Note: The following table appears in the printed Annual Report on the facing page of the Chairman's Letter and is referred to in that letter.
Berkshire's Corporate Performance vs. the S&P 500

Year		Annual Percentage Change		Relative Results (1)-(2)
		in Per-Share Book Value of Berkshire (1)	in S&P 500 with Dividends Included (2)
1965		23.8	10.0	13.8
1966		20.3	(11.7)	32.0
1967		11.0	30.9	(19.9)
1968		19.0	11.0	8.0
1969		16.2	(8.4)	24.6
1970		12.0	3.9	8.1
1971		16.4	14.6	1.8
1972		21.7	18.9	2.8
1973		4.7	(14.8)	19.5
1974		5.5	(26.4)	31.9
1975		21.9	37.2	(15.3)
1976		59.3	23.6	35.7
1977		31.9	(7.4)	39.3
1978		24.0	6.4	17.6
1979		35.7	18.2	17.5
1980		19.3	32.3	(13.0)
1981		31.4	(5.0)	36.4
1982		40.0	21.4	18.6
1983		32.3	22.4	9.9
1984		13.6	6.1	7.5
1985		48.2	31.6	16.6
1986		26.1	18.6	7.5
1987		19.5	5.1	14.4
1988		20.1	16.6	3.5
1989		44.4	31.7	12.7
1990		7.4	(3.1)	10.5
1991		39.6	30.5	9.1
1992		20.3	7.6	12.7
1993		14.3	10.1	4.2
1994		13.9	1.3	12.6
1995		43.1	37.6	5.5
1996		31.8	23.0	8.8
1997		34.1	33.4	.7
1998		48.3	28.6	19.7
1999		.5	21.0	(20.5)
2000		6.5	(9.1)	15.6
2001		(6.2)	(11.9)	5.7
2002		10.0	(22.1)	32.1
2003		21.0	28.7	(7.7)
2004		10.5	10.9	(.4)
Average Annual Gain — 1965-2004		21.9	10.4	11.5
Overall Gain — 1964-2004		286,865	5,318

Notes: Data are for calendar years with these exceptions: 1965 and 1966, year ended 9/30; 1967, 15 months ended 12/31.

Starting in 1979, accounting rules required insurance companies to value the equity securities they hold at market rather than at the lower of cost or market, which was previously the requirement. In this table, Berkshire's results through 1978 have been restated to conform to the changed rules. In all other respects, the results are calculated using the numbers originally reported.

The S&P 500 numbers are pre-tax whereas the Berkshire numbers are after-tax. If a corporation such as Berkshire were simply to have owned the S&P 500 and accrued the appropriate taxes, its results would have lagged the S&P 500 in years when that index showed a positive return, but would have exceeded the S&P in years when the index showed a negative return. Over the years, the tax costs would have caused the aggregate lag to be substantial.

BERKSHIRE HATHAWAY INC.

To the Shareholders of Berkshire Hathaway Inc.:

Our gain in net worth during 2004 was \$8.3 billion, which increased the per-share book value of both our Class A and Class B stock by 10.5%. Over the last 40 years (that is, since present management took over) book value has grown from \$19 to \$55,824, a rate of 21.9% compounded annually.*

It's per-share intrinsic value that counts, however, not book value. Here, the news is good: Between 1964 and 2004, Berkshire morphed from a struggling northern textile business whose intrinsic value was less than book into a diversified enterprise worth far more than book. Our 40-year gain in intrinsic value has therefore somewhat exceeded our $21.9\%$ gain in book. (For an explanation of intrinsic value and the economic principles that guide Charlie Munger, my partner and Berkshire's vice-chairman, and me in running Berkshire, please read our Owner's Manual, beginning on page 73.)

Despite their shortcomings, yearly calculations of book value are useful at Berkshire as a slightly understated gauge for measuring the long-term rate of increase in our intrinsic value. The calculations are less relevant, however, than they once were in rating any single year's performance versus the S&P 500 index (a comparison we display on the facing page). Our equity holdings (including convertible preferreds) have fallen considerably as a percentage of our net worth, from an average of $114\%$ in the 1980s, for example, to less than $50\%$ in recent years. Therefore, yearly movements in the stock market now affect a much smaller portion of our net worth than was once the case, a fact that will normally cause us to underperform in years when stocks rise substantially and overperform in years when they fall.

However the yearly comparisons work out, Berkshire's long-term performance versus the S&P remains all-important. Our shareholders can buy the S&P through an index fund at very low cost. Unless we achieve gains in per-share intrinsic value in the future that outdo the S&P, Charlie and I will be adding nothing to what you can accomplish on your own.

Last year, Berkshire's book-value gain of $10.5\%$ fell short of the index's $10.9\%$ return. Our lackluster performance was not due to any stumbles by the CEOs of our operating businesses: As always, they pulled more than their share of the load. My message to them is simple: Run your business as if it were the only asset your family will own over the next hundred years. Almost invariably they do just that and, after taking care of the needs of their business, send excess cash to Omaha for me to deploy.

I didn't do that job very well last year. My hope was to make several multi-billion dollar acquisitions that would add new and significant streams of earnings to the many we already have. But I struck out. Additionally, I found very few attractive securities to buy. Berkshire therefore ended the year with \$43 billion of cash equivalents, not a happy position. Charlie and I will work to translate some of this hoard into more interesting assets during 2005, though we can't promise success.

In one respect, 2004 was a remarkable year for the stock market, a fact buried in the maze of numbers on page 2. If you examine the 35 years since the 1960s ended, you will find that an investor's return, including dividends, from owning the S&P has averaged 11.2% annually (well above what we expect future returns to be). But if you look for years with returns anywhere close to that 11.2% – say, between 8% and 14% – you will find only one before 2004. In other words, last year's “normal” return is anything but.

Over the 35 years, American business has delivered terrific results. It should therefore have been easy for investors to earn juicy returns: All they had to do was piggyback Corporate America in a diversified, low-expense way. An index fund that they never touched would have done the job. Instead many investors have had experiences ranging from mediocre to disastrous.

There have been three primary causes: first, high costs, usually because investors traded excessively or spent far too much on investment management; second, portfolio decisions based on tips and fads rather than on thoughtful, quantified evaluation of businesses; and third, a start-and-stop approach to the market marked by untimely entries (after an advance has been long underway) and exits (after periods of stagnation or decline). Investors should remember that excitement and expenses are their enemies. And if they insist on trying to time their participation in equities, they should try to be fearful when others are greedy and greedy only when others are fearful.

