Copy what works: Pabrai's method and the research behind it

Key ideas

Pick one person who's already winning at what you want to do, and study them before anyone else.
Go deep: learn the principles behind their decisions, not a summary of their results.
Copy the whole system, including the dull daily habits, not just the one flashy insight.
Commit fully. Trying an idea for a week and dropping it isn't copying.
Adapt only after you've mastered the basics, the way an apprentice watches before working alone.
Borrowing from several proven sources and combining them is how a copier ends up original.

Mohnish Pabrai built his investing career by copying Warren Buffett on purpose, from the way Buffett set up his first funds to how much he reads, and he’s said so openly for decades. These are my notes on a video about his approach, plus what I found when I looked for the research behind it: studies of how people and companies learn by copying, why it works, and how to do it well.

How Pabrai copied Buffett

Pabrai was running an IT company when he picked up a Peter Lynch book in London in 1994, and that’s where he first read about Buffett. He’s said in interviews that Buffett’s approach made more sense to him than anything else in the industry, so he set out to learn it rather than improve on it.

What he copied, by his own account and the video’s:

The fund structure. When he started Pabrai Investment Funds in 1999, he modeled them on Buffett’s partnerships from the 1950s.
The habits. Long hours reading and thinking, and working mostly alone.
The schooling. He went to Berkshire Hathaway’s annual meeting year after year and read Buffett’s letters closely.
Other teachers, later. Once he had Buffett down, he added Charlie Munger’s way of thinking and Benjamin Graham’s principles.

In June 2007 he and his friend Guy Spier bid $650,100 at a charity auction for lunch with Buffett, which they had in June 2008. Pabrai called it his guru dakshina, a thank-you to a teacher.

Why copying works

The short version: people mostly show their best moves, so a copier picks up ideas that have already been tested.

Finding	Study	What it found
Copying filters for good ideas	Rendell and others, Science, 2010	In a simulated tournament with 104 entries, the winning learning strategies relied almost entirely on copying others.
Being first rarely keeps you on top	Golder and Tellis, 1993	Of about 500 brands, close to half of market pioneers failed, and the market leaders were often not the first ones in. Its oft-quoted 8% failure rate belongs to early market leaders, a group picked because it succeeded, so it doesn’t measure followers in general. I’ve seen these figures in summaries only.
Copying a skilled investor can match them	Frank and others, Copycat Funds, 2004	Funds that copied the published holdings of other funds earned about the same returns after costs, in a sample of high-fee funds from the 1990s.
Copying is cheaper and faster	Oded Shenkar, Copycats, 2010	Copying costs a company roughly 65% to 75% of what the original innovation cost and takes about 70% of the time.

Others who built on copying

Sam Walton. In his autobiography, Made in America, he says most of what he did was copied, calls Kmart his laboratory, and describes visiting rival stores with a legal pad and a tape recorder. His habit was to look for what competitors did right.
Fast followers in Shenkar’s book. McDonald’s built on what White Castle started, and Honda and Toyota followed Chrysler into minivans.
The video’s example. The narrator describes a friend whose recipe channel grew once she copied the format and topic choices of the biggest channel in her niche, then started experimenting after that.

How to copy well

The research and the examples point the same way:

Pick one model. The best person or company at exactly what you want to do.
Study the system, not the highlights. Shenkar’s warning is that a surface copy usually fails. What carries over is the reasoning and the routine.
Look for what they do right. Walton’s habit, and the opposite of how most people study a competitor.
Commit long enough to learn it. Pabrai went to the meetings for years, not once.
Then adapt and combine. Add a second and third source once the first is second nature.

Keep survivorship bias in mind when you pick a model: the people you can see are the ones it worked for.

Key facts

Fact	Figure	Source
Pabrai first reads about Buffett	1994, in a Peter Lynch book picked up in London	Columbia interview, 2008
Pabrai’s fund, mid-1999 to September 2024	12.9% a year after fees, against 7.8% for the S&P 500	Forbes
Pabrai’s main fund, 2000 to 2018	1,204% in total, against 159% for the S&P 500	William Green
Pabrai’s fund in 2008	Down about two thirds, against 37% for the S&P 500	Pabrai, Cambridge talk, 2025
Charity lunch with Buffett	$650,100 bid in June 2007, lunch in June 2008, for the Glide Foundation	NBC News, CNBC
Buffett’s partnership, 1957 to 1969	About 29% to 31% a year before fees, about 24.5% after	Novel Investor
Berkshire Hathaway, 1965 to 2024	19.9% a year	Berkshire’s 2024 letter
Pabrai’s net worth	No sourced figure; published estimates mostly fall between $150 million and $165 million	Estimate sites only

Sources

Key terms

Cloning: Pabrai's word for deliberately copying another investor's methods and habits, which he does openly.
Guru dakshina: In Hindu tradition, a gift to a teacher when your education is complete. Pabrai used the term for his charity lunch bid.
Social learning: Learning by watching what others do instead of by your own trial and error.
Market pioneer: The first company to sell a product in a new category.
Survivorship bias: Drawing lessons only from the people who succeeded, because the ones who tried the same thing and failed aren't visible.

Quotes from the talk

I'm a shameless copycat.

Mohnish Pabrai, quoted by William Green, quoted in the talk