David R. Snyder, CFA
4/28/25
Warren Buffett gained control over Berkshire Hathaway (BRKA 797,750) in 1965 and has managed the insurance company and its investments for the last 60 years (1965-2024). The average annual compound return of BRK during that time period was 19.9% vs. 10.4% for the total return of the S&P 500 (for this paper when I refer to the returns of the S&P 500 it includes dividends reinvested). He is widely regarded as the greatest long term investor during that time period.
Investment returns are measured relative to risk, with the Sharpe ratio being the gold standard for risk adjusted returns. Despite taking on more risk (24% volatility vs.15% for the S&P 500) BRK’s Sharpe ratio for the 1977-2016 time period was .79 vs .49 for the S&P 500, the most appropriate benchmark. That is significant outperformance especially considering that 97% of money managers have Sharpe ratios below the S&P 500 Sharpe ratio for the last 50 years.
In 2018 the Financial Analysts Journal (FAJ) published an excellent study, “Buffett’s Alpha” that was an in depth analysis of the factors contributing to BRK’s outperformance during the period from October of 1976 through March of 2017. When they break down the attribution of BRK’s outperformance, the numbers are not nearly as flattering. Because BRK is an insurance company that collects premiums and then pays out claims years later, the company has in effect continuous loans and can make money on the float. The average annual cost of BRK’s insurance float has been only 1.72%, 3 percentage point below the average T-bill rate. BRK can make significant money on the float as its hurdle rate is only 1.72% and its investments returns have far exceeded that rate. They estimated Buffett’s average leverage to be 1.7 to 1.
BRK’s average excess return over Treasury Bills during the 1976-2017 period was 18.6% vs. 7.5% for the S&P 500. However adjusting for leverage leaves only an excess return of 11.8% vs. 7.5% for the S&P 500 and the Sharpe ratio drops from .79 to an estimated (my estimates based on data) of .64. The leverage increased risk but was more than offset by excess returns. The information ratio drops from .64 to about .47 without the leverage. These are all still fantastic risk-adjusted returns and would be in the 99th to 100th percentile for all funds alive in 1976 and 2017. If all fund managers over varying long term periods are included, BRK would still be in the 97th to 100th percentile on risk-adjusted returns.
BRK maintained a market beta of only .69 during this time period yet its volatility was over 50% higher than the S&P 500 (23.5% vs. 15.3%). This is because of leverage and high idiosyncratic risk that easily can be lowered by diversifying investments. Buffet has paid a price for concentrated portfolios when performance is measured in risk-adjusted returns. The more concentrated the portfolio the higher the returns required to achieve higher risk-adjusted returns. Risk-adjusted returns are not just an academic exercise. Risk affects investors liquidity (much less money if had to withdraw during downturns), and emotional well-being. Risk has consequences. Intuitively there should be higher required returns to compensate for risk. From June of 1998 through February of 2000, BRK lost 44% of its market value while the S&P 500 rose 32%. In 1974 BRK lost 46% of its value vs. 26% for the S&P 500. The next year BRK gained only 3% vs. 37% for the S&P 500. These kind of negative returns and extreme underperformance can cause very high anxiety even among the most loyal Buffett investors.
Both BRK’s public and private holdings returns have exceeded the S&P 500 but the public holdings have done far better. BRK’s private holdings have increased from less than 20% of the investment portfolio in the early 1980’s to more than 78% in 2017 according to the authors. The average private holdings have been 65%. Without leverage, BRK’s private holdings average annual excess return was only 9.3% vs. 7.5% for the S&P 500 from 1984-2017. In fact the Sharpe ratio for the unlevered private holdings was .45, below the .49 of the S&P 500. Negative alpha. In contrast his public holdings during the same time period produced 12.0% average annual excess returns vs. 7.5% for the S&P 500 without leverage. Thus his change to prioritizing private holdings was detrimental to shareholder value.
The authors determined the annual value of BRK’s private holdings by subtracting the market values of BRK’s public equity holdings and cash from BRK’s stock valuation. This valuation method assumes there is no premium or discount being assigned to the private companies or to the total value of BRK’s public and private holdings. Because of Buffet’s strong reputation, there is most likely a Buffett premium imbedded in BRK’s stock price which would be revealed in a higher than warranted valuation of the private companies using the authors valuation method. Thus the annualized returns of the private companies cited in the study are most likely overstated.