Sector Results

As managers, Charlie and I want to give our owners the financial information and commentary we would wish to receive if our roles were reversed. To do this with both clarity and reasonable brevity becomes more difficult as Berkshire's scope widens. Some of our businesses have vastly different economic characteristics from others, which means that our consolidated statements, with their jumble of figures, make useful analysis almost impossible.

On the following pages, therefore, we will present some balance sheet and earnings figures from our four major categories of businesses along with commentary about each. We particularly want you to understand the limited circumstances under which we will use debt, given that we typically shun it. We will not, however, inundate you with data that has no real value in estimating Berkshire's intrinsic value. Doing so would tend to obfuscate the facts that count.

Regulated Utility Businesses

We have an 80.5% (fully diluted) interest in MidAmerican Energy Holdings, which owns a wide variety of utility operations. The largest of these are (1) Yorkshire Electricity and Northern Electric, whose 3.7 million electric customers make it the third largest distributor of electricity in the U.K.; (2) MidAmerican Energy, which serves 698,000 electric customers, primarily in Iowa; and (3) Kern River and Northern Natural pipelines, which carry 7.9% of the natural gas consumed in the U.S.

The remaining 19.5% of MidAmerican is owned by three partners of ours: Dave Sokol and Greg Abel, the brilliant managers of these businesses, and Walter Scott, a long-time friend of mine who introduced me to the company. Because MidAmerican is subject to the Public Utility Holding Company Act (“PUHCA”), Berkshire’s voting interest is limited to 9.9%. Voting control rests with Walter.

Our limited voting interest forces us to account for MidAmerican in an abbreviated manner. Instead of our fully incorporating the company's assets, liabilities, revenues and expenses into Berkshire's statements, we make one-line entries only in both our balance sheet and income account. It's likely, though, that PUHCA will someday – perhaps soon – be repealed or that accounting rules will change. Berkshire's consolidated figures would then incorporate all of MidAmerican, including the substantial debt it utilizes (though this debt is not now, nor will it ever be, an obligation of Berkshire).

At yearend, \$1.478 billion of MidAmerican’s junior debt was payable to Berkshire. This debt has allowed acquisitions to be financed without our partners needing to increase their already substantial investments in MidAmerican. By charging 11% interest, Berkshire is compensated fairly for putting up the funds needed for purchases, while our partners are spared dilution of their equity interests. Because MidAmerican made no large acquisitions last year, it paid down \$100 million of what it owes us.

MidAmerican also owns a significant non-utility business, HomeServices of America, the second largest real estate broker in the country. Unlike our utility operations, this business is highly cyclical, but nevertheless one we view enthusiastically. We have an exceptional manager, Ron Peltier, who through both his acquisition and operational skills is building a brokerage powerhouse.

HomeServices participated in \$59.8 billion of transactions in 2004, a gain of \$11.2 billion from 2003. About 24% of the increase came from six acquisitions made during the year. Through our 17 brokerage firms – all of which retain their local identities – we employ more than 18,000 brokers in 18 states. HomeServices is almost certain to grow substantially in the next decade as we continue to acquire leading localized operations.

Last year MidAmerican wrote off a major investment in a zinc recovery project that was initiated in 1998 and became operational in 2002. Large quantities of zinc are present in the brine produced by our California geothermal operations, and we believed we could profitably extract the metal. For many months, it appeared that commercially-viable recoveries were imminent. But in mining, just as in oil exploration, prospects have a way of “teasing” their developers, and every time one problem was solved, another popped up. In September, we threw in the towel.

Our failure here illustrates the importance of a guideline – stay with simple propositions – that we usually apply in investments as well as operations. If only one variable is key to a decision, and the variable has a 90% chance of going your way, the chance for a successful outcome is obviously 90%. But if ten independent variables need to break favorably for a successful result, and each has a 90% probability of success, the likelihood of having a winner is only 35%. In our zinc venture, we solved most of the problems. But one proved intractable, and that was one too many. Since a chain is no stronger than its weakest link, it makes sense to look for – if you’ll excuse an oxymoron – mono-linked chains.

A breakdown of MidAmerican's results follows. In 2004, the “other” category includes a \$72.2 million profit from sale of an Enron receivable that was thrown in when we purchased Northern Natural two years earlier. Walter, Dave and I, as natives of Omaha, view this unanticipated gain as war reparations – partial compensation for the loss our city suffered in 1986 when Ken Lay moved Northern to Houston, after promising to leave the company here. (For details, see Berkshire’s 2002 annual report.)

Here are some key figures on MidAmerican's operations:

	Earnings (in $ millions)
	2004	2003
U.K. utilities	$326	$289
Iowa utility	268	269
Pipelines	288	261
HomeServices	130	113
Other (net)	172	190
Loss from zinc project	(579)	(46)
Earnings before corporate interest and taxes	605	1,076
Interest, other than to Berkshire	(212)	(225)
Interest on Berkshire junior debt	(170)	(184)
Income tax	(53)	(251)
Net earnings	$170	$416
Earnings applicable to Berkshire*	$237	$429
Debt owed to others	10,528	10,296
Debt owed to Berkshire	1,478	1,578

*Includes interest earned by Berkshire (net of related income taxes) of \$110 in 2004 and \$118 in 2003.

Insurance

Since Berkshire purchased National Indemnity (“NICO”) in 1967, property-casualty insurance has been our core business and the propellant of our growth. Insurance has provided a fountain of funds with which we’ve acquired the securities and businesses that now give us an ever-widening variety of earnings streams. So in this section, I will be spending a little time telling you how we got where we are.

The source of our insurance funds is “float,” which is money that doesn’t belong to us but that we temporarily hold. Most of our float arises because (1) premiums are paid upfront though the service we provide – insurance protection – is delivered over a period that usually covers a year and; (2) loss events that occur today do not always result in our immediately paying claims, because it sometimes takes many years for losses to be reported (asbestos losses would be an example), negotiated and settled. The \$20 million of float that came with our 1967 purchase has now increased – both by way of internal growth and acquisitions – to \$46.1 billion.

Float is wonderful – if it doesn't come at a high price. Its cost is determined by underwriting results, meaning how the expenses and losses we will ultimately pay compare with the premiums we have received. When an underwriting profit is achieved – as has been the case at Berkshire in about half of the 38 years we have been in the insurance business – float is better than free. In such years, we are actually paid for holding other people's money. For most insurers, however, life has been far more difficult: In aggregate, the property-casualty industry almost invariably operates at an underwriting loss. When that loss is large, float becomes expensive, sometimes devastatingly so.

