager's 13F-HR filings directly. Since 2013 the holdings have been in a structured XML format that any spreadsheet can import. The SEC also publishes a quarterly list of the securities that count as 13F-reportable, which is useful for understanding why a known position is missing.
If you want to track a handful of managers, a spreadsheet and a calendar reminder for the filing deadline is enough. Compare each new filing with the previous one, flag new positions and exits, and ignore small changes in share counts that often reflect flows rather than conviction.
Where a paid tool earns its keep is in doing that comparison across hundreds of managers at once, flagging when several independent managers moved into the same name in the same quarter, and filtering out the noise from index-like funds whose holdings carry no information. That consensus view is something a single person cannot build by hand in the window between filing and the next quarter.
What a reasonable reader should take away
The evidence does not support the claim that copying hedge funds is a free lunch. It also does not support the claim that the 45-day lag makes the filings worthless. The truth is in the middle, and it is specific.
Cloning works tolerably well for a narrow set of managers: concentrated, low-turnover, long-biased, with a track record long enough to separate skill from noise. It fails for everyone else. It never shows you the hedges, the shorts or the trades that opened and closed within a quarter. And the biggest single benefit it offers is not stock-picking alpha but fee avoidance.
If you want to see which large managers are converging on the same stocks each quarter, without paying for a dozen newsletters or parsing XML yourself, the Smart-Money 13F Consensus report scores every quarter's filings for overlap, conviction and persistence, and lists the names where independent managers agree.
Want the signal instead of the raw filings? Get the free 13F Consensus preview. Prefer the tool to the write-up? Browse all data feeds or connect the free MCP server.