Take advice only from investors with a track record

Why should I only take advice from investors with a track record?

Because talk is free and capital is not. An investor with a long record has been forced to be specific: to buy at a price, to size the position, to hold through a decline, and to sell. A commentator has done none of those things and pays nothing for being wrong. The filter is not intelligence or eloquence, it is whether the person has compounded real money over a long enough period that luck has had time to wash out. Once you apply that filter, the useful question changes from what they say to what they own.

The asymmetry is worth sitting with. A person who has run capital for thirty years has been wrong in public, repeatedly, with money on the line, and has had to keep going anyway. Every position they hold today survived that process. A person who writes about markets has been wrong just as often, without the cost, and their record is whatever they choose to remember. Both may be intelligent. Only one has been graded.

There is a second reason to prefer holdings over commentary, and it is less about honesty than about resolution. Opinions are vague by nature. A position is exact. It has a size, which tells you conviction. It has a date, which tells you what price looked acceptable. It appears next to the other positions, which tells you what it is competing with. You learn more from one line in a filing than from an hour of talk.

What is cloning in investing?

Cloning means taking your ideas from investors who have already proven they can pick stocks, rather than generating every idea yourself. Mohnish Pabrai made the term famous and is blunt about it: he cloned Buffett, who cloned Graham, who was refined by Fisher and Munger. The research supports the idea more strongly than most people expect. Martin and Puthenpurackal found that copying Berkshire Hathaway from its public 13F filings, buying at the start of the month after each disclosure, beat the S&P 500 by roughly 10.75% a year from 1976 to 2006, which suggests the market is slow to price in what a skilled investor has already concluded.

Pabrai's framing is that originality is overrated in this business, because good ideas are scarce and there is no prize for having found one yourself. The lineage he points to is real: Buffett took Graham's framework and adapted it, Munger pushed him toward quality, and Pabrai took the result. Cloning done properly is not copying a trade. It is copying a way of thinking, and using another investor's conclusions as a filter on where to spend your research hours.

The evidence on the narrower version, copying the trades themselves, is stronger than intuition suggests. The Martin and Puthenpurackal study of Berkshire's filings covers thirty years and a delay long enough that anyone could have acted on it, and still finds material excess return. The mechanism they propose is simple: the market underreacts to the news that a skilled investor has bought, and takes months to finish pricing it in. That is a slow-moving edge available to anyone patient enough to use it.

What is a 13F filing and what does it show?

A 13F is the quarterly holdings report every institutional manager with more than $100m in US equities must file with the SEC. It lists their long positions in US-listed stocks at the close of the quarter, with share counts and market values. Managers have up to 45 days after quarter end to file, so a filing you read in mid-August shows positions as of 30 June. It is the only regular window into what serious investors actually did, as opposed to what they said on television.

Nine investors are tracked here, and the filter is deliberately narrow. Each one has compounded real capital over a long period in a way that can be checked. The list is dominated by value and quality investors because those are the styles whose holding periods survive the filing delay, with one macro investor included precisely because his book looks nothing like the others.

All nine filings are dated 30 June 2026 and the returns are read on 29 August 2026, so this is two months of a single quarter, which is far too short to rank anybody. What the table is good for is shape. Pabrai holds four names and 43% of the book sits in one of them. Li Lu holds seven, with 48% in Alphabet. Davis holds 112 and Druckenmiller 95. The concentrated books are the ones where a filing tells you something, because a 1% position is a maybe and a 43% position is a conviction.

What does a 13F not show?

A great deal. No short positions, so a long you are copying might be one leg of a hedge. No cash, so you cannot see whether the manager is fully invested or sitting on half a book of cash. No foreign-listed stocks, no bonds, no private holdings. Options appear only in limited form and are reported at underlying notional value, which can make a hedge look like a huge conviction bet. And no leverage: the filing tells you what was held, never how much was borrowed to hold it. Phil Town’s filing in our data is the perfect illustration: 77% of the reported book is a put on the S&P 500, so only 13% of it is ordinary priceable stock.

