Intrinsic Valuation of Berkshire Hathaway
Intrinsic Valuation of Berkshire Hathaway
Based on a sum-of-the-parts framework
In order to value Berkshire Hathaway, you need to first get into the head of Warren Buffett. Luckily, Mr. Buffett provides a wealth of clues in his annual reports stretching back multiple decades. Plus there are tons of books written on Warren Buffett and Berkshire Hathaway. Once you see how he views it, you then need to apply your own lens and perspective.
Last March I attempted to tackle the question of intrinsic value. First, there was a lot of reading. I read all of his annual letters going back to 1969 as well as his old partnership letters pre-dating Berkshire Hathaway. I read much of the literature written on Buffett, including the original Lowenstein biography, the new Schroeder one and the Essays of Warren Buffett, which is a good distillation of the raw material provided in the aforementioned annual letters. I subscribed to a number of Buffett related blogs, including Everything Warren Buffett, which does it good job of aggregating anything Buffett related in the media.
Insurance Businesses
Understanding Berkshire Hathaway's insurance business was at first a daunting challenge. I had never really studied the insurance industry in depth. The benefit is that I have a fresh look and Buffett and Munger have always been good at breaking down what seem like complex things into simple concepts that make sense (Munger is a big fan of "reductionism"). (Fairfax Financial has been another good teacher in this area).
There are two things you really need to do well in the insurance business. First, you need to "write" good business. By this, you need to know enough about the risk you are undertaking and price accordingly so that the premiums you receive more than cover 1) the losses that you eventually pay out and 2) the expenses you incur while underwriting. Second, you need to know how to invest all of the “float” that is generated. In the insurance business, you collect premiums first and pay out losses later, sometimes far into the future. In some cases like automobiles, you may collect up front and pay out over the course of the next year. In others, like asbestos liabilities, the losses get paid out over a very long period of time as it takes time for people to develop asbestos-related symptons. (Interestingly, Johns-Manville was one of the largest companies to file for bankruptcy in 1982, because of asbestos-related liabilities. Berkshire acquired Johns-Manville in 2001). In both cases, the money that you collect up front is called "float" and you have the privilege of investing it and reaping the rewards.
Another way to look at float is as a permanent source of funding that is costless and in some cases actually generates money. If you under-write poorly, though, this source of funding may end up costing you a lot more than if you had relied on, say, long-term bonds. And it can also be volatile. Berkshire Hathaway's track record in this area has been extremely good. "Float" has proven to be a source of funding for Berkshire Hathaway that has over the long run 1) actually made money and 2) grown. And it is largely due to a culture of conservative investing, underwriting discipline and a long-term orientation.
Berkshire groups its insurance businesses in four categories:
GEICO - today one of the largest providers of automobile insurance, GEICO is one of Berkshire's most recognizable businesses. Its competitive advantage is a direct-to-consumer market which gives GEICO a 15% pricing advantage over the traditional agent model ("Fifteen minutes can save you fifteen percent or more on car insurance!"). GEICO, which started marketing to government employees through the mail in 1936, is still doing the same thing, except relying heavily on the Internet and branded advertising. As automobile insurance is relatively short-term (one year), GEICO looks to underwrite at a large and stable profit, but nevertheless has provided steadily growing float for Berkshire to invest. Since 1993, GEICO has generated $8 billion in pre-tax underwriting profits and grown its float from $1.9 billion to $9.6 billion. Their growth is robust, and they are applying the business model to tangential areas like motorcycle and boat insurance.
General Re - one of the largest providers of reinsurance in the world. Reinsurance is basically insurance provided to other insurers so risk can be shared and spread out. General Re (which includes Cologne Re for international coverage) focuses on reinsurance of property/casualty, life and health insurance. Berkshire's competitive advantage in this area is their high rating (until they were downgraded last year, Berkshire was the only Triple-A rated reinsurance business), discipline (again, avoiding bad underwriting) and capital availability. Because many of these reinsurance contracts are long-term, the General Re generates significant float, $21.0 billion as of December 31, 2009.
Berkshire Hathaway Reinsurance Group - similar to General Re, this is the group led by Ajit Jain and takes on exceptionally large event-driven risks such as earthquakes, hurricanes and other natural disasters. Berkshire's competitive advantages in this area are again similar to General Re: high rating, disciplined under-writing and access to capital. There are only a handful of firms that can write this type of insurance. Similar to General Re, float is significant - $26.2 billion as of December 31, 2009.