Most of BRK’s unlevered public and private companies outperformance can be explained by three factors: low beta and low volatility, value and low price-to-book, and high quality profitable, stable and growing earnings with high payout ratios. BRK’s unlevered portfolio of publicly traded stocks from 4/1980-3/2017 only had an excess annualized return of 0.3% when adjusted for these factors. Its unlevered private holdings during this time period had an excess annualized return of 3.5%, but this is statistically insignificant. Interesting that BRK’s annualized excess return was only reduced from 13.4% to 5.4% (still statistically insignificant) after adjusting for these factors, above the combined annualized excess returns of BRK’s public and private holdings during that time period of about 2%. Part of this is explained by leverage and possibly other factors that weren’t part of the regression. The total 5.4% excess return could also have been explained in part by the Buffet premium which would have resulted in overstated private portfolio returns.
Almost all of BRK’s investment outperformance can be attributed to the 1965-1995 time period. Over the last 30 calendar years (ending in December 2024) BRK’s public and private investments have only outperformed the S&P 500 by 1.5% annualized, which is statistically insignificant (if he charged an annual management fee of 1% similar to his peers managing money, he would basically be even with S&P 500). Although volatility was reduced somewhat vs. the first 30 years (probably due to more diversification), BRK’s risk-adjusted investment returns were below the S&P 500 for the last 30 years. BRK’s investments have underperformed the S&P 500 for the last 10 and 15 years, and matched the S&P 500 for the last 20 years. On a risk-adjusted basis the relative performance is even worse.
The lackluster performance over the last 30 years is most likely due somewhat to the larger size of BRK as he invested in a higher percentage of smaller companies in his first 30 years. However small and mid caps have only modestly outperformed large caps over the last 20 and 30 years. In defense of Buffett, even if small caps didn’t significantly outperform over last 30 years, there were more inefficiencies within the small cap sector than large cap that he couldn’t exploit due to BRK’s large size.
BRK’s avoidance of technology has also hindered its performance as tech has significantly outperformed the S&P 500 over the last 30 years even adjusted for risk. There are also fewer monopolies in the last 30 years than there were from the prior 30 years. Buffet loved to invest in monopolies such as newspapers (Washington Post was a home run) and broadcasting (Capital Cities Broadcasting was a spectacular investment). Those industries have dramatically increased competition today. The only monopolies today are in technology and communication services, sectors Buffet has avoided except for monopolist Apple (AAPL 221).
His first major investment in technology was IBM (IBM 236) which significantly underperformed during his seven year holding period. His investments in AAPL from 2016-18 have reaped phenomenal gains and have saved BRK from the underperformance of his other holdings. He allowed AAPL to become over 50% of its publicly traded portfolio and 23% of its market cap which is irresponsible in my opinion as he knows many investors own BRK as a fund and not a stock. No fund should have a quarter of its portfolio in one stock.
Never understood why he never owned Accenture (ACN 293) as this is the perfect stock to own to gain exposure to tech without the risk. They are a high value added consultant and business processing specialist who helps clients with their technology needs. Very little tech obsolescence risk. ACN has produced annual compound earnings growth of over 15% with only one down year since they went public in 2001. Very predictable earnings. Their returns on assets and equity have averaged over 30%. The stock has increased more than twentyfold since the IPO.
Buffet’s refusal to sell Coca-Cola (KO 72) after it became extremely overvalued also significantly affected the last 30 years performance. KO has dramatically underperformed the S&P 500 since it peaked in 1998 when it was BRK’s biggest holding (10% of its portfolio). Its relative P/E was more than double that of the S&P 500 in 1998 and its earnings growth through the 1990’s after Buffet purchased it was enhanced by accounting gimmicks. KO was able to compound annual earnings growth at close to 20% through the aid of annual selling of a portion of its bottling operations, which should have been classified as one time gains. As time progressed there were fewer bottling operations to sell, and fewer earnings assets remained. As a result earnings growth decelerated significantly. Don’t know why Buffett didn’t recognize these accounting items.