Insurers have generally earned poor returns for a simple reason: They sell a commodity-like product. Policy forms are standard, and the product is available from many suppliers, some of whom are mutual companies (“owned” by policyholders rather than stockholders) with profit goals that are limited. Moreover, most insureds don’t care from whom they buy. Customers by the millions say “I need some Gillette blades” or “I’ll have a Coke” but we wait in vain for “I’d like a National Indemnity policy, please.” Consequently, price competition in insurance is usually fierce. Think airline seats.

So, you may ask, how do Berkshire's insurance operations overcome the dismal economics of the industry and achieve some measure of enduring competitive advantage? We've attacked that problem in several ways. Let's look first at NICO's strategy.

When we purchased the company – a specialist in commercial auto and general liability insurance – it did not appear to have any attributes that would overcome the industry’s chronic troubles. It was not well-known, had no informational advantage (the company has never had an actuary), was not a low-cost operator, and sold through general agents, a method many people thought outdated. Nevertheless, for almost all of the past 38 years, NICO has been a star performer. Indeed, had we not made this acquisition, Berkshire would be lucky to be worth half of what it is today.

What we've had going for us is a managerial mindset that most insurers find impossible to replicate. Take a look at the facing page. Can you imagine any public company embracing a business model that would lead to the decline in revenue that we experienced from 1986 through 1999? That colossal slide, it should be emphasized, did not occur because business was unobtainable. Many billions of premium dollars were readily available to NICO had we only been willing to cut prices. But we instead consistently priced to make a profit, not to match our most optimistic competitor. We never left customers – but they left us.

Most American businesses harbor an “institutional imperative” that rejects extended decreases in volume. What CEO wants to report to his shareholders that not only did business contract last year but that it will continue to drop? In insurance, the urge to keep writing business is also intensified because the consequences of foolishly-priced policies may not become apparent for some time. If an insurer is optimistic in its reserving, reported earnings will be overstated, and years may pass before true loss costs are revealed (a form of self-deception that nearly destroyed GEICO in the early 1970s).

Portrait of a Disciplined Underwriter
National Indemnity Company

Year		Written Premium(In $ millions)	No. ofEmployees atYear-End	Ratio ofOperating Expenses toWritten Premium	Underwriting Profit(Loss) as a Percentage of Premiums(Calculated as ofyear end 2004)*
1980	......	$79.6	372	32.3%	8.2%
1981	......	59.9	353	36.1%	(.8%)
1982	......	52.5	323	36.7%	(15.3%)
1983	......	58.2	308	35.6%	(18.7%)
1984	......	62.2	342	35.5%	(17.0%)
1985	......	160.7	380	28.0%	1.9%
1986	......	366.2	403	25.9%	30.7%
1987	......	232.3	368	29.5%	27.3%
1988	......	139.9	347	31.7%	24.8%
1989	......	98.4	320	35.9%	14.8%
1990	......	87.8	289	37.4%	7.0%
1991	......	88.3	284	35.7%	13.0%
1992	......	82.7	277	37.9%	5.2%
1993	......	86.8	279	36.1%	11.3%
1994	......	85.9	263	34.6%	4.6%
1995	......	78.0	258	36.6%	9.2%
1996	......	74.0	243	36.5%	6.8%
1997	......	65.3	240	40.4%	6.2%
1998	......	56.8	231	40.4%	9.4%
1999	......	54.5	222	41.2%	4.5%
2000	......	68.1	230	38.4%	2.9%
2001	......	161.3	254	28.8%	(11.6%)
2002	......	343.5	313	24.0%	16.8%
2003	......	594.5	337	22.2%	18.1%
2004	......	605.6	340	22.5%	5.1%

*It takes a long time to learn the true profitability of any given year. First, many claims are received after the end of the year, and we must estimate how many of these there will be and what they will cost. (In insurance jargon, these claims are termed IBNR – incurred but not reported.) Second, claims often take years, or even decades, to settle, which means there can be many surprises along the way.

For these reasons, the results in this column simply represent our best estimate at the end of 2004 as to how we have done in prior years. Profit margins for the years through 1999 are probably close to correct because these years are “mature,” in the sense that they have few claims still outstanding. The more recent the year, the more guesswork is involved. In particular, the results shown for 2003 and 2004 are apt to change significantly.

Finally, there is a fear factor at work, in that a shrinking business usually leads to layoffs. To avoid pink slips, employees will rationalize inadequate pricing, telling themselves that poorly-priced business must be tolerated in order to keep the organization intact and the distribution system happy. If this course isn't followed, these employees will argue, the company will not participate in the recovery that they invariably feel is just around the corner.

To combat employees' natural tendency to save their own skins, we have always promised NICO's workforce that no one will be fired because of declining volume, however severe the contraction. (This is not Donald Trump's sort of place.) NICO is not labor-intensive, and, as the table suggests, can live with excess overhead. It can't live, however, with underpriced business and the breakdown in underwriting discipline that accompanies it. An insurance organization that doesn't care deeply about underwriting at a profit this year is unlikely to care next year either.

Naturally, a business that follows a no-layoff policy must be especially careful to avoid overstaffing when times are good. Thirty years ago Tom Murphy, then CEO of Cap Cities, drove this point home to me with a hypothetical tale about an employee who asked his boss for permission to hire an assistant. The employee assumed that adding \$20,000 to the annual payroll would be inconsequential. But his boss told him the proposal should be evaluated as a \$3 million decision, given that an additional person would probably cost at least that amount over his lifetime, factoring in raises, benefits and other expenses (more people, more toilet paper). And unless the company fell on very hard times, the employee added would be unlikely to be dismissed, however marginal his contribution to the business.

It takes real fortitude – embedded deep within a company’s culture – to operate as NICO does. Anyone examining the table can scan the years from 1986 to 1999 quickly. But living day after day with dwindling volume – while competitors are boasting of growth and reaping Wall Street’s applause – is an experience few managers can tolerate. NICO, however, has had four CEOs since its formation in 1940 and none have bent. (It should be noted that only one of the four graduated from college. Our experience tells us that extraordinary business ability is largely innate.)

The current managerial star – make that superstar – at NICO is Don Wurster (yes, he’s “the graduate”), who has been running things since 1989. His slugging percentage is right up there with Barry Bonds’ because, like Barry, Don will accept a walk rather than swing at a bad pitch. Don has now amassed \$950 million of float at NICO that over time is almost certain to be proved the negative-cost kind. Because insurance prices are falling, Don’s volume will soon decline very significantly and, as it does, Charlie and I will applaud him ever more loudly.

