Warren Buffett: The Greatest Factor Investor of All Time?
Saints and Star Investors
The Vatican waits at least five years after a person dies before it begins considering whether they are worthy of sainthood. The canonization process can take decades, even centuries, to complete as the church meticulously investigates the lives of candidates for saintly virtues, associated miracles, and other indicators of divine influence.
The investment industry would be well advised to follow a similarly rigorous process before elevating particular investors to superstar status. Many investors, Bill Ackman, David Einhorn, Bill Miller, and Neil Woodford, among them, have prematurely become icons before tumbling from the heights. Many more have risen, fallen, and already been forgotten.
Analyses of these star investors’ track records often reveal exposure to certain styles like Value or Growth. Since factors are just as cyclical as equity markets, track records tend to rise and fall with factor performance.
Warren Buffett is among those few investors who have achieved iconic status and never relinquished it. The pedestal to which he has ascended has barely budged over the last several decades. Although investors have experienced perfectly normal periods of underperformance when investing in Berkshire Hathaway, Buffett has avoided the major pitfalls.
So from a factor perspective, what has driven Berkshire Hathaway’s outperformance?
Berkshire Hathaway vs. The S&P 500
While Buffett has controlled Berkshire Hathaway (BRK.A) since 1964, his track record over roughly the last three decades offers a robust enough sample size to draw some conclusions about what has contributed to his success. His outperformance relative to the S&P 500 has been especially significant since 1991 and created immense value for BRK.A shareholders, albeit with higher volatility due to implied leverage.
Though there were some periods of underperformance, during the Tech Bubble in 2000, for example, these often turned out to be based on wise investment decisions that translated into superior returns later on.
Berkshire Hathaway vs. The S&P 500
Source: FactorResearch
Berkshire Hathaway: Factor Exposure Analysis
We next conduct a factor exposure analysis of Berkshire Hathaway by measuring the betas to common equity factors through a regression analysis. The factor betas are time-varying and the analysis is reminiscent of a modern art piece. It does little to answer the question of whether Buffett’s success is the result of factor bets.
Berkshire Hathaway Factor Betas
Source: FactorResearch
However, by averaging the factor betas over the last three decades, a clear picture emerges: BRK.A had positive exposure to the Value, Size, Low Volatility, and Quality factors as well as negative exposure to Growth and Dividend Yield.
Naturally, this reflects Buffett’s well-known preference for cheap and high-quality firms over high-growth technology companies. Despite recent investments in Apple and other tech stocks, such additions are more akin to existing blue-chip holdings as Coca-Cola rather than pure growth companies like Netflix.
Average Berkshire Hathaway Factor Betas, 1991–2018
Source: FactorResearch
Replicating Berkshire Hathaway Performance
Because of this better understanding of Berkshire Hathaway’s factor exposure, we can apply these data points to replicate BRK.A with a factor-mimicking portfolio. We measure the factor exposure on a monthly basis and then allocate to long-short factor portfolios based on BRK.A’s exposure for the next month.
The factor-mimicking portfolio follows the trend of Berkshire Hathaway relatively closely and therefore achieves a similar outperformance relative to the S&P 500. Of course, this portfolio is based on historical data that is reflected in BRK.A’s stock price and is lagging in nature, but Berkshire Hathaway’s portfolio of investments does not change too frequently.
Replicating Berkshire Hathaway via a Factor-Mimicking Portfolio
Source: FactorResearch
Berkshire Hathaway’s Outperformance: Alpha vs. Factor Returns
Why replicate Berkshire Hathaway performance through factors rather than gaining exposure to Buffett’s expertise by simply buying the stock? Factor-mimicking portfolios are typically used to gain access to something otherwise inaccessible. Before 1996, most retail investors could not afford Berkshire Hathaway stock, which was selling at more than $30,000 a share. Buffett solved that problem by issuing B-class shares with a lower nominal value.
The factor-mimicking portfolio helps investors distinguish between alpha and factor returns given Berkshire Hathaway’s significant exposure to common equity factors. The analysis demonstrates that most of the outperformance relative to the S&P 500 is explained by factor exposure, with little alpha generated over time.
Berkshire Hathaway fans might be surprised, even offended, by our suggestion of a lack of alpha. But we could also contend that Buffett may be the best factor investor ever based on the wealth he created. While Buffett is unlikely to characterize his investment style in these terms, he has demonstrated unprecedented skill in selecting factors, constructing a multi-factor portfolio, and adapting the factor mix over time.
Berkshire Hathaway’s Outperformance: Alpha vs. Factor Returns
Source: FactorResearch
Further Thoughts
More than $100 billion annually has been invested in factor-focused products in recent years. But selecting factors and combining them into multi-factor portfolios is a challenging endeavor.