After increasing tenfold over 10 years from Buffett’s initial purchase in 1988, KO has only increased 62% over the last 26 years. Thus over the entire time period BRK has owned KO (1988-2024), KO has only slightly outperformed the S&P 500 even with dividends reinvested. Buffet gave back almost the entire outperformance over the first 10 years of ownership. And he says he will never sell KO.
Buffet has made many other dud investments over the last 25 years. Kraft Heinz (KHC 29) has been a disaster since he bought it although he made a lot of money on Heinz before the merger with Kraft. He had unsuccessful investments in Verizon (VZ 42), General Motors (GM 47), UPS (UPS 97), Walmart (WMT 95) and Kroger (KR 70). Procter and Gamble (PG 162) was just a slightly better than average performer from 2005 until he sold it in 2023, although Buffett did well with Gillette before it was sold to PG in 2005. Conoco-Phillips COP 92) was an absolute loser when he owned it from 2008-2012. Paramount Global (PARA 12) was a big loser for Buffett although a small position. Wells Fargo (WFC 69) was a strong performer after Buffett purchased it in 1990, but underperformed significantly since 2007. Buffett sold most of its stake in 2021 and WFC almost tripled since his last sales.
Buffett’s bank investments such as US Bancorp (USB 40), and PNC (PNC 160) were underperformers from 2006-2022. The Bank of America (BAC 40) 5% preferred stock with long term common stock warrants to buy common stock at $7.15 (no premium) purchase in 2011 was a big outperformer but Buffett would have never agreed to the deal if it was only for common stock at $7.14. It was a no brainer special deal for Buffett. Sanofi (SNY 53) was a significant underperformer while he owned it. His investment in Tesco PLC (TSCO 357) in 2012 resulted in a substantial realized loss four years later. His purchase of Energy Future Holdings (coal generator) high yield bonds in 2007 lost most of its value.
His recent investments in Chevron (CVX 140) and Occidental Petroleum (OXY 40) have been significant underperformers with significant unrealized losses. He purchased a lot of the shares after oil stocks had significantly outperformed in 2021-22.
His private holdings since the early 1990’s have been even worse. Buffett admitted that he significantly overpaid for General Reinsurance and Lubrizol. Buffett’s purchase of Dexter Shoes in 1993 resulted in almost a total loss of his investment. Buffett’s biggest private investment of $32 billion in Precision Castparts in 2016 was followed by a $10 billion write-down a few years later and it is still not performing well. Buffett’s purchase of truck-stop operator Pilot Cos. in three stages from 2017 to 2024 for $13 billion has been a big disappointment with pretax earnings declining 42% in 2024.
Buffett has made some successful private investments to help offset some of the poor performers. His acquisitions of railroads and utilities over the last 20 years through his Energy unit have been successful. He also made some timely public investments in Chinese EV maker BYD as well as a few Japanese trading companies but they are small positions. GEICO has performed well overall although it has had periods of poor results and executed a strong turnaround over the last two years after a period of losses. However Progressive (PRG 284) has been taking market share and been more profitable than GEICO over the last decade.
BRK’s relative poor performance over the last 30 years can’t be explained by the factors that correlated with his prior 30 year outperformance going out of favor. The authors of the FAJ article created a systematic Buffett style portfolio (buying safe, low beta, low volatility, low price -to -book, high quality stocks) for the relevant time period (1977-2017) which outperformed BRK substantially from 1995-2017 even though there was still high correlation (although less correlated). Thus BRK actually underperformed its factor exposure, adding no value with his individual stock selection during the 1995-2017 period.
Buffett had the advantage of not charging a management fee when comparing his record to other money managers. He also has the luxury of having new money come in every month to his company from insurance premiums. Thus when the stock market is in a bear market he has access to new cash to invest in contrast to other money managers who are fully invested and have to just hope their current holdings rebound. Most mutual funds and money managers have withdraws during bear markets, which is even worse.