* * * * * * * * * * * *

Another way to prosper in a commodity-type business is to be the low-cost operator. Among auto insurers operating on a broad scale, GEICO holds that cherished title. For NICO, as we have seen, an ebb-and-flow business model makes sense. But a company holding a low-cost advantage must pursue an unrelenting foot-to-the-floor strategy. And that's just what we do at GEICO.

A century ago, when autos first appeared, the property-casualty industry operated as a cartel. The major companies, most of which were based in the Northeast, established “bureau” rates and that was it. No one cut prices to attract business. Instead, insurers competed for strong, well-regarded agents, a focus that produced high commissions for agents and high prices for consumers.

In 1922, State Farm was formed by George Mecherle, a farmer from Merna, Illinois, who aimed to take advantage of the pricing umbrella maintained by the high-cost giants of the industry. State Farm employed a “captive” agency force, a system keeping its acquisition costs lower than those incurred by the bureau insurers (whose “independent” agents successfully played off one company against another). With its low-cost structure, State Farm eventually captured about 25% of the personal lines (auto and homeowners) business, far outdistancing its once-mighty competitors. Allstate, formed in 1931, put a similar distribution system into place and soon became the runner-up in personal lines to State Farm. Capitalism had worked its magic, and these low-cost operations looked unstoppable.

But a man named Leo Goodwin had an idea for an even more efficient auto insurer and, with a skimpy \$200,000, started GEICO in 1936. Goodwin's plan was to eliminate the agent entirely and to deal instead directly with the auto owner. Why, he asked himself, should there be any unnecessary and expensive links in the distribution mechanism when the product, auto insurance, was both mandatory and costly. Purchasers of business insurance, he reasoned, might well require professional advice, but most consumers knew what they needed in an auto policy. That was a powerful insight.

Originally, GEICO mailed its low-cost message to a limited audience of government employees. Later, it widened its horizons and shifted its marketing emphasis to the phone, working inquiries that came from broadcast and print advertising. And today the Internet is coming on strong.

Between 1936 and 1975, GEICO grew from a standing start to a $4\%$ market share, becoming the country's fourth largest auto insurer. During most of this period, the company was superbly managed, achieving both excellent volume gains and high profits. It looked unstoppable. But after my friend and hero Lorimer Davidson retired as CEO in 1970, his successors soon made a huge mistake by under-reserving for losses. This produced faulty cost information, which in turn produced inadequate pricing. By 1976, GEICO was on the brink of failure.

Jack Byrne then joined GEICO as CEO and, almost single-handedly, saved the company by heroic efforts that included major price increases. Though GEICO's survival required these, policyholders fled the company, and by 1980 its market share had fallen to $1.8\%$ . Subsequently, the company embarked on some unwise diversification moves. This shift of emphasis away from its extraordinary core business stunted GEICO's growth, and by 1993 its market share had grown only fractionally, to $1.9\%$ . Then Tony Nicely took charge.

And what a difference that's made: In 2005 GEICO will probably secure a 6% market share. Better yet, Tony has matched growth with profitability. Indeed, GEICO delivers all of its constituents major benefits: In 2004 its customers saved \$1 billion or so compared to what they would otherwise have paid for coverage, its associates earned a \$191 million profit-sharing bonus that averaged 24.3% of salary, and its owner – that's us – enjoyed excellent financial returns.

There's more good news. When Jack Byrne was rescuing the company in 1976, New Jersey refused to grant him the rates he needed to operate profitably. He therefore promptly – and properly – withdrew from the state. Subsequently, GEICO avoided both New Jersey and Massachusetts, recognizing them as two jurisdictions in which insurers were destined to struggle.

In 2003, however, New Jersey took a new look at its chronic auto-insurance problems and enacted legislation that would curb fraud and allow insurers a fair playing field. Even so, one might have expected the state's bureaucracy to make change slow and difficult.

But just the opposite occurred. Holly Bakke, the New Jersey insurance commissioner, who would be a success in any line of work, was determined to turn the law's intent into reality. With her staff's cooperation, GEICO ironed out the details for re-entering the state and was licensed last August. Since then, we've received a response from New Jersey drivers that is multiples of my expectations.

We are now serving 140,000 policyholders – about 4% of the New Jersey market – and saving them substantial sums (as we do drivers everywhere). Word-of-mouth recommendations within the state are causing inquiries to pour in. And once we hear from a New Jersey prospect, our closure rate – the percentage of policies issued to inquiries received – is far higher in the state than it is nationally.

We make no claim, of course, that we can save everyone money. Some companies, using rating systems that are different from ours, will offer certain classes of drivers a lower rate than we do. But we believe GEICO offers the lowest price more often than any other national company that serves all segments of the public. In addition, in most states, including New Jersey, Berkshire shareholders receive an 8% discount. So gamble fifteen minutes of your time and go to GEICO.com – or call 800-847-7536 – to see whether you can save big money (which you might want to use, of course, to buy other Berkshire products).

* * * * * * * * * * * *

Reinsurance – insurance sold to other insurers who wish to lay off part of the risks they have assumed – should not be a commodity product. At bottom, any insurance policy is simply a promise, and as everyone knows, promises vary enormously in their quality.

At the primary insurance level, nevertheless, just who makes the promise is often of minor importance. In personal-lines insurance, for example, states levy assessments on solvent companies to pay the policyholders of companies that go broke. In the business-insurance field, the same arrangement applies to workers' compensation policies. "Protected" policies of these types account for about 60% of the property-casualty industry's volume. Prudently-run insurers are irritated by the need to subsidize poor or reckless management elsewhere, but that's the way it is.

Other forms of business insurance at the primary level involve promises that carry greater risks for the insured. When Reliance Insurance and Home Insurance were run into the ground, for example, their promises proved to be worthless. Consequently, many holders of their business policies (other than those covering workers' compensation) suffered painful losses.

The solvency risk in primary policies, however, pales in comparison to that lurking in reinsurance policies. When a reinsurer goes broke, staggering losses almost always strike the primary companies it has dealt with. This risk is far from minor: GEICO has suffered tens of millions in losses from its careless selection of reinsurers in the early 1980s.

Were a true mega-catastrophe to occur in the next decade or two – and that’s a real possibility – some reinsurers would not survive. The largest insured loss to date is the World Trade Center disaster, which cost the insurance industry an estimated \$35 billion. Hurricane Andrew cost insurers about \$15.5 billion in 1992 (though that loss would be far higher in today’s dollars). Both events rocked the insurance and reinsurance world. But a \$100 billion event, or even a larger catastrophe, remains a possibility if either a particularly severe earthquake or hurricane hits just the wrong place. Four significant hurricanes struck Florida during 2004, causing an aggregate of \$25 billion or so in insured losses. Two of these – Charley and Ivan – could have done at least three times the damage they did had they entered the U.S. not far from their actual landing points.

