Mohnish Pabrai: Only about a dozen truly great decisions have driven results over 58 years, averaging one every five years
Mohnish Pabrai analyzes Warren Buffett's capital allocation record at Berkshire Hathaway, pointing out that only about a dozen truly great decisions have driven results over 58 years, averaging one every five years. The key commonality of these decisions is to buy and hold long-term without selling.
Buffett explicitly stated in his 2022 shareholder letter that most capital allocation decisions are mediocre, with satisfactory results coming from about a dozen truly good decisions, combined with the advantages of long-term investors; Pabrai estimates that Berkshire made about 300 significant decisions during this period (acquisitions, stocks, and key hires), with only about 4% truly driving results, and these winners must be continuously held rather than sold.
See's Candies was acquired by Berkshire for $25 million in 1972 (the seller's original asking price was $30 million), with pre-tax profits at the time of about $2 million to $4 million. It has been held for over 50 years, contributing over $2 billion in pre-tax profits to Berkshire, requiring only an additional capital investment of about $32 million, with these cash flows used for acquiring other businesses.
Coca-Cola is one of Berkshire's longest-held public stocks, with a cumulative investment of about $1.3 billion between 1988 and 1990 (then about a quarter of Berkshire's book value), and the core position has been held for nearly 40 years without selling; in recent years, the market value of this position has significantly increased and continues to contribute to growing dividends.
Pabrai refers to this approach of buying great companies and then doing nothing as "waiting for the paint to dry," emphasizing that successful investors need a combination of extreme patience and extreme decisiveness, and after buying, they should engage in other activities rather than frequent trading. He cites Pascal's view that much of human suffering comes from the inability to be quietly alone and inactive.
In market mechanisms, long-term holders (like Berkshire) are the net buyers locking in quality assets, while event-driven traders and short-term funds sell due to an inability to tolerate price stagnation or declines; funds flow from those willing to "watch the paint dry" to those willing to pay high prices, and long-term shareholders of great companies benefit from compounding, while frequent traders are pressured by transaction costs and opportunity costs.
Berkshire continuously obtains reinvestable cash flow through cash cow businesses like See's, used to expand insurance float and for other acquisitions, creating positive feedback; among publicly held equity positions, the long-held core positions contribute the vast majority of excess returns.
Source: Public information
ABAB AI Insight
Mohnish Pabrai started by replicating Warren Buffett and Charlie Munger's partnership structure and value investing approach, aiming to grow his initial $1 million capital to $1 billion over about 30 years, publicly acknowledging that most attempts failed but with limited downside, and the returns from successful cases covered dozens of failures; he has long studied Berkshire's capital allocation error rate, noting that even Buffett himself considers about half of his acquisitions as mistakes, and he almost passed on the See's deal due to a $5 million price difference.
In terms of capital path, Pabrai concentrates funds on companies with high returns, low capital needs, and durable moats, gaining cognitive advantages through "shameless cloning" of Buffett's path rather than pursuing originality; See's excess cash flow has been used by Berkshire for subsequent acquisitions and stock investments (including Coca-Cola), forming resource reallocation, motivated by the asymmetric returns of a few truly great decisions that can cover a large number of mediocre decisions.
In analogy, similar to Buffett's early shift from Ben Graham-style cheap stocks to high-quality brand companies, Pabrai views See's as a key case for unlocking pricing power recognition, while Coca-Cola reflects globally replicable consumer habit moats; current value investing is transitioning from merely undervalued to identifying long-term compounding machines, emphasizing holding rather than frequent rebalancing.
Structurally, it combines capital concentration and pricing power transfer: a few long-term holders lock in ownership of quality assets by not selling, transforming the selling pressure during short-term fluctuations into buying opportunities for themselves, with the mechanism being that the intrinsic value of great companies grows over time while market prices exhibit significant short-term noise, and only capital willing to wait can continue to accumulate rather than being eroded by transaction costs and emotions.
ABAB News · Laws of Cognition
Doing nothing after buying is harder than doing anything.
Wealth = a few correct decisions × long-term holding.
Those who watch the paint dry will eventually own the paint factory.