The Motley Fool Discussion Boards
Stocks B / Berkshire Hathaway
|Subject: The Lesson||Date: 6/20/2003 1:24 PM|
|Author: EliasFardo||Number: 78783 of 224508|
If you are going to hold a common stock forever, if you are never going to sell an operating division, even if its results are unsatisfactory, how do you change? How do you make adjustments in your portfolio? The world changes around you. How do you take advantage of that? In an effort to answer those questions, and learn a few other lessons along the way, lets look at Berkshire from 1980 to 2002.
At the end of 1980, Berkshire Hathaway was a much different creature than what we know today. It is not just that it was much smaller; it only had assets of $1 billion and equity of $400 million. And it is not that it was not then an insurance company. Of its $53 million of after tax profits, nearly $40 million was either underwriting income, investment income or realized security gains; all or nearly all of which were insurance related. At December 31, 1980, the insurance group compromised 70% of the total consolidated assets of Berkshire Hathaway compared to 67% at the end of 2002; which is remarkably consistent. So it was still an insurance company. And it is not that it did not then have substantial noninsurance operations, for it did. But still, it was very different.
I am going to look at how it is different by examining the equity investments, the noninsurance companies and the insurance businesses.
Marketable Equity Securities:
At December 31, 1980, Berkshire owned marketable equity securities that were worth a full 133% of stockholders' equity. At December 31, 2002, equities comprised only 44% of stockholders' equity. In total, at the end of 1980, Berkshire had 193% of its equity invested in liquid, marketable investments compared to 119% at the end of 2002.
When comparing the equity marketable securities owned to the total market value of Berkshire Hathaway at December 31 for both 1980 and 2002, the difference is staggering. At December 31, 1980, the market value of Berkshire Hathaway was $420 million, and it owned $526 million of marketable equity securities, or $1.25 of equities for every dollar of market value. For that same percentage relationship to exist at the end of 2002, Berkshire Hathaway would need to own $140 billion of equities, instead of the actual amount of $28 billion.
The large exposure in 1980 to the equity markets was especially important. During the following year, Buffett placed another $30 million into equities, which along with that year's equity appreciation increased the total value of equities to $641 million at December 31, 1981. So going in 1982 and the beginning of one of the greatest bull markets of all time, Berkshire Hathaway had a total market value of $552 million, but held $641 million of marketable equity securities.
The change in the names of the equity investments over this period is dramatic. The schedule of marketable equity securities as of December 31, 1980 included 18 companies by name. Of these 18 names, Berkshire Hathaway is currently associated with only two: GEICO and The Washington Post. The four largest holdings by market value were GEICO at $105 million, General Foods at $60 million, SAFECO at $45 million and The Washington Post at $42 million. These four companies equaled 44% of the total $526 million of equities. At December 31, 2002, the largest four equity positions were Coca-Cola at $8,768 million, American Express at $5,359 million, Gillette at $2,915 million and Wells Fargo at $2,497 million. These four companies equaled 69% of the $28,363 million of equities. So over this 20 year plus period, not only did the value of marketable equity securities decrease as a percentage of stockholders' equity and market capitalization, but it became focused into fewer names as well. The equity portfolio got more concentrated even as it got bigger.
During 1980, there were ten different operations with enough significant activity to be listed under "Sources of Reported Earnings" in the Shareholders' Letter. It is instructive to look at the changes that occurred in these ten units. The largest unit, the insurance group, is the only one of the ten whose ownership was subsequently neither increased or decreased. Of the other nine, Berkshire-Waumbec Textiles was slowly liquidated, Associated Retail Stores was sold, See's Candies, 60% owned in 1980 had its ownership subsequently increased to 100%, Buffalo Evening News, 60% owned in 1980 had its ownership subsequently increased to 100%, Blue Chip Stamps - Parent, 60% owned in 1980 had its ownership subsequently increased to 100% and is currently insignificant, Illinois National Bank was spunoff, Wesco Financial - Parent, 48% owned in 1980 had its ownership subsequently increased to 80% and is currently insignificant, Mutual Savings and Loan, 48% owned in 1980 had its ownership subsequently increased to 80% and was subsequently sold, and Precision Steel, 48% owned in 1980 had its ownership subsequently increased to 80% and is currently insignificant. Of the nine noninsurance sources of earnings for 1980, two were sold, one was liquidated, one was spunoff and three are currently insignificant in regard to assets or earnings. The only two noninsurance operations which still have a significant financial influence on the parent company are See's Candies and the Buffalo Evening News.