Many insurers regard a \$100 billion industry loss as “unthinkable” and won’t even plan for it. But at Berkshire, we are fully prepared. Our share of the loss would probably be 3% to 5%, and earnings from our investments and other businesses would comfortably exceed that cost. When “the day after” arrives, Berkshire’s checks will clear.

Though the hurricanes hit us with a \$1.25 billion loss, our reinsurance operations did well last year. At General Re, Joe Brandon has restored a long-admired culture of underwriting discipline that, for a time, had lost its way. The excellent results he realized in 2004 on current business, however, were offset by adverse developments from the years before he took the helm. At NICO's reinsurance operation, Ajit Jain continues to successfully underwrite huge risks that no other reinsurer is willing or able to accept. Ajit's value to Berkshire is enormous.

* * * * * * * * * * * *

Our insurance managers, maximizing the competitive strengths I've mentioned in this section, again delivered first-class underwriting results last year. As a consequence, our float was better than costless. Here's the scorecard:

	(in $ millions)
	Underwriting Profit	Yearend Float
Insurance Operations	2004	2004	2003
General Re	$ 3	$23,120	$23,654
B-H Reinsurance	417	15,278	13,948
GEICO	970	5,960	5,287
Other Primary*	161	1,736	1,331
Total	$1,551	$46,094	$44,220

*Includes, in addition to National Indemnity, a variety of other exceptional insurance businesses, run by Rod Eldred, John Kizer, Tom Nerney and Don Towle.

Berkshire’s float increased \$1.9 billion in 2004, even though a few insureds opted to commute (that is, unwind) certain reinsurance contracts. We agree to such commutations only when we believe the economics are favorable to us (after giving due weight to what we might earn in the future on the money we are returning).

To summarize, last year we were paid more than \$1.5 billion to hold an average of about \$45.2 billion. In 2005 pricing will be less attractive than it has been. Nevertheless, absent a mega-catastrophe, we have a decent chance of achieving no-cost float again this year.

Finance and Finance Products

Last year in this section we discussed a potpourri of activities. In this report, we'll skip over several that are now of lesser importance: Berkadia is down to tag ends; Value Capital has added other investors, negating our expectation that we would need to consolidate its financials into ours; and the trading operation that I run continues to shrink.

• Both of Berkshire's leasing operations rebounded last year. At CORT (office furniture), earnings remain inadequate, but are trending upward. XTRA disposed of its container and intermodal businesses in order to concentrate on trailer leasing, long its strong suit. Overhead has been reduced, asset utilization is up and decent profits are now being achieved under Bill Franz, the company's new CEO.
• The wind-down of Gen Re Securities continues. We decided to exit this derivative operation three years ago, but getting out is easier said than done. Though derivative instruments are purported to be highly liquid – and though we have had the benefit of a benign market while liquidating ours – we still had 2,890 contracts outstanding at yearend, down from 23,218 at the peak. Like Hell, derivative trading is easy to enter but difficult to leave. (Other similarities come to mind as well.)

Gen Re's derivative contracts have always been required to be marked to market, and I believe the company's management conscientiously tried to make realistic “marks.” The market prices of derivatives, however, can be very fuzzy in a world in which settlement of a transaction is sometimes decades away and often involves multiple variables as well. In the interim the marks influence the managerial and trading bonuses that are paid annually. It’s small wonder that phantom profits are often recorded.

Investors should understand that in all types of financial institutions, rapid growth sometimes masks major underlying problems (and occasionally fraud). The real test of the earning power of a derivatives operation is what it achieves after operating for an extended period in a no-growth mode. You only learn who has been swimming naked when the tide goes out.

- After 40 years, we've finally generated a little synergy at Berkshire: Clayton Homes is doing well and that's in part due to its association with Berkshire. The manufactured home industry continues to reside in the intensive care unit of Corporate America, having sold less than 135,000 new homes last year, about the same as in 2003. Volume in these years was the lowest since 1962, and it was also only about $40\%$ of annual sales during the years 1995-99. That era, characterized by irresponsible financing and naïve funders, was a fool's paradise for the industry.

Because one major lender after another has fled the field, financing continues to bedevil manufacturers, retailers and purchasers of manufactured homes. Here Berkshire's support has proven valuable to Clayton. We stand ready to fund whatever makes sense, and last year Clayton's management found much that qualified.

As we explained in our 2003 report, we believe in using borrowed money to support profitable, interest-bearing receivables. At the beginning of last year, we had borrowed \$2 billion to relend to Clayton (at a one percentage-point markup) and by January 2005 the total was \$7.35 billion. Most of the dollars added were borrowed by us on January 4, 2005, to finance a seasoned portfolio that Clayton purchased on December 30, 2004 from a bank exiting the business.

We now have two additional portfolio purchases in the works, totaling about \$1.6 billion, but it's quite unlikely that we will secure others of any significance. Therefore, Clayton's receivables (in which originations will roughly offset payoffs) will probably hover around \$9 billion for some time and should deliver steady earnings. This pattern will be far different from that of the past, in which Clayton, like all major players in its industry, “securitized” its receivables, causing earnings to be front-ended. In the last two years, the securitization market has dried up. The limited funds available today come only at higher cost and with harsh terms. Had Clayton remained independent in this period, it would have had mediocre earnings as it struggled with financing.

In April, Clayton completed the acquisition of Oakwood Homes and is now the industry's largest producer and retailer of manufactured homes. We love putting more assets in the hands of Kevin Clayton, the company's CEO. He is a prototype Berkshire manager. Today, Clayton has 11,837 employees, up from 7,136 when we purchased it, and Charlie and I are pleased that Berkshire has been useful in facilitating this growth.

For simplicity's sake, we include all of Clayton's earnings in this sector, though a sizable portion of these are derived from areas other than consumer finance.

