Profiting from Hedge Funds
John Konnayil Vincent · 2013 · 291 pages
A mechanical system for building a stock portfolio out of the quarterly SEC filings that hedge fund managers are legally forced to disclose. The method is real and reproducible; the book's own backtests show it barely beating the S&P 500 over a period too short to prove anything.
The one idea
Every institutional manager with over $100 million in US securities must file Form 13F with the SEC within 45 days of quarter-end, disclosing every long stock position it holds. The book's entire premise is that ordinary investors can build a spreadsheet from these free public filings and legally clone what Warren Buffett, Bill Ackman, David Einhorn, and other value-oriented managers are doing, without paying a hedge fund's 2-and-20 fee or meeting its accredited-investor minimum.
The method has two moving parts: manager selection and portfolio construction. Selection matters more than construction. The book screens out managers whose 13F is a poor proxy for what they actually do: those focused on distressed debt, market-neutral pairs trades, quant strategies, non-US holdings, or heavy hedging, because none of that shows up cleanly in the filing. It narrows twelve profiled managers down to three usable ones (Ackman, Buffett, Einhorn) on the grounds that their US long positions are concentrated, long-held, and represent real conviction rather than a hedge or a sliver of a bigger global book.
Construction is a matter of picking an allocation formula (equal weight, weighted by position size, or a 10-5-2 tiered weighting) and a sentiment filter (largest existing positions versus largest new additions), then rebalancing on a trigger rather than a calendar: the book's rule is to rebalance only when the largest position in the model grows to 50 percent more than the smallest. Six combinations of allocation and sentiment choice are demonstrated on a single manager's 13F (Glenn Greenberg's Brave Warrior Capital) before the book runs three-year live backtests on Ackman, Buffett, and Einhorn individually and combined.
What it actually teaches
The book builds a cloned portfolio in a fixed sequence: pick managers whose 13F is trustworthy, choose how to divide cash among their positions, choose which positions to track, then rebalance only when the model drifts.
SCREEN MANAGERS FOR 13F FIDELITY
Before touching a spreadsheet, eliminate managers whose 13F does not represent their real book: distressed-debt specialists (their debt holdings never appear in the filing), market-neutral and pairs traders (the short leg is invisible), quant funds (positions are stale by the time they post), and anyone with heavy non-US or hedged exposure. Of the twelve managers profiled in Part One, the book keeps only three for the mechanical models: Bill Ackman, Warren Buffett, and David Einhorn, each running concentrated, long-term, mostly-US long books.
PICK AN ASSET ALLOCATION FORMULA
Equal allocation splits cash evenly across however many positions you track (five stocks, $50,000 capital, $10,000 each). Weighted allocation sizes each position by its share of the manager's disclosed portfolio, approximated with a fixed ratio; the book's example uses 3:2:1 across a manager's top five positions, so the largest position (17.4 percent of the source portfolio in the worked example) gets 33 percent of the model's cash. The 10-5-2 allocation formalizes a common manager pattern: a very small number of positions over 10 percent, an average number over 5 percent, the rest under 2 percent, demonstrated with a 10:5:2 weighting ratio.
CHOOSE A SENTIMENT CAPTURE FILTER
Top holdings tracks the manager's largest existing positions, which captures conviction but can hold overvalued stocks the manager is loyal to. Best ideas tracks the largest new positions added quarter over quarter, which captures current sentiment but requires more trading and can pick up a position a value manager nibbles at once and abandons. The book explicitly discards a third option, following the most popular positions across many managers, on the grounds that consensus-chasing works against you (a University of Toledo 2012 paper cited in the book found rising aggregate institutional demand for a stock predicts falling forward returns).
REBALANCE ON DRIFT, NOT ON A CALENDAR
Update the model every quarter as soon as new 13Fs post (45 days after quarter-end), replacing positions the manager dropped. Separately, rebalance the dollar allocation whenever the largest position in the model exceeds 150 percent of the smallest position's dollar value, rather than on a fixed schedule. In the book's three-year Ackman, Buffett, and combined backtests, this trigger fired only once or twice across three years, keeping trading to under one transaction per quarter on average.
What it looks like Monday morning
A reader who wants to try this pulls a target manager's 13F-HR filing from SEC EDGAR (search by firm name, filter to 13F-HR, download the Information Table), builds the spreadsheet columns the book lays out (name, market value, share count, percent of portfolio), and compares it to the prior quarter's filing to see what was added, trimmed, or dropped. From there it's a matter of picking one allocation formula and one sentiment filter, sizing a paper or small live position, and setting a calendar reminder for 45 days after the next quarter-end when the fresh 13F posts.