BRK’s claims he doesn’t attempt to time the stock market, but his cash holdings tell a different story. His cash position varies significantly and when his cash levels are high he is often quoted as saying that he can’t find any good values in the stock market. That is veiled market timing. His success with varying his cash positions has been mixed over the years and really has not added much if any value. During the Covid bear market in 2020, Buffett didn’t invest any of his $130 billion in cash near the bottom, missing one of the greatest buying opportunities of the last 30 years. He even sold his airline holdings near the Covid lows. Buffet did make one successful market timing call. In October of 2008 during the financial crisis he sold long term puts on the S&P 500 (which is a bet that the S&P 500 would rise over the designated time period) when it fell below 1000.
Buffett also made a great generational call on housing when he said that investors should buy as many houses as possible in 2012 as housing prices were bottoming. Buffett is not a big fan of real estate so this was a bold contrarian call.
Buffett also has been able to negotiate extraordinary investment deals with public companies who were in need of capital such as GE (GE 201) and BAC preferred with warrants during the financial crisis. Every fund manager in the world would have taken the BAC deal that Buffett was offered. Other fund managers don’t have the same opportunities to buy these special deals.
Buffett has been a very astute purchaser of BRK’s stock, unlike the rest of corporate America that just mindlessly buys back stock regardless of valuations. This has limited volatility on the downside, another advantage over his peer fund managers who don’t have the same opportunity.
BRK has also been a big beneficiary of one of the strongest ever property and casualty pricing cycles over the last couple of years due to its high insurance exposure. Thus we are evaluating BRK equity returns at the top of a pricing cycle, with BRK trading at its highest price to book value since before the financial crisis. This would bias the annualized returns to the upside.
Most of the investment returns discussed above don’t adjust for taxes. Because Buffett is a long term investor with little turnover, his tax-adjusted returns would be better relative to his money management peers for taxable accounts.
In summary all of BRK’s outperformance over the last 60 years can be explained by his use of leverage (1.7 to 1 on average) and three factors: low beta low volatility stocks, value stocks with low price-to-book ratios, and high quality stable profitable stocks with high payout ratios. The authors created systematic Buffett style portfolios based on these three factors and produced very highly correlated returns with unlevered BRK portfolios and even outperformed BRK for the last 30 years. Thus Buffett didn’t produce any statistical significant alpha selecting individual stocks.
In defense of Buffett it is very difficult to be a successful long term investor, especially over the last 30 years when changes in the economy and sectors seem to happen with more speed.
Buffett’s returns over the last 30 years have roughly matched the S&P 500 adjusted for a 1% management fee despite having high leverage and BRK’s Sharpe ratio (risk-adjusted returns) during this period was below the S&P 500 Sharpe ratio. His unlevered public and private portfolios underperformed the S&P 500 for the last 30 years, especially his private holdings.
Buffett didn’t discover that buying value, low beta and quality stocks created alpha. It has been well documented in research papers and well known in the financial community since Buffett began managing BRK in 1965. So why didn’t other money managers follow this same strategy? Remember theoretically an investor could have matched BRK’s unlevered excess returns by just following a statistical model with no need for doing in depth individual stock analysis and selection. In layman’s terms an investor could have duplicated BRK’s returns by just screening for low beta, low price to book, high quality stocks with steady earnings growth without knowing anything about the companies and then add 70% leverage (obviously it would have to be done within an insurance company vehicle to match BRK’s low cost leverage). Could it be that his peers just didn’t have the patience and discipline to follow such a model?
I have also always been skeptical of retroactive vs. real time models. There is always the possibility of bias with retroactive analysis. But the FAJ is a very highly regarded publication with some of the most respected researchers in the finance industry. So I am not going to question their methodology.
The other issue is if BRK’s success was mostly due to the unique advantages of operating as an insurance company as contended (especially to obtain very low cost leverage), then why didn’t more money managers use the insurance vehicle to manage their investments? Most insurance companies have been very conservative with their investments and don’t allocate nearly as much of their investments to equities as BRK. It is necessary for insurance companies to maintain statutory capital at minimum regulatory levels, thus very important to have profitable underwriting margins to be able to take more risk on investments.
Buffett was able to do very well on the insurance side to create more buying power for his investments. Thus Buffet’s real success may be his ability to manage an insurance and/or reinsurance companies.