Other Insurance - Berkshire owns a handful of other insurance businesses that insure anything from worker's compensation to construction. Insurance businesses are not difficult to start if you have the capital and the right people. For example, in 2008, Berkshire activated BH Assurance Company to start under-writing tax-exempt municipal bonds after they identified a perceived market dislocation with all that was happening to the likes of Ambac Financial and MBIA (Buffett has since remarked that the risk here is worse than he had originally thought - more to emphasize how much risk these guys were taking).
I valued each of these businesses based off "normalized" pre-tax operating income that factored 1) earnings from under-writing and 2) income from the "float" attributable to that particular business. So take GEICO. Over the last 10 years, GEICO has averaged 10.8% pre-tax operating margins on premium income. In addition, they generate income on the "float" that they generate. I assumed a 7.00% earnings, which based on Berkshire's track record is very conservative. This bumps operating margin to 15.3%. In 2009, they generated $13.6 billion of premium income, representing over 9.6 million policies. If you apply a 15.0% margin to that revenue, you get pre-tax profit of $2.0 billion. Because GEICO still has a lot of growth left (8% YoY growth in policies), I applied an 8.0x valuation multiple, or around $16.3 billion.
For the reinsurance businesses, I smoothed out (but did not ignore) catastrophe losses and adjustments to reflect the short-term volatility of this business. I used lower valuation multiples (4-5x) for the reinsurance businesses due to the lumpy nature of these businesses, but note that much of the value here comes from the large float that they generate. All in all, I valued General Re, BH Reinsurance and the Other Insurance businesses at $26.4 billion.
So that brings me to around $42.7 billion for the four insurance groups. However, this does not include what is the biggest value held in the insurance group.
Insurance businesses throw off a lot of cash. In addition to the "float", Berkshire's insurance businesses typically underwrite at a profit and have historically invested very well. As Berkshire does not pay out a dividend and rarely seeks to buy back stock, all of this excess cash needs to be invested. In Berkshire's case, it goes into either cash, equity securities, fixed income securities or operating businesses.
In general, most of the equity and fixed income securities at Berkshire are grouped into the Insurance group. As of December 31, 2009, Berkshire held:
Cash and cash equivalents - $27.9 billion
Fixed income securities - $32.5 billion
Equity securities of $56.6 billion
Additional investments of $29.0 billion
In aggregate, this is $146.0 billion in cash and investments. This is against "float" of $61.6 billion. Even though "float" represents liabilities, Berkshire has consistently grown and made money off its "float". I captured this value by assuming that they will continue to generate a return on this float (7%). In excess of its "float", Berkshire holds cash and investments in the insurance group of $84.4 billion if valued at market prices as of December 31, 2009. With the S&P 500 up about 5% since then, there is probably another $4 billion unrealized fair market value gain.
Included in this group of investments are high-quality businesses that Berkshire has invested in. Notable ones include American Express, Coca-Cola, Proctor & Gamble and Wells Fargo. They also include the preferred and fixed income investments Berkshire made in companies like GE, Goldman Sachs, Wrigley and Harley-Davidson during the market dislocation of 2008/2009, when Berkshire looked like the only available source of capital left standing. And for the purposes of this bottoms-up valuation, it also includes Berkshire's 22% stake in Burlington Northern Santa Fe (acquisition did not close until Q1 2010).
Operating Businesses
Even though Buffett is more well-known for his stock-picking prowess, Berkshire does not just re-invest its extra cashflow into stocks and fixed income securities. In fact, Buffett has for a long time signaled an increasing concentration on owning entire operating companies, because with wholly owned subsidiaries, you can fully control the allocation of free cash flow (either back into the business or collect it to make other investments). Berkshire groups these companies into various buckets that shed light onto how he views these businesses. Berkshire's investment preferences have evolved over time due to a number of factors which I will attempt to break down.