So Berkshire Hathaway might be an alternative to multi-factor products. After all, the company is supported by a fund manager with a decades-long track record of skillfully managing factor exposure. The management fee is competitive compared to smart beta exchange-traded funds (ETFs) and Buffett’s stake in the company ensures that his interests are aligned with those of his investors.
Investors just have to hope that the contrarian diet of Cherry Coke and burgers will sustain the fund manager for a few more years. Whatever happens, that diet has already secured Buffett’s place in the canon of great investors.
For more insight into how Warren Buffett does it, see “Buffett’s Alpha,” by Andrea Frazzini, David Kabiller, CFA, and Lasse Heje Pedersen, from the Financial Analysts Journal. For more insights from Nicolas Rabener and the Factor Research team, sign up for their email newsletter.
If you liked this post, don’t forget to subscribe to the Enterprising Investor.
All posts are the opinion of the author. As such, they should not be construed as investment advice, nor do the opinions expressed necessarily reflect the views of CFA Institute or the author’s employer.
Image credit: ©Getty Images/Bill Pugliano/Stringer
Professional Learning for CFA Institute Members
CFA Institute members are empowered to self-determine and self-report professional learning (PL) credits earned, including content on Enterprising Investor. Members can record credits easily using their online PL tracker.
Very interesting. Is there a way to add leverage as a factor? As the “Buffett’s Alpha” paper notes, Buffett used about 1.7x leverage. I actually think it’s materially higher in the earlier BRK years, but over time it has come down. But that’s a big difference if you buy KO that returns about 10% a year versus holding KO with leverage and getting 17% a year (not even counting the negative cost of that 1.7x leverage in the form of underwriting profits, which would boost the return even higher). The way I think about it – with KO he’s doing a 50-year LBO at about 2x leverage on a massive company that’s way too big for any PE shop to take out. So he’s the only marginal buyer with leverage who finds KO attractive, even at higher valuations. Even the individual investor or hedge fund will get margin-called and pay ridiculously high interest to hold KO using broker leverage, whereas he won’t. Thus, KO and other low beta stocks are consistently underbid.
Hi NJC, thanks for your thoughts. Other researchers like AQR have done similar analysis and also concluded that leverage played a large role in BRK’s outperformance. We could add leverage as a factor directly, but indirectly capture it via the factor-mimicking portfolio as the factor weights can be more than 100%. And yes, Warren Buffett has been harvesting the low volatility factor via investments like KO, which requires leverage, way before it became popular. He is one of a kind.
This is an interesting article; enjoyed reading it. Perhaps I misunderstood, but it seems to me that the author may have it backwards with respect to alpha. The factor profile is a RESULT of the investment selection process followed by Buffett (and his managers) as opposed to these factors driving his investment process. I do not think that is a subtle distinction; rather a critical difference in typical factor based portfolios versus bottom up investing processes as it relates to the generation of returns above a benchmark. All that said, investing to beat a benchmark is a creation of the investment industry in the first place and while a convenient mechanism, it ought to be irrelevant compared to seeking generally consistent long-term absolute returns.
Hi Brett, thanks for your perspective. I think the question if Warren Buffett generated alpha depends on the definition. We took the academic definition of alpha, which mans unexplained returns, but we can explain his outperformance via exposure to certain factors like Value and Low Volatility.
However, we would naturally agree that he is highly skilled in portfolio construction. The folks at AQR call that craftmanship alpha, which probably a good way of describing Warren Buffet’s skill and alternative alpha definition.
Nice article with deep factor analysis of Buffett’s alpha. What I do not understand at all: Why is there “a lack of alpha”? Did it not clearly show up by these sophisticated factor evolutions over time, expressing real mastery? And if there is a lack, why “selecting factors and combining them into multi-factor portfolios is a challenging endeavor”? If there were “a lack of alpha”, then it would have been plain passive investing, what Buffett did!?
Or are these contradictions another “feature not a bug” as is the case with the whole factor theory, which one of the most successful HF managers and deep thinker David Harding calls “hardly the height of good scientific practice”, because it is just “retrospectively introducing extra variables and fitting the data to them” as has obviously been done here as well? See: https://www.winton.com/davids-views/July-2018/wintons-david-harding-on-turning-away-from-trend-following-risk-net
Hi Norbert, thanks for your comments. We applied the academic definition of alpha representing unexplained returns, but can actually explain most of BRK’s outperformance by exposure to factors like Value or Low Volatility. These have been known for decades, although Warren Buffett was one of the early adopters.
Our perspective is that Warren Buffett is highly skilled in portfolio construction and harvesting factor returns, which might be called craftsmanship alpha. We’re not critical of his success, just provide a different perspective.