Why the fading away of so many companies? In the 1985 annual report, Buffett wrote, "I won't close down businesses of sub-normal profitability merely to add a fraction of a point to our corporate rate of return. However, I also feel it inappropriate for even an exceptionally profitable company to fund an operation once it appears to have unending losses in prospect. Adam Smith would disagree with my first proposition, and Karl Marx would disagree with my second; the middle ground is the only position that leaves me comfortable." If a company does not provide adequate returns on its capital, Buffett will not kill it, but it will not put it on indefinite life support either.
At the 2000 annual meeting, Charlie discussed this very thing: "I think it's in the nature of things for some businesses to die. It's also in the nature of things that in some cases, you shouldn't fight it. There's no logical answer in some cases except to wring the money out and go elsewhere."
Berkshire would not have been near as profitable an enterprise if Buffett had not been willing to let go of underperforming units; not necessarily killing them, but not encouraging them either. This is, unfortunately, a major departure from the way many business enterprises are operated. Once a company gets associated with doing something, it tends to continue to do that something, even if returns are unsatisfactory. And sometimes, holding on to an underperforming unit pays off.
This story is of the Buffalo News. The News was purchased by Blue Chip Stamps in 1977 for $34 million. It proceeded to lose money, about $12 million through the end of 1982. In 1977, the News started a Sunday edition, and as a result suffered interlocutory injunctions from its Sunday competitor. Before these injunctions were reversed in 1979, they created circulation and promotion problems for the News.
Then in 1980, the News took a short strike. Losses, lawsuits and labor unrest put Munger in a pensive mood. In the 1981 annual report for Blue Chip Stamps, he wrote, "If we hadn't purchased the News in 1977 but had simply earned returns on the unspent purchase price comparable with the average earnings power of the rest of our shareholders' equity, we would now have about $70 million in value of other assets, earning over $10 million per year, in place of the Buffalo Evening News and its current red ink. No matter what happens in the future in Buffalo we are about 100% sure to have an economic place lower than we would have occupied if we had not made our purchase."
Charlie need not be so despondent. For in 1982, the Courier Express, the News' competitor, failed. This left the News as the only area-wide daily newspaper in Buffalo. Enjoying its newfound monopoly status, the Buffalo News experienced pretax earnings of $19 million in 1983, and $27 million in 1984. In the 1982 annual report for Blue Chip Stamps, Munger writes, "Finally, our shareholders should recognize that if our 1977 purchase of the News has now worked out acceptably from their viewpoint, which contrary to our prediction last year may now be true even after taking into account time delays, the conclusion does not follow that we made a sound managerial decision buying the News when we did for the price we paid." In other words, "even though I was wrong last year when I said that 'we are about 100% sure' that the purchase of the News was a mistake, its purchase was still a bad managerial decision." Or, " even though it was not a mistake, and therefor a mistake to say it was a mistake, it was still a mistake." Buying the News may have been a mistake. But once purchased, holding on was not.
The surge in the profitability of the News was in part responsible for an unusual period of time for Berkshire. The period from 1983 to 1986 was the last time that the noninsurance operations of Berkshire were more profitable than the insurance group; at least when realized and unrealized capital gains are eliminated, until 2001. During the four years ending in 1986, the insurance group experienced significant underwriting losses, something which was repeated in 2001.
With the purchase and continued ownership of companies with substantial moats, such as See's Candies, the Buffalo paper, the Nebraska Furniture Mart and Borsheims, and the with the slow abandoning of enterprises without moats, such as Associated Retail Stores and the textile operations, the Berkshire of "substantial competitive advantages" was being formed. These early moats shared at least one commonality: geography. They all had geographical moats, dominating one city such as Omaha or Buffalo, or entrenching themselves in one region, such as the West for See's. This is not to say that geography was the only moat enjoyed by these companies, for it isn't. But geography may be the most easily identified and recognizable moat. It was left to subsequent acquisitions to more fully flesh out the range of ways in which businesses can enjoy substantial competitive advantages.
These early operating entities had another commonality: they weren't growing very quickly. For Buffett, this is not a problem. At the 1994 annual meeting, he said, "We're willing to buy companies that aren't going to grow at all - assuming we get enough for our money when we do it. But you can certainly have a situation where there's absolutely no growth in a business and it's a much better investment than some company that's going to grow at very substantial rates - particularly if they're going to need capital in order to grow. There's a huge difference between the business that grows and requires lots of capital to do so and the business that grows and doesn't require capital." He will use the cash flow from these companies in the same way in which he used the cash flow from the textiles businesses and Blue Chip Stamps: investing it in other operations. This whole mode of business operation is not often practiced, and not even widely understood.