	(in $ millions)
	Pre-Tax Earnings		Interest-Bearing Liabilities
	2004	2003	2004	2003
Trading – ordinary income	$ 264	$ 355	$5,751	$7,826
Gen Re Securities	(44)	(99)	5,437*	8,041*
Life and annuity operation	(57)	85	2,467	2,331
Value Capital	30	31	N/A	N/A
Berkadia	1	101	—	525
Leasing operations	92	34	391	482
Manufactured housing finance (Clayton)	220	37**	3,636	2,032
Other	78	75	N/A	N/A
Income before capital gains	584	619
Trading – capital gains	1,750	1,215
Total	$2,334	$1,834

* Includes all liabilities
** From date of acquisition, August 7, 2003

Manufacturing, Service and Retailing Operations

Our activities in this category cover the waterfront. But let's look at a summary balance sheet and earnings statement consolidating the entire group.

Balance Sheet 12/31/04 (in \$ millions)

Assets		Liabilities and Equity
Cash and equivalents	$899	Notes payable	$1,143
Accounts and notes receivable	3,074	Other current liabilities	4,685
Inventory	3,842	Total current liabilities	5,828
Other current assets	254
Total current assets	8,069
Goodwill and other intangibles	8,362	Deferred taxes	248
Fixed assets	6,161	Term debt and other liabilities	1,965
Other assets	1,044	Equity	15,595
	$23,636		$23,636

Earnings Statement (in \$ millions)

	2004	2003
Revenues	$44,142	$32,106
Operating expenses (including depreciation of $676 in 2004 and $605 in 2003)	41,604	29,885
Interest expense (net)	57	64
Pre-tax earnings	2,481	2,157
Income taxes	941	813
Net income	$1,540	$1,344

This eclectic group, which sells products ranging from Dilly Bars to fractional interests in Boeing 737s, earned a very respectable 21.7% on average tangible net worth last year, compared to 20.7% in 2003. It’s noteworthy that these operations used only minor financial leverage in achieving these returns. Clearly, we own some very good businesses. We purchased many of them, however, at substantial premiums to net worth – a matter that is reflected in the goodwill item shown on the balance sheet – and that fact reduces the earnings on our average carrying value to 9.9%.

Here are the pre-tax earnings for the larger categories or units.

	Pre-Tax Earnings(in $ millions)
	2004	2003
Building Products	$643	$559
Shaw Industries	466	436
Apparel & Footwear	325	289
Retailing of Jewelry, Home Furnishings and Candy	215	224
Flight Services	191	72
McLane	228	150*
Other businesses	413	427
	$2,481	$2,157

* From date of acquisition, May 23, 2003.

- In the building-products sector and at Shaw, we’ve experienced staggering cost increases for both raw-materials and energy. By December, for example, steel costs at MiTek (whose primary business is connectors for roof trusses) were running 100% over a year earlier. And MiTek uses 665 million pounds of steel every year. Nevertheless, the company continues to be an outstanding performer.

Since we purchased MiTek in 2001, Gene Toombs, its CEO, has made some brilliant “bolt-on” acquisitions and is on his way to creating a mini-Berkshire.

Shaw fielded a barrage of price increases in its main fiber materials during the year, a hit that added more than \$300 million to its costs. (When you walk on carpet you are, in effect, stepping on processed oil.) Though we followed these hikes in costs with price increases of our own, there was an inevitable lag. Therefore, margins narrowed as the year progressed and remain under pressure today. Despite these roadblocks, Shaw, led by Bob Shaw and Julian Saul, earned an outstanding 25.6% on tangible equity in 2004. The company is a powerhouse and has a bright future.

- In apparel, Fruit of the Loom increased unit sales by 10 million dozen, or $14\%$ , with shipments of intimate apparel for women and girls growing by $31\%$ . Charlie, who is far more knowledgeable than I am on this subject, assures me that women are not wearing more underwear. With this expert input, I can only conclude that our market share in the women's category must be growing rapidly. Thanks to John Holland, Fruit is on the move.

A smaller operation, Garan, also had an excellent year. Led by Seymour Lichtenstein and Jerry Kamiel, this company manufactures the popular Garanimals line for children. Next time you are in a Wal-Mart, check out this imaginative product.

- Among our retailers, Ben Bridge (jewelry) and R. C. Willey (home furnishings) were particular standouts last year.

At Ben Bridge same-store sales grew $11.4\%$ , the best gain among the publicly-held jewelers whose reports I have seen. Additionally, the company's profit margin widened. Last year was not a fluke: During the past decade, the same-store sales gains of the company have averaged $8.8\%$ .

Ed and Jon Bridge are fourth-generation managers and run the business exactly as if it were their own – which it is in every respect except for Berkshire’s name on the stock certificates. The Bridges have expanded successfully by securing the right locations and, more importantly, by staffing these stores with enthusiastic and knowledgeable associates. We will move into Minneapolis-St. Paul this year.

At Utah-based R. C. Willey, the gains from expansion have been even more dramatic, with 41.9% of 2004 sales coming from out-of-state stores that didn't exist before 1999. The company also improved its profit margin in 2004, propelled by its two new stores in Las Vegas.

I would like to tell you that these stores were my idea. In truth, I thought they were mistakes. I knew, of course, how brilliantly Bill Child had run the R. C. Willey operation in Utah, where its market share had long been huge. But I felt our closed-on-Sunday policy would prove disastrous away from home. Even our first out-of-state store in Boise, which was highly successful, left me unconvinced. I kept asking whether Las Vegas residents, conditioned to seven-day-a-week retailers, would adjust to us. Our first Las Vegas store, opened in 2001, answered this question in a resounding manner, immediately becoming our number one unit.

Bill and Scott Hymas, his successor as CEO, then proposed a second Las Vegas store, only about 20 minutes away. I felt this expansion would cannibalize the first unit, adding significant costs but only modest sales. The result? Each store is now doing about 26% more volume than any other store in the chain and is consistently showing large year-over-year gains.

R. C. Willey will soon open in Reno. Before making this commitment, Bill and Scott again asked for my advice. Initially, I was pretty puffed up about the fact that they were consulting me. But then it dawned on me that the opinion of someone who is always wrong has its own special utility to decision-makers.

- Earnings improved in flight services. At FlightSafety, the world's leader in pilot training, profits rose as corporate aviation rebounded and our business with regional airlines increased. We now operate 283 simulators with an original cost of \$1.2 billion. Pilots are trained one at a time on this expensive equipment. This means that as much as \$3.50 of capital investment is required to produce \$1 of annual revenue. With this level of capital intensity, FlightSafety requires very high operating margins in order to obtain reasonable returns on capital, which means that utilization rates are all-important. Last year, FlightSafety's return on tangible equity improved to 15.1% from 8.4% in 2003.