One question that people have asked is whether Berkshire Hathaway deserves a conglomerate discount for holding so many unrelated companies. My answer is no based on two major factors. First, the problem with conglomerates was that they justified their purchases of new businesses by using operating synergies, which meant that they needed to centralize certain functions. These synergies turned out to be illusory. Berkshire makes no such attempt to centralize operations, and it is evidenced in the size of their corporate office which has barely growth through the years. Berkshire allows its companies to run completely independently and only really steps in to decide what to do with its cashflow (re-invest in the business or make available elsewhere) and how to compensate management. Berkshire seeks to only invest in businesses where they think management will buy into this system. Second, conglomerates typically provide less accounting disclosure as they group formerly separate businesses together. That happens to a certain extent at Berkshire, but based on going through decades of financials, I can say that Berkshire is 1) very consistent and 2) provides you with information that is truly relevant. Basically, I trust them, but that trust is based on a mountain of evidence.
Larger Operating Businesses
These businesses are large enough within the Berkshire Hathaway organization to warrant breaking out their financials and getting more discussion and airtime.
Marmon Holdings - This is a collection of 130 manufacturing and service businesses that operate independently within 11 diverse business segments. Comprises anything from engineered wire & cable to transportation services, retail store fixtures and food service equipment. Berkshire acquired 60% of Marmon from the Pritzker family for $4.5 billion and will purchase the remainder from 2011-2014 based on future earnings. These niche industrial businesses tend to be stable and provide decent returns on capital. Revenue and earnings did decline significant from 2008 to 2009 due to the economic downturn, but the intrinsic value of these businesses should still be intact. Using a 9x valuation multiple on 2009 earnings (depressed), I value Berkshire's 63.6% stake in Marmon at around $4 billion, which is around the original purchase price. I viewed this acquisition as a purchase of a good collection of businesses where the deal dynamics led to an attractive purchase price.
McLane - Operates 20 grocery and 18 foodservice distribution centers across the country. Formerly a non-core business unit within Wal-Mart which still accounts for about a third of their business. Their competitive advantages are based largely on leveraging their scale to provide low-cost distribution. Growth comes from increasing their footprint and the breadth of products that they distribute. Acquired in 2003 for $1.45 billion, the business has steadily grown and was one of Berkshire's best-performing operating businesses in 2009, growing pre-tax earnings from $276 million to $344 million. I value this business using a 7x multiple at $2.4 billion.
Shaw Industries - Shaw is the largest manufacturer of tufted broadloom carpets and is also a full-service flooring company. They were acquired by Berkshire in 2001-2002 for about $2.5 billion in aggregate and have since continued to make acquisitions in this space. For example, in 2005, Shaw spent $550 million to acquire two companies providing carpet backing and nylon fiber manufacuring. This business has been hit hard by the housing downturn. After peaking at $594 million of pre-tax income in 2006, in 2009 they were down to $245 million in earnings (adjusting for plant closure costs). Because of the cyclical nature of this business, I smoothed out the long-term operating margins (7.5%) and applied it to 2009 revenue of $4.0 billion. Applying a 7x multiple, I calculated intrinsic value of $2.1 billion.
Utilities and Energy
Although this group is entitled "Utilities and Energy" in the 2009 report, the reason why Burlington Northern is going to henceforth be grouped here is that it is a business with high capex requirements and regulated returns - very similar to the utilities business. This is an important concept and there is a reason why Berkshire started investing in these businesses relatively recently (Berkshire first invested in MidAmerican in 1999). Berkshire had previously favored "great businesses" like See's Candies or Coca-Cola. A "great business" in Buffett's mind is one that has very high returns on capital (or no capital requirements at all) and could grow earnings without any incremental capital, usually through pricing power. The "great" thing about these businesses is that you could take all of the cashflow generated from that business and invest it elsewhere yet still grow the intrinsic value of the business over time.
The only problem with these "great" businesses is 1) they are not easy to find in whole pieces (i.e. not passive equity) and 2) they are even harder to find as the sums of money you want to invest increase. So when Berkshire was generating cashflow in the tens of millions, it was worth Buffett's effort to dig up these businesses. Now that Berkshire's cashflow generation is in the tens of billions, it has become increasingly difficult to put that money to work in only "great" businesses. In order to move the needle for Berkshire, they need to buy businesses with at least $75 million of pre-tax operating income (and this threshold keeps rising - it was $50 million in 2003). One could argue that at some point Berkshire should really start issuing dividends, but as long as Buffett thinks he can out-perform the indices - and clearly he still thinks he can - I would not expect that. This is a separate topic for discussion down the road - my sense is that Buffett views the book value (as a proxy for intrinsic value) as a scorecard. Book value is "pure" as it represents a continuous compounding of value over time (Berkshire has not paid a dividend pretty since its days as a pure textile manufacturer).