Peter Lynch is another legendary investor who I believe is the greatest ever long only money manager. From May of 1977 through May of 1990, Lynch’s Magellan Fund returned 29.06% annualized vs. 15.52% for the S&P 500, a 13.54% annualized outperformance. The value of his fund increased twenty-nine fold vs. fourteenfold for BRK’s unlevered stock and private holdings portfolio during the same time period. Using the same regression analysis with the same factors as was used for the FAJ study on BRK’s performance reveals that 7.94% of the 13.54% annualized outperformance by Lynch was not explained by these factors vs. virtually all of BRK’s outperformance (other than leverage). The 7.94% annualized alpha is statistically significant superior stock selection. And of course Lynch didn’t have the benefit of BRK’s leverage to enhance returns even more.
Interesting that Lynch didn’t benefit from the same factors that benefitted BRK. Lynch’s fund was basically neutral with regard to quality and value (low price to book) that loaded with BRK. Also Lynch benefitted moderately from higher beta (1.12 beta) while BRK was low beta loaded. Lynch was somewhat momentum loaded unlike BRK. Contrary to conventional wisdom, Lynch was neutral with regard to market cap (not small cap loaded) but need to segment the time periods when small cap outperformed to get a better reading. It also shows that there is more than one way to achieve outperformance, although Lynch’s 7.94% annualized alpha not explained by factors is assumed to be from individual stock selection unlike BRK’s outperformance.
Lynch’s Sharpe ratio was 0.98 with an information ratio of 1.78, dramatically higher than BRK’s unlevered Sharpe ratio of 0.64 and information ratio of 0.47, although BRK’s ratios would have been moderately higher but still well below Lynch’s if measured for the same period Lynch managed his fund. Part of the reason for the higher risk-adjusted returns was that Lynch eliminated idiosyncratic risk by diversifying in as many as 1200 stocks.
Many people have tried to discredit Lynch’s remarkable performance by arguing that the stock market wasn’t as efficient back in the 1970’s and 1980’s especially in small caps. The argument is that Lynch and other money managers had access to information by talking to companies executives that wasn’t made public at the same time. Since 2002 SEC rules and regulations have tightened release of material information to investors. Lynch would have benefitted more than Buffett from this loophole because Buffett didn’t talk as much with management teams of his public companies in which he had equity stakes.
While there may be possibly some truth with disclosures, it is greatly exaggerated and not as commonplace as believed. I was managing money as early as the mid to late 1980’s and spoke to a lot of small companies and they were very strict with disclosures. Also the fact is that there were many heavyweight fund companies and money managers with the same resources back in the 1970’s and 1980’s who had the same exact opportunities as Lynch. Yet Lynch dramatically outperformed all of his peers.
The other argument that there was less analyst coverage and less easily accessible computerized public information of small caps would have benefited Buffett and Lynch equally. Yes there was more inefficient pricing of small caps to exploit through research even without speaking to management, but again Lynch significantly outperformed his peers and BRK with the same opportunities.
Studies show that active management outperformance has been reduced in half from 1967 through 2012. So certainly Lynch would not perform as well today as he did in the late 1970’s and 1980’s. Neither would Buffett and he is sure proving that over the last 30 years. But that would only account for a portion of Lynch’s alpha.
The truth is that Lynch just outworked his peers and was open to all opportunities. He invested in all categories and was mostly factor neutral. He was just as good at evaluating a turnaround as a growth company and read the financials as well as talked to every company he owned. His written deep analyses of why he invested in Chrysler and Fannie Mae when they were distressed were brilliant. All of the information he cited with these two companies was available to any money manager by way of public documents. He successfully invested in manufactured home stocks in the late 1970’s when no other fund manager wanted to touch them because they were perceived as risky and unsightly.
The most remarkable feature of Lynch’s money management skills is that he was able to achieve such incredible results while owning a very large amount of stocks. It takes incredible skill to produce such significant outperformance while diversifying as much as Lynch did during his career, although his top 50 holdings were more concentrated. The exact opposite of Buffett’s strategy. This is a big advantage for Lynch over Buffett as Lynch was able to achieve higher risk adjusted returns for the same outperformance by eliminating idiosyncratic risk.
Lynch was the greatest money manager by a wide margin for long only accounts. The facts and analysis speak for themselves. The only knock is that he only managed money for 14 years, although that period included a broad range of investment and economic environments.