Given Warren Buffett’s long-term record of outperformance via Berkshire Hathaway, rational investors should consider following his lead as the the company discloses its holdings. Why would the market not immediately discount his moves as announced? In their July 2010 paper entitled “Overconfidence, Under-Reaction, and Warren Buffett’s Investments”, John Hughes, Jing Liu and Mingshan Zhang investigate how other experts/large traders contribute to market underreaction to Berkshire Hathaway’s moves. Using return, analyst recommendation, insider trading and institutional holdings data for publicly traded stocks listed in Berkshire Hathaway’s quarterly SEC Form 13F filings during 1980-2006 (2,140 quarter-stock observations), they find that:
- Berkshire Hathaway tends to invest in large firms with low book-to-market ratios and large accounting accruals, while avoiding firms with high asset growth and poor past returns.
- Holdings are concentrated and tilted toward banking, business services, insurance and publishing. The number of stocks held ranges from 5 to 95. The average numbers of stocks held during the 1980s, 1990s and 2000s are 22, 12 and 33, respectively.
- The median holding period is one year, with approximately 20% (30%) of stocks held for more than two years (less than six months).
- Over the 1980-2006 sample period, Berkshire Hathaway’s annualized abnormal return from stock holdings (adjusted for market, size, book-to-market and momentum factors) is 7.2%. These returns are substantially independent of those for well-known pricing anomalies, suggesting that Warren Buffett has unique insights regarding future returns.
- A value-weighted (equal-weighted) portfolio that mimics Berkshire Hathaway’s holdings, reformed quarterly based on company filings, generates an annualized abnormal return of 6% (6.6%) over the entire sample period.
- When Berkshire Hathaway announces an increase in a stock position, the average market-adjusted return for the stock is 0.7% to 0.9% over the next five days and two weeks, respectively. The immediate price reaction therefore tends to be very incomplete.
- Net stock sales by insiders (officers, directors, and major stockholders) of companies in which Berkshire Hathaway has a position tend to decrease when Berkshire Hathaway increases its positions, indicating shared private information.
- Contrarily, the behavior of analysts and fund managers indicates overconfidence by disagreeing with Warren Buffett:
- Analyst recommendations are somewhat lower for Berkshire Hathaway’s holdings than for the S&P 500, and analysts tend to downgrade these recommendations following increases in Berkshire Hathaway’s holdings.
- Funds, even those in the top fifth of past performance, appear to take the other side of Berkshire Hathaway’s trades by selling when Berkshire Hathaway is buying.
In summary, evidence indicates that investors can capture a large portion of Berkshire Hathaway’s long-term outperformance by mimicking holdings described in the company’s SEC disclosures, because many other large traders do not.
Frequently Asked Questions
What is a Berkshire Hathaway 13F mimicking strategy?
A Berkshire Hathaway 13F mimicking strategy is an approach that copies the stock holdings Berkshire Hathaway discloses in its quarterly SEC Form 13F filings. The portfolio is reformed each quarter based on those filings, and can be value-weighted or equal-weighted. The study covered 1980-2006 and 2,140 quarter-stock observations.
Does copying Warren Buffett's stock picks beat the market?
Copying Warren Buffett's stock picks via Berkshire Hathaway's disclosed holdings generated abnormal returns in the 1980-2006 sample. A value-weighted mimicking portfolio earned 6.0% annualized and an equal-weighted portfolio earned 6.6%, adjusted for market, size, book-to-market and momentum factors. The evidence indicates investors can capture a large portion of Berkshire's outperformance because many other large traders do not follow it.
How do researchers test whether following Berkshire Hathaway's holdings generates abnormal returns?
Researchers tested Berkshire Hathaway's holdings using return, analyst recommendation, insider trading and institutional holdings data for stocks listed in its quarterly Form 13F filings during 1980-2006. They measured abnormal returns adjusted for market, size, book-to-market and momentum factors, and built value-weighted and equal-weighted mimicking portfolios reformed quarterly. They also measured average market-adjusted stock returns after Berkshire announced position increases, finding 0.7% over five days and 0.9% over two weeks.
How do analysts, insiders and fund managers react differently when Warren Buffett buys a stock?
Insiders, analysts and fund managers react differently when Berkshire Hathaway increases a stock position. Insider net stock sales tend to decrease, indicating shared private information. Analysts tend to downgrade their recommendations, and funds, even those in the top fifth of past performance, tend to sell when Berkshire is buying, which indicates overconfidence.
What are the limitations of copying Berkshire Hathaway's disclosed stock holdings?
Copying Berkshire Hathaway's disclosed holdings relies on filings that reveal positions only after the fact, and the immediate price reaction to announced increases is very incomplete rather than absent (0.7% to 0.9% over five days to two weeks). Results come from a limited 1980-2006 sample of 2,140 quarter-stock observations with concentrated holdings that ranged from 5 to 95 stocks. The portfolio also tilts toward large, low book-to-market firms in sectors such as banking, insurance, business services and publishing.