More importantly, Buffett clearly does believe that he can continue to out-perform the indices, even after he dies. The way to do it is through these regulated businesses. And the way to ensure that all that tens of billions of excess cashflow can continue to be invested is through capital expenditures in these regulated businesses. While you cannot expect equity returns much above the low teens, if management does its job correctly, you know that the money will earn a decent return and continue to offer opportunities to compound, albeit at a lower rate. And if you look at the long-term track record of the S&P 500, particularly after including the dreadful returns of the last ten years, a 15% return on equity, in a monopoly business compounded over decades, is not bad at all.
Buffett would have never invested in these types of businesses 20 years ago. But today, there aren't many businesses out there that can suck up capital in such great quantities and deliver this level of return. As he has said over and over again, percentage returns are certain to decline over time.
MidAmerican - Berkshire first invested in MidAmerican in October 1999. In 2002, Berkshire helped fund the acquisition of natural gas pipeline assets that had become distressed in the post-Enron fallout. In 2006, Berkshire funded the acquisition of a large power utility Pacificorp. MidAmerican also pounced on Constellation Energy when it got in trouble in 2008 and ended up making a billion dollars in less than a year. But most of the investment in this business is going to come from future capital expenditures. For example, MidAmerican is the largest builder of wind farms in the country. They have also approached it in a relatively risk-averse way - building wind farms adds to their regulatory asset base which increases their earnings. Clearly there is a lot of need over the next fifty, hundred years to upgrade infrastructure and that means plenty of opportunity for billions of reinvestment. I do not expect MidAmerican to return cash to corporate for a long time - it will all be invested back into the business.
I valued the various businesses based on a valuation multiple of 7-10x pre-tax earnings. That pegged the regulated utilities (MidAmerican, CalEnergy and Pacificorp) at $14.5 billion, the gas pipelines at $3.4 billion and - somewhat out of place - the real estate brokerage assets at $0.4 billion. Chop of $5.4 billion in corporate debt and you arrive at a valuation of $15.3 billion. Berkshire owns 89.5% of this, or $13.7 billion.
Burlington Northern Santa Fe Railroads - Berkshire acquired a 22% stake in BNSF over a few years and Buffett has probably been following the railroad sector since the 1950s. However, since my valuation is based on figures as of December 31, 2009, the value of Berkshire's stake in BNSF is included in the equity securities categorized under Insurance.
Other Businesses
Many of the businesses that are grouped together today used to be large enough in the context of Berkshire to warrant their own entries. For example, Buffalo News used to be one of the larger contributors of earnings for Berkshire. But a decline in the newspaper industry coupled with the growth of Berkshire meant that Buffalo News was no longer a significant part of Berkshire's vast intrinsic value. In fact the last time Buffalo News was separated out was in 1999. That is not meant to reduce the significance Buffalo News had on Berkshire. Without the close to $1 billion of cashflow generated from 1981 to today, Berkshire would not have been able to re-deploy capital into other businesses and investments that are worth multiples of that today.
Other Manufacturing - This category includes apparel manufacturers Fruit of the Loom and Russell, metalworking company Iscar, construction-related companies Acme Building Brands, Johns Manville and MiTek, and a number of other industrial businesses. These businesses were generally hurt by the economic downturn in 2009 and saw revenues and earnings decline 16% and 51%, respectively. Because 2009 earnings are depressed and not necessarily representative, I used 2008 earnings as a better proxy and applied a 6x valuation multiple, arriving at a valuation of about $9.6 billion.
Other Service Businesses - This category includes a wide variety of service businesses such as aviation simulator FlightSafety, fractional airplane ownership company NetJets, information services provider Business Wire and kitchen tools seller the Pampered Chef. Using a similar valuation multiple and methodology, I arrived at a valuation of about $5.7 billion.
Retailing - This includes See's Candies, a number of jewelry and home furnishing retailers. Using similar valuation multiples and methodologies, I arrived at a valuation for this collection of companies of $1.5 billion.