In another 2004 event, Al Ueltschi, who founded FlightSafety in 1951 with \$10,000, turned over the CEO position to Bruce Whitman, a 43-year veteran at the company. (But Al's not going anywhere; I won't let him.) Bruce shares Al's conviction that flying an aircraft is a privilege to be extended only to people who regularly receive the highest quality of training and are undeniably competent. A few years ago, Charlie was asked to intervene with Al on behalf of a tycoon friend whom FlightSafety had flunked. Al's reply to Charlie: "Tell your pal he belongs in the back of the plane, not the cockpit."

FlightSafety's number one customer is NetJets, our aircraft fractional-ownership subsidiary. Its 2,100 pilots spend an average of 18 days a year in training. Additionally, these pilots fly only one aircraft type whereas many flight operations juggle pilots among several types. NetJets' high standards on both fronts are two of the reasons I signed up with the company years before Berkshire bought it.

Fully as important in my decisions to both use and buy NetJets, however, was the fact that the company was managed by Rich Santulli, the creator of the fractional-ownership industry and a fanatic about safety and service. I viewed the selection of a flight provider as akin to picking a brain surgeon: you simply want the best. (Let someone else experiment with the low bidder.)

Last year NetJets again gained about 70% of the net new business (measured by dollar value) going to the four companies that dominate the industry. A portion of our growth came from the 25-hour card offered by Marquis Jet Partners. Marquis is not owned by NetJets, but is instead a customer that repackages the purchases it makes from us into smaller packages that it sells through its card. Marquis deals exclusively with NetJets, utilizing the power of our reputation in its marketing.

Our U.S. contracts, including Marquis customers, grew from 3,877 to 4,967 in 2004 (versus approximately 1,200 contracts when Berkshire bought NetJets in 1998). Some clients (including me) enter into multiple contracts because they wish to use more than one type of aircraft, selecting for any given trip whichever type best fits the mission at hand.

NetJets earned a modest amount in the U.S. last year. But what we earned domestically was largely offset by losses in Europe. We are now, however, generating real momentum abroad. Contracts (including 25-hour cards that we ourselves market in Europe) increased from 364 to 693 during the year. We will again have a very significant European loss in 2005, but domestic earnings will likely put us in the black overall.

Europe has been expensive for NetJets – far more expensive than I anticipated – but it is essential to building a flight operation that will forever be in a class by itself. Our U.S. owners already want a quality service wherever they travel and their wish for flight hours abroad is certain to grow dramatically in the decades ahead. Last year, U.S. owners made 2,003 flights in Europe, up 22% from the previous year and 137% from 2000. Just as important, our European owners made 1,067 flights in the U.S., up 65% from 2003 and 239% from 2000.

Investments

We show below our common stock investments. Those that had a market value of more than \$600 million at the end of 2004 are itemized.

Shares	Company	Percentage of Company Owned	12/31/04
			Cost*	Market(in $ millions)
151,610,700	American Express Company	12.1	$1,470	$ 8,546
200,000,000	The Coca-Cola Company	8.3	1,299	8,328
96,000,000	The Gillette Company	9.7	600	4,299
14,350,600	H&R Block, Inc	8.7	223	703
6,708,760	M&T Bank Corporation	5.8	103	723
24,000,000	Moody’s Corporation	16.2	499	2,084
2,338,961,000	PetroChina “H” shares (or equivalents)	1.3	488	1,249
1,727,765	The Washington Post Company	18.1	11	1,698
56,448,380	Wells Fargo & Company	3.3	463	3,508
1,724,200	White Mountains Insurance	16.0	369	1,114
	Others		3,531	5,465
	Total Common Stocks		$9,056	$37,717

*This is our actual purchase price and also our tax basis; GAAP “cost” differs in a few cases because of write-ups or write-downs that have been required.

Some people may look at this table and view it as a list of stocks to be bought and sold based upon chart patterns, brokers' opinions, or estimates of near-term earnings. Charlie and I ignore such distractions and instead view our holdings as fractional ownerships in businesses. This is an important distinction. Indeed, this thinking has been the cornerstone of my investment behavior since I was 19. At that time I read Ben Graham's The Intelligent Investor, and the scales fell from my eyes. (Previously, I had been entranced by the stock market, but didn't have a clue about how to invest.)

Let's look at how the $businesses$ of our “Big Four” – American Express, Coca-Cola, Gillette and Wells Fargo – have fared since we bought into these companies. As the table shows, we invested \$3.83 billion in the four, by way of multiple transactions between May 1988 and October 2003. On a composite basis, our dollar-weighted purchase date is July 1992. By yearend 2004, therefore, we had held these “business interests,” on a weighted basis, about 12 $^{1/2}$ years.

In 2004, Berkshire's share of the group's earnings amounted to \$1.2 billion. These earnings might legitimately be considered “normal.” True, they were swelled because Gillette and Wells Fargo omitted option costs in their presentation of earnings; but on the other hand they were reduced because Coke had a non-recurring write-off.

Our share of the earnings of these four companies has grown almost every year, and now amounts to about 31.3% of our cost. Their cash distributions to us have also grown consistently, totaling \$434 million in 2004, or about 11.3% of cost. All in all, the Big Four have delivered us a satisfactory, though far from spectacular, business result.

That's true as well of our experience in the market with the group. Since our original purchases, valuation gains have somewhat exceeded earnings growth because price/earnings ratios have increased. On a year-to-year basis, however, the business and market performances have often diverged, sometimes to an extraordinary degree. During The Great Bubble, market-value gains far outstripped the performance of the businesses. In the aftermath of the Bubble, the reverse was true.

Clearly, Berkshire's results would have been far better if I had caught this swing of the pendulum. That may seem easy to do when one looks through an always-clean, rear-view mirror. Unfortunately, however, it's the windshield through which investors must peer, and that glass is invariably fogged. Our huge positions add to the difficulty of our nimbly dancing in and out of holdings as valuations swing.

Nevertheless, I can properly be criticized for merely clucking about nose-bleed valuations during the Bubble rather than acting on my views. Though I said at the time that certain of the stocks we held were priced ahead of themselves, I underestimated just how severe the overvaluation was. I talked when I should have walked.

What Charlie and I would like is a little action now. We don't enjoy sitting on \$43 billion of cash equivalents that are earning paltry returns. Instead, we yearn to buy more fractional interests similar to those we now own or – better still – more large businesses outright. We will do either, however, only when purchases can be made at prices that offer us the prospect of a reasonable return on our investment.