I think the article on Winton makes a case for factor investing as they are moving away from trend following towards risk premia strategies.
Hi Nicolas,
Thank you for your comment. Yes, craftsmanship alpha is correct. But extraordinary craftsmanship or skill is the essential base for every alpha, which you cannot get from the market without. If the outcome can be explained by factors, that is fine and interesting, but does not change the individual alpha performance of Buffett or other extraordinary investor, e.g. quant investor Dunn, who actually surpassed Buffett’s Alpha persistently for decades already:
https://www.markettamer.com/blog/who-is-the-greatest-investor
Thus, even though I am not a particular fan of Buffett, but learned a lot from his basic principles for my own predominantly quant pure alpha/beta investment style, I still cannot understand, why you suggest “a lack of alpha”? Didn’t you impressively demonstrate his high alpha skills by analyzing ex post his really smart dynamic ex ante factor selection skill based on real track record results?
This can never be compared with these pure marketing gags of so called smart beta funds, offered mainly for charging increased fees in order to compensate the shrinking profits from active mutual equity funds!? Why are they also rightly so called “dumb alpha”, in the first place? Is there any smart beta fund with a real track record with a high alpha performance similar to value investor Buffet or quant investor Dunn? I don’t know any. But if Buffet had a “lack of alpha”, mimicking his style and success should not have been such a rare thing, if everybody can pick it up for sure in the ETF offer.
Certainly, Winton makes a case for factor investing, but certainly NOT for so-called Smart Beta/dumb alpha by moving away from trend following towards risk premia strategies, BUT by adapting and optimizing the allocation of all essential investment factors, such as trend, value, momentum…, in the only sensible active way in order to keep producing extraordinary alpha, favorably uncorrelated to broad market beta, in structurally changing market times.
I am not at all apposed to factor investing, if it is done in the right active way as Buffett, Dunn and Winton are doing this, but against the tale of superior passive factor investing, serving for less honorable goals and misguiding more bounded rational investors.
That is what infuriates Harding as well, when he writes in his view, referring to Fama’s flawed later work, dis-improving his great research results in the beginning with valuable simple formulations of the EMH: ” “Seeing your hypothesis falsified and then trying to defend it by retrospectively introducing extra variables and fitting the data to them is hardly the height of good scientific practice.”
(https://www.winton.com/davids-views/July-2018/wintons-david-harding-on-turning-away-from-trend-following-risk-net)
Hi Norbert, thanks for your additional comments and thoughts.
I think we agree that Warren Buffett is highly skilled at generating excess returns, but simply apply slightly different definitions to alpha. For example, if we would run a factor exposure analysis on RenTech’s Medallion fund, then we would unlikely to be able to attribute the returns to common equity factors. However, in the case of BRK we measured structural factor exposure, therefore can explain the returns. Strictly speaking, the outperformance is not unexplained and does not represent alpha. Naturally this doesn’t mean that it’s easy to replicate what Warren Buffett has achieved. If you compare it to Formula One, we know the car, but it doesn’t mean we can drive it like Lewis Hamilton.
We share most of your views on smart beta and have summarized some of these in the following piece, which might be of interest to you:
https://blogs.cfainstitute.org/investor/2019/02/11/smart-beta-broken-by-design/
Did you go and check to and see whether your “factor portfolios” would have had enough capacity to do what Buffett did? One problem with some quantitative analyses is that they assume infinite market depth, zero transaction costs, strategies have no market impact costs, etc.
In short, I think there are a lot of things that you missed.
Hi David, thanks for your thoughtful questions. The analysis used long-short factors for explaining and then replicating the performance of BRK using these as building blocks. The factors are constructed via well-diversified portfolios of liquid US stocks and are not particular capacity constrained. The factor mimicking portfolio doesn’t have a high turnover and we included transaction costs. The objective of the analysis was simply aimed to see if factors explain the outperformance of BRK, which they do.
Hi! Buffet often proposes a different model of stock selection: Predicting the long term earnings, profitability, and compounding potential, assessing the competence of management of a company, generating a minimum likely future price, and buying if there is a significant margin of safety and you’re capable of understanding the business. The theory espouses is that in the long term, investment returns track to the earnings of the business. Turtle Creek?
Certain people, such as Monish Pabrai and Turtle Creek, have been able to replicate this methodology with success, so it seems to be valid empirically. How do you distinguish between the causal validity of the academic factor model (low price = risky, risky = discount) and this approach, which simply assumes markets make enough pricing mistakes? Do you think they are equivalent?
This whole article is the best example I have ever seen of what Charlie Munger says: “To the man with a hammer, everything looks like a nail”.