Finance and Financial Products
The Finance and Financial Products group is comprised primarily of manufactured housing provider Clayton Homes and two leasing businesses. The inclusion of Clayton Homes in "Finance and Financial Products" may seem odd at first glance. However, when you delve into the business, you realize that it really makes its money on loans provided to finance the purchase of these manufactured homes. For example, despite significant shrinkage in the manufactured homes market, Clayton Homes consistently increased its earnings from 2003 through 2007. Based on a valuation multiple of approximately 6x pre-tax operating earnings, I estimated the value for Clayton Homes and the two leasing businesses at about $3.4 billion.
I referenced Clayton Homes in an earlier post (http://minoritycapital.blogspot.com/2009/03/irrational-cds-pricing-at-berkshire.html).
Value of Derivative Bets
From 2004 to 2008, Berkshire wrote equity index put option contracts on four major equity indices. This was well-publicized because Buffett had famously remarked that derivatives were "weapons of mass destruction" and these put option contracts were seen as a direct contradiction of that statement. However, if you looked a little closer, it became clear that Berkshire's derivative bets were really more like catastrophe insurance. First, Berkshire collected the premiums up front (and could invest them over the duration of the contract, similar to "float" - in the case of the equity index put option contracts, $4.9 billion), so there was no counter-party risk. Second, Berkshire had minimal collateral requirements, so they would never be in danger of being forced to sell.
Just as Buffett has said that they may lose money on some of their catastrophe insurance bets, Berkshire could be wrong in its calculated bet that markets will be higher ten to twenty years from now than today. In this case, there is a chance that they could pay out up to $38 billion in premiums (though that is only if they go to zero, in which case we may have larger issues to worry about!). Meanwhile, they can invest the premiums they collected up front and likely compound that amount over time. In the best case, they get to invest $4.9 billion for 15-20 years and the option contracts expire with no loss payouts. By the way, to get to $38 billion in 15 years, you need to compound $4.9 billion at approximately 14.6%. Over twenty years, that figure is 10.8%.
At the end of December 2008, Berkshire was being lambasted for having entered into these equity index put option contracts. With equity markets around the world near multi-decade lows, accountants calculated an unrealized liability of $10 billion related to these contracts. Even though this is really just an accounting entry, it spooked some enough to fuel the eventual downgrade of Berkshire from its long-time Triple-A rating and send CDS protection on its bonds skyrocketing to irrational levels. Fast-forward to December 2009, and with equity markets much improved, the unrealized book liability on these contracts is down to $7.3 billion. However, just as the decline in book value attributable in 2008 to these contracts was not really meaningful, the increase in book value attributable to 2009 is also not very meaningful.
Berkshire also entered into a number of other derivative contracts. Some of them have not worked out that well and have been settled with payouts.
I had excluded the balance sheet from Finance and Financial Products in my earlier calculation of Investments at FMV in excess of float. I estimated the net asset value here to be $3 billion and the true liability of the various derivatives to be around $3.6 billion. So net-net, that is around $0.6 billion of liability.
Final Tally
Insurance Businesses - $42.7 billion
Excess of Investments over Float - $84.3 billion
MidAmerican - $13.7 billion
Marmon, McLane, Shaw - $8.5 billion
Finance Products - $3.4 billion
Other Manufacturing - $9.6 billion
Other Services - $5.7 billion
Retailing - $1.5 billion
Financial Hedges - $(0.6) billion
Total Intrinsic Valuation - $169 billion
With approximately 1.55 million Series A shares outstanding, I calculate intrinsic value at around $108,950 per share, which is less than the current share price of around $122,000. However, two points of note. First, the investment portfolio on a mark-to-market basis is up from December 31, 2009 which probably adds another $4-$6 billion to the book value of Berkshire. Second, I used fairly conservative valuation multiples (4-8x) across-the-board to value Berkshire's operating businesses.
Nevertheless, I would conclude that there is not a significant amount of "margin of safety" in Berkshire's stock price. That said, given Berkshire's conservative structure and collection of high-quality assets, I think there is also much lower risk. I am not dying to buy more shares of Berkshire at the moment, and see other assets that I think are more attractively priced. However, I would definitely not sell at the current price.
Disclosure: I own shares of Berkshire Hathaway.
This was originally published in a personal blog in March 2010.