The priced column in the table above exists for exactly this reason. For Town, only 13% of the filed book is ordinary stock we can price, because the rest is index puts reported at notional value. Reading his filing as a stock list would tell you almost nothing about his positioning, and a naive return calculation on it would be fiction. Druckenmiller sits at 64% because his book carries options and unresolved securities alongside the equities. Coverage under 90% appears on the surface as a label rather than being quietly rounded away.

This is the general lesson about coattail investing. A 13F is a photograph of one leg of a strategy, taken 45 days ago, with the rest of the frame cropped out. You are looking at the longs and inferring a thesis you cannot see. That is workable for a long-horizon owner whose entire strategy is the longs. It is close to useless for a hedged or levered book.

How does Invest Board handle super-investor portfolios?

Nine investors are tracked, chosen because each has a long record of compounding real capital: Buffett, Ackman, Klarman, Li Lu, Akre, Pabrai, Druckenmiller, Town and Davis. Their latest 13F holdings load as a board you can read like your own portfolio, with our scores, valuation signal, quality metrics and price history beside every position. Return since the filing date is computed on the priced basis only, summing current values against filed values for holdings we can actually price, and each investor carries a coverage percentage so a thin read is visible rather than hidden.

That last detail is the reason to load a filing into your own analysis rather than just reading the names. Pabrai's biggest position at 43% of the book scores 6 out of 10 here and carries an overvalued signal. His smallest scores 9 and reads as undervalued. Neither of those readings is a verdict on Pabrai, who bought at prices we no longer see and understands the coal cycle far better than a screen does. They are a prompt. Where the machine disagrees with a proven investor, one of them is missing something, and finding out which is the most productive research you can do.

The mechanics are ordinary. Every holding is matched to a company in the database, run through the same composite score, valuation signal and quality metrics as anything on your own board, and priced daily at the close. Return since filing is computed on the priced basis, summing current values against filed values across holdings we can price, rather than pretending the unpriceable lines are worth zero. That detail matters: computed naively, an options-heavy book prints a catastrophic loss that never happened.

Is cloning a super-investor a good strategy?

It is a good sourcing strategy and a bad substitute for judgment. The filing tells you a skilled investor bought something, at a price you can no longer get, for reasons you do not know, with a time horizon and a risk tolerance that are not yours. It cannot tell you why they bought, what would make them sell, or how the position fits the rest of their book. Use it to shorten the list of things worth studying. Then do the work: understand the business, form your own view of value, and buy only if you have the conviction to hold it through a decline.

The failure mode is easy to picture. You copy a position because a great investor owns it, the price falls 30%, and you have nothing to hold on to. You do not know why they bought, so you cannot tell whether the fall invalidates the thesis or improves the opportunity. You did not choose the size, so it may be far larger or smaller than your conviction warrants. At that point you are not cloning an investor, you are borrowing their courage, and it does not transfer.

Used properly, the filing does one job well: it shortens the list. Out of thousands of listed companies, here are the few dozen that people with real records are willing to own right now. That is a better starting universe than a stock screen, because a screen selects on numbers and a filing selects on judgment. Then the ordinary work begins, and it is the same work as always: understand the business, decide what it is worth, and buy with a margin of safety if the price allows.

How current are 13F holdings?

They are always stale, by design. Up to 45 days pass between the quarter end and the filing, and the manager may have sold before you read it. This matters least for the investors worth cloning, because their holding periods are measured in years rather than weeks. It matters most for anyone trading around catalysts. If a manager turns their book over every quarter, their 13F is a historical document rather than an idea list, which is one reason the nine we track are weighted toward long-horizon owners.

Signals get stronger when several of them agree. A name that appears across two or three of these books, bought independently by investors with different styles, is a different kind of signal than a single position. Alphabet is currently held by Buffett, Li Lu and Klarman. That does not make it cheap, and it does not make it right. It makes it worth a morning of your attention before you look at anything else.

Which brings the argument back to where it started. The reason to follow holdings rather than opinions is not that great investors are always right. It is that their holdings are the only part of their thinking that has been priced, sized, dated and paid for. Everything else they say is free, and free things are worth checking. General investment analysis, not personalised advice.