For many companies, it is almost predetermined how their ready capital will be deployed. They do a particular thing, they are going to continue to do that particular thing, and they are going to attempt to get bigger at doing that particular thing until even the slowest thinker recognizes the need to make a change. So, the value of a company can be limited by the preconceived notions vested in its future capital expenditures. Not so at Berkshire. At the 2001 annual meeting, Buffett said, "It's relatively easy to figure out the present value of most of our businesses. But the question becomes what do we do with the money as it comes in? That will have a huge impact on the value 10 years from now. " What that "huge impact" will be is a central question of valuation. And since Buffett understands the importance of this impact, he makes it central to business practice.
From 1980 to 1994, noninsurance acquisitions came at an average of about one every two years. I was not until 1995 before Buffett seemed to get serious about expanding Berkshire through acquisitions. Below are the acquisitions by year, based on the year the purchase closed, of noninsurance companies:
1980 to 1984 - 1
1985 to 1989 - 3
1900 to 1994 - 3
1995 to 1999 - 7
2000 to 2002 - 16
An examination of the individual purchases during these periods is telling. Of the four acquisitions between 1980 and 1989, two, Nebraska Furniture Mart and Borsheim's are in Omaha, and are operations with which Buffett would have personal familiarity. A third is Scott Fetzer, a slam dunk acquisition. The fourth, Fechheimer is not a very large company. And all three of the purchases during the period of 1990 to 1994 were shoe companies. Even the purchases in 1995 must have been comfortable for Buffett, being jewelry and furniture store chains. The balance of GEICO was acquired in 1996, but this was a company with which Buffett had a long relationship; not exactly groundbreaking.
Buffett's perceived circle of competence, a least with acquisitions of entire companies, seemed to contain little real adventure. Outside of reinsurance, he was comfortable selling newspapers, shoes, candy, furniture, jewelry, vacuum cleaners, encyclopedias and individual car insurance policies. Many of these companies had geographical moats, and all of them sold products directly to individual consumers, not to businesses.
If you look at this pattern of making comfortable purchases as a spinning wheel, and want to know the name of the cog that was placed in the wheel's spokes to stop the spinning, it would be FlightSafety International, purchased in 1996. It is a - gasp - INTERNATIONAL company, just look at its name. It sells a highly technological products; and it sells to businesses, not individuals. This is not an Omaha furniture store.
This does not mark the end of the old Warren Buffett; he still buys the occasional jewelry or furniture store chain. And he eventually even goes back to selling products that are similar to the textile roots of Berkshire. But, from here on, the acquisitions at Berkshire are increasingly eclectic, and impossible to anticipate. This is mostly by necessity, there are just so many furniture and jewelry operations worth owning. And the growing amount of investable funds requires looking at larger and larger targets of opportunity.
However, it is still all about moats. From the 2000 annual meeting: "If you're evaluating a business year-to-year, the number one question you want to ask yourself is whether the competitive advantage has been made stronger and more durable. And that's more important than the P&L for a given year." Moats trump P&L. So, when Buffett goes shopping, he is always looking for sustainable competitive advantages. A common moat of many of the more recent acquisitions is market dominance. The companies Buffett has purchased recently are usually either a leader, or the leader in their industry. For instance, Albecca is the U. S. leader in custom-made picture frames, Fruit of the Loom produces about 33.3% of male underwear in the U. S., MiTek is the world's leading producer of connector plates for roofing trusses, XTRA is a leading lessor of truck trailers, Justin Industries is the leading maker of Western boots, Acme is the leading producer of bricks in its geographical area, Shaw Industries is the world's largest carpet manufacturer, Johns Manville is the nation's leading producer of commercial and industrial insulation and FlightSafety International is the world's leader in pilot training. And of course, NetJets is not only is the worldwide leader, but it actually created the industry.
Market dominance itself may not be the important moat. It may actually be evidence of other advantages that in turn create the domination. But once achieved, market dominance becomes one more nasty moat to harry the competition. It is the width and depth of these moats, and the ferocity of its resident vermin, that will determine much of the future success of Berkshire. Also from the 2000 annual meeting, "So we think in terms of that moat and the ability to keep its width and its impossibility of being crossed as the primary criterion of a great business. And we tell our managers we want the moat widened every year. That doesn't necessarily mean the profit will be more this year than it was last year because it won't be sometimes. However, if the moat is widened every year, the business will do very well."