The more useful Monday-morning habit the book pushes is the manager-selection screen itself, independent of the mechanical models: before following anyone's 13F, check whether their style survives the filing's blind spots. A manager who hedges, shorts, trades distressed debt, or invests mostly outside the US is not a candidate no matter how good their track record looks, because the 13F will show you a distorted fraction of what they actually own.
| Model | Ending value | Total return | S&P 500 return (same window) |
|---|---|---|---|
| Bill Ackman, largest three positions | $157,797 | 57.80% | 43.46% |
| Warren Buffett, largest three positions | $156,310 | 56.31% | 43.46% |
| David Einhorn, largest three positions | $140,017 | 40.02% | 43.46% |
| Combined, largest position from each manager | $177,178 | 77.18% | 43.46% |
| Combined, largest new position from each manager | $157,186 | 57.19% | 43.46% |
Einhorn's three-stock model underperformed the index; the book attributes the combined portfolio's stronger result largely to a single Apple position carrying the final year.
Picking managers to clone is more art than science.John Konnayil Vincent, Profiting from Hedge Funds
Where it fails
- The backtests are three years, and the author admits it. The book's own summary of the equal allocation chapter states plainly that three years is 'obviously way too small to prove the mettle of these models.' The Ackman three-stock model returned 57.8 percent against the S&P 500's 43.46 percent over August 2009 to August 2012, Buffett's returned 56.31 percent, and Einhorn's underperformed at 40.02 percent. That is one bull-market stretch off the 2009 bottom, not a track record across a bear market or a rate-hiking cycle, and the book never revisits the models over a longer window.
- Every backtest excludes dividends, and the comparison index does not. The book states outright that 'dividend income was ignored to provide a precise comparison with the S&P 500 Index performance,' but the S&P 500 total-return figures typically quoted (and likely what 43.46 percent approximates) already bake dividends in for the index while the model portfolios do not collect any. That asymmetry flatters the index's side of the comparison, not the model's, so the real gap between the models and a dividend-reinvested benchmark is understated in the book's favor, not against it.
- The single-manager models are openly acknowledged as too small to trust. The three-stock Einhorn model spent its first year underwater and was then 'literally carried by the vast outperformance' of a single Apple position, per the book's own text. The author states real implementations need at least ten positions and five or more managers combined, meaning the worked examples that carry most of the book's page count are demonstrations of a toy version of the method, not the version the author actually recommends anyone run.
- The method assumes 45-day-old information doesn't matter, which the book concedes it sometimes does. 13F data is disclosed up to 45 days after quarter-end, meaning a trade made on day one of a quarter can be four and a half months stale by the time it is public. The book calls this fine for patient value managers but flags, without resolving, that it 'renders cloning useless' for anyone whose strategy is time-sensitive, and that confidential-treatment requests let some managers delay disclosure up to a year on individual positions.
- The reader needs a brokerage account, spreadsheet discipline, and a live EDGAR habit the book doesn't supply. Nothing here is push-button. The reader has to manually pull 13F-HR filings from EDGAR each quarter, reconcile column formats that differ manager to manager (the book shows Pabrai's clean eight-column filing against Buffett's messier one with duplicate rows for options and warrants), and keep a running spreadsheet current. There's no companion tool, template file, or website bundled with the book; AlphaClone, the commercial product built by a researcher the book cites approvingly, existed at time of writing to do this automatically for a fee, which somewhat undercuts the book's DIY pitch.
- 2013-era manager selections have aged out; three of the four featured managers no longer look like the book's case for them. The book builds its strongest argument for Ackman around his activist, high-conviction 13F disclosure. Ackman's Pershing Square subsequently had well-documented rough multi-year stretches (Valeant and Herbalife positions are widely known to have gone badly wrong after this book's 2013 publication) that a 2009-2012 backtest could not have captured. A reader relying on the book's manager profiles as current judgment of who is safe to clone is relying on a decade-plus-stale read.
Who it is for
Buy it if
You already know how to read a 13F filing (or are willing to learn) and want a disciplined, low-turnover way to build a long-only equity sleeve around a handful of value managers you trust, using nothing but free EDGAR data and a spreadsheet. It also suits a reader who wants the mechanics of 13F cloning explained clearly enough to build their own screening rules rather than trust a commercial cloning product.
Skip it if
You want a validated, market-tested strategy: the book's own backtest window is three years, ends in 2012, and excludes dividends from the model side of every comparison it makes. Skip it too if you were hoping for an automated tool or template rather than a spreadsheet method you build yourself quarter by quarter from raw SEC filings.
As an Amazon Associate I earn from qualifying purchases. It costs you nothing extra, and it does not change what gets recommended or what the verdict says.