But, Buffett is not willing to build the moat. It must come preconstructed, and it must not be messy. Another common characteristic of some of the recent purchases is a good, preacquisition cleaning. Justin Industries and Shaw Industries had both gone through a period of restructuring before Berkshire's purchase. They had consolidated and disposed of some operations. Johns Manville and Fruit of the Loom came scrubbed clean out of bankruptcy. So, the best way to predict an acquisition of Berkshire is to identify a company that has been recently restructured in bankruptcy, leads the world in its industry, has management in place - and is a chain of furniture stores.
To see the full extent of the changes in noninsurance business from 1980 to 2002, look at the changes in "Sources of Reported Earnings" in the annual report. None of the sources from 1980 are still being individually identified. Even as recently as 1996, Fechheimer, Kirby, World Book, See's Candies, Scott Fetzer, Home Furnishings, Buffalo News, Jewelry and the Shoe Group were named as sources of reported earnings. Today, instead, there is Apparel, Building Products, Flight Services, Finance and Financial Products Business, MidAmerican Energy, Retail Operations, Scott Fetzer, Shaw Industries and Other. The only surviving reported source of earnings from 1996 is Scott Fetzer. And it is just a few acquisitions away from joining See's and the News in "Other."
Most of these sources of reported earnings were acquired in the last eight years. If acquisitions in the next eight years are of sufficient size and profitability to make a similar change in the listed sources, I have a hard time thinking what they could be. If that happens, Berkshire's noninsurance operations will have grown to an enormous size. But it may. The pace of acquisitions has been accelerating, with about half of the acquisitions made in the last 23 being closed in the last 3 years alone. And this doesn't even take insurance into consideration. And, Berkshire is still primarily an insurance company.
In 1980, the insurance premium income of Berkshire Hathaway was $185 million. In 1984, is was only $140 million. The business was moribund. This came from Buffett at the 1996 annual meeting, but it explains why Berkshire is willing to suffer falling volume in insurance: "We have promised people at all of our insurance operations that we will never have layoffs because of a drop in volume. We do not want the people who run our insurance operations to feel like they have to write $X in order to keep everybody there. We can afford some overhead costing us a little money through lack of using our operation at full capacity - because it isn't that much relative to the size of our insurance operation. What we can't afford is people feeling some internal compulsion to keep writing business in order to keep their jobs. So we have a strong policy on that."
In 1985, premium income started to grow, and topped out at $825 million in 1987 before falling to a low of $395 million in 1989 when a major quota share treaty expired. All of this period was plagued by underwriting losses, which started appearing in 1982 and persisted until 1992. To illustrate just how bad it was, the combined ratio was a painful 121 in 1983, and a devastating 134 in 1984. When the first underwriting profits in a decade were finally enjoyed in 1993, insurance premiums had grown back to $650 million.
Reading the insurance discussion in the annual reports for this period is a very sour experience, each year's commentary more dismal than the previous. Buffett continually pounded the table, insisting that a 10% increase in written premiums was necessary to just keep the combined ratio unchanged. And he annually posted a table showing that such percentage increase was not reached in the previous year. At times, the reader could rightly wonder why Berkshire even remained in the insurance business. In 1987, Buffett wrote, "The insurance industry is cursed with a set of dismal economic characteristics that make for a poor long-term outlook: hundreds of competitors, ease of entry, and a product that cannot be differentiated in any meaningful way. In such a commodity-like business, only a very low-cost operator or someone operating in a protected, and usually small, niche can sustain high profitability levels."
Warren Buffett was clearly looking for a moat for the insurance operations of Berkshire Hathaway. According to him, some moats just simply are not possible in this business. There are hundreds of competitors, so forget monopoly pricing. There is an ease of entry by new competitors; so even if you knock off some of your competitors, new ones spring up to replace them. The product cannot be differentiated in any meaningful way. Your offered insurance policy is no different and no more desirable than the policies of one of your many competitors; so forget brand names. He identifies only two possible moats: being the low cost producer and finding a niche.
"At Berkshire," continues Buffett in the same 1987 annual report, "we work to escape the industry's commodity economics in two ways. First, we differentiate our product by our financial strength, which exceeds that of all others in the industry." OK, he identifies three moats: being the low cost producer, finding a niche, and having market leading financial strength. "Our second method of differentiating ourselves is the total indifference to volume that we maintain. In 1989, we will be perfectly willing to write five times as much business as we write in 1988 - or only one-fifth as much." OK, so he identifies four moats: being the low cost producer, finding a niche, having market leading financial strength and maintaining pricing discipline even at the expense of volume - or "The Refusal To Shoot One's Own Foot" moat.
Throw in one more moat, providing superior customer service, and we have identified the moats on which the various Berkshire Hathaway insurance companies attempt to make their fortune. GEICO is predicated on customer service and being the low cost producer, and the rest of the insurance businesses employ niches, financial strength and pricing discipline at the expense of volume. And how to put this superior financial strength to work? Again, from the 1987 annual report, "Our insurance business has also made some important non-financial gains during the last few years. Mike Goldberg, its manager, has assembled a group of talented professionals to write larger risks and unusual coverages. His operation is now well equipped to handle the lines of business that will occasionally offer us major opportunities." This paragraph was a foretelling. In the 1988 annual report, the now familiar name of Ajit Jain first appears. And the 1989 annual report contains the first discussion of Cat covers. The now immensely profitable Berkshire Hathaway Reinsurance Group was being constructed.
Things got better. In 1992 Buffett wrote, "Charlie and I continue to like the insurance business, which we expect to be our main source of earnings for decades to come. The industry is huge; in certain sectors we can compete world-wide; and Berkshire possesses an important competitive advantage. We will look for ways to expand our participation in the business, either indirectly as we have done through GEICO or directly as we did by acquiring Central States Indemnity." Those 1992 comments reflect a more sanguine Warren Buffett. And by 1994, when Berkshire produced an $130 million underwriting gain, the man was almost giddy. And events subsequently proved that his confidence in the future profitability and size of the Berkshire insurance business was not misplaced. In every year from 1993 forward, the Super Cat business of Berkshire has produced a substantial underwriting gain. In 1996, Berkshire paid $2.3 billion to buy the other 49% of GEICO it did not already own. But, in 1998 came the return of horrible underwriting results with the purchase of General Re.
Selling insurance is like buying a common stock, only in reverse. When buying a stock, you know the amount of your cash outflow; but not your cash inflow. When selling insurance, you know the amount of your cash inflow, but not your cash outflow. But, the emotional skills of selecting a company to buy will transfer to selecting a risk to assume. You still need to stay within your circle of competence, you need to be disciplined, you need to be grounded in reality, you need to determine value, and you need to price with a margin of safety. In both cases, you need to compute the present value of a future stream of cash flows. It is all the same thing.
In both investing in common stocks and selling reinsurance, excess returns will flow to those who are the most skillful at exploiting market inefficiencies. As Buffett said at the 2000 annual meeting, "Reinsurance is not the world's most efficient business - and it never will be - because it's not strictly actuarial. All excess returns will not be competed away. There will be people who earn very sub-normal returns in the business. And there will be people who get killed in the business. That means there will be quite a deviation from the mean in terms of the results of individual insurers." So, reinsurance, and General Re, is an excellent business for an old stock picker like Buffett. Just as in buying marketable securities, it provides opportunities to profit from misspriced transactions.
The early years of ownership of General Re provided one of those personal growth opportunities we have all have learned to hate and fear. In contrast to almost all other acquisitions at Berkshire, it was large enough to be material, it was done entirely with Berkshire common stock, and it was immediately a problem. Certainly, no acquisition during this period under study had raised as many legitimate questions as General Re. While other factors played their part, especially market prices for equity securities, since its acquisition, General Re was the primary drag on growth at Berkshire. In 1999, the first full year after the purchase of General Re, book value per share grew by .5%. In 2000, 6.5%; in 2001 a negative 6.2%; and in 2002, 10%. For a company that had historically grown book value per share at a rate in excess of 20% a year, this is a significant reduction. It was not until early 2003 before it appeared that General Re might contribute the kind of earnings expected of a Berkshire Hathaway unit.
Berkshire Hathaway is still primarily an insurance company. Berkshire's three most important units are GEICO, General Re and The Berkshire Hathaway Reinsurance Group. All insurance. Insurance still generates more profits than the rest of Berkshire, and those reported earnings do not include the stealth earnings represented by appreciating securities held by the insurance companies. Plus, the insurance companies hold a large and growing amount of float. At the 2001 annual meeting, Buffett said, "But I think Berkshire is easier to value than most businesses, actually - because we give you all the information ( that is important to us at least ) in valuing it. Then the biggest judgment you have to make is how well capital will be deployed in the future." Back to the future. That capital Berkshire will be employing in the future will come mostly from the insurance group of Berkshire. No matter how much growth in the noninsurance business is chronicled here, it still gets back to the future of the insurance business, especially the deployment of its cash flow.
For being run