* * * * * * * * * * * *

We’ve repeatedly emphasized that the “realized” gains that we report quarterly or annually are meaningless for analytical purposes. We have a huge amount of unrealized gains on our books, and our thinking about when, and if, to cash them depends not at all on a desire to report earnings at one specific time or another. A further complication in our reported gains occurs because GAAP requires that foreign exchange contracts be marked to market, a stipulation that causes unrealized gains or losses in these holdings to flow through our published earnings as if we had sold our positions.

Despite the problems enumerated, you may be interested in a breakdown of the gains we reported in 2003 and 2004. The data reflect actual sales except in the case of currency gains, which are a combination of sales and marks to market.

Category	Pre-Tax Gain (in $ millions)
	2004	2003
Common Stocks	$ 870	$ 448
U.S. Government Bonds	104	1,485
Junk Bonds	730	1,138
Foreign Exchange Contracts	1,839	825
Other	(47)	233
Total	$3,496	$4,129

The junk bond profits include a foreign exchange component. When we bought these bonds in 2001 and 2002, we focused first, of course, on the credit quality of the issuers, all of which were American corporations. Some of these companies, however, had issued bonds denominated in foreign currencies. Because of our views on the dollar, we favored these for purchase when they were available.

As an example, we bought €254 million of Level 3 bonds (10 $\frac{3}{4}$ % of 2008) in 2001 at 51.7% of par, and sold these at 85% of par in December 2004. This issue was traded in Euros that cost us 88¢ at the time of purchase but that brought \$1.29 when we sold. Thus, of our \$163 million overall gain, about \$85 million came from the market's revised opinion about Level 3's credit quality, with the remaining \$78 million resulting from the appreciation of the Euro. (In addition, we received cash interest during our holding period that amounted to about 25% annually on our dollar cost.)

* * * * * * * * * * * *

The media continue to report that “Buffett buys” this or that stock. Statements like these are almost always based on filings Berkshire makes with the SEC and are therefore wrong. As I’ve said before, the stories should say “Berkshire buys.”

Portrait of a Disciplined Investor

Lou Simpson

Year		Return from GEICO Equities	S&P Return	Relative Results
1980	....	23.7%	32.3%	(8.6%)
1981	....	5.4%	(5.0%)	10.4%
1982	....	45.8%	21.4%	24.4%
1983	....	36.0%	22.4%	13.6%
1984	....	21.8%	6.1%	15.7%
1985	....	45.8%	31.6%	14.2%
1986	....	38.7%	18.6%	20.1%
1987	....	(10.0%)	5.1%	(15.1%)
1988	....	30.0%	16.6%	13.4%
1989	....	36.1%	31.7%	4.4%
1990	....	(9.9%)	(3.1%)	(6.8%)
1991	....	56.5%	30.5%	26.0%
1992	....	10.8%	7.6%	3.2%
1993	....	4.6%	10.1%	(5.5%)
1994	....	13.4%	1.3%	12.1%
1995	....	39.8%	37.6%	2.2%
1996	....	29.2%	23.0%	6.2%
1997	....	24.6%	33.4%	(8.8%)
1998	....	18.6%	28.6%	(10.0%)
1999	....	7.2%	21.0%	(13.8%)
2000	....	20.9%	(9.1%)	30.0%
2001	....	5.2%	(11.9%)	17.1%
2002	....	(8.1%)	(22.1%)	14.0%
2003	....	38.3%	28.7%	9.6%
2004	....	16.9%	10.9%	6.0%
Average Annual Gain 1980-2004		20.3%	13.5%	6.8%

Even then, it is typically not I who make the buying decisions. Lou Simpson manages about $2 \frac{1}{2}$ billion of equities that are held by GEICO, and it is his transactions that Berkshire is usually reporting. Customarily his purchases are in the \$200-\$300 million range and are in companies that are smaller than the ones I focus on. Take a look at the facing page to see why Lou is a cinch to be inducted into the investment Hall of Fame.

You may be surprised to learn that Lou does not necessarily inform me about what he is doing. When Charlie and I assign responsibility, we truly hand over the baton – and we give it to Lou just as we do to our operating managers. Therefore, I typically learn of Lou’s transactions about ten days after the end of each month. Sometimes, it should be added, I silently disagree with his decisions. But he’s usually right.

Foreign Currencies

Berkshire owned about \$21.4 billion of foreign exchange contracts at yearend, spread among 12 currencies. As I mentioned last year, holdings of this kind are a decided change for us. Before March 2002, neither Berkshire nor I had ever traded in currencies. But the evidence grows that our trade policies will put unremitting pressure on the dollar for many years to come – so since 2002 we’ve heeded that warning in setting our investment course. (As W.C. Fields once said when asked for a handout: “Sorry, son, all my money’s tied up in currency.”)

Be clear on one point: In no way does our thinking about currencies rest on doubts about America. We live in an extraordinarily rich country, the product of a system that values market economics, the rule of law and equality of opportunity. Our economy is far and away the strongest in the world and will continue to be. We are lucky to live here.

But as I argued in a November 10, 2003 article in Fortune, (available at berkshirehathaway.com), our country's trade practices are weighing down the dollar. The decline in its value has already been substantial, but is nevertheless likely to continue. Without policy changes, currency markets could even become disorderly and generate spillover effects, both political and financial. No one knows whether these problems will materialize. But such a scenario is a far-from-remote possibility that policymakers should be considering now. Their bent, however, is to lean toward not-so-benign neglect: A 318-page Congressional study of the consequences of unremitting trade deficits was published in November 2000 and has been gathering dust ever since. The study was ordered after the deficit hit a then-alarming \$263 billion in 1999; by last year it had risen to \$618 billion.

Charlie and I, it should be emphasized, believe that true trade – that is, the exchange of goods and services with other countries – is enormously beneficial for both us and them. Last year we had \$1.15 trillion of such honest-to-God trade and the more of this, the better. But, as noted, our country also purchased an additional \$618 billion in goods and services from the rest of the world that was unreciprocated. That is a staggering figure and one that has important consequences.

The balancing item to this one-way pseudo-trade — in economics there is always an offset — is a transfer of wealth from the U.S. to the rest of the world. The transfer may materialize in the form of IOUs our private or governmental institutions give to foreigners, or by way of their assuming ownership of our assets, such as stocks and real estate. In either case, Americans end up owning a reduced portion of our country while non-Americans own a greater part. This force-feeding of American wealth to the rest of the world is now proceeding at the rate of \$1.8 billion daily, an increase of 20% since I wrote you last year. Consequently, other countries and their citizens now own a net of about \$3 trillion of the U.S. A decade ago their net ownership was negligible.

The mention of trillions numbs most brains. A further source of confusion is that the current account deficit (the sum of three items, the most important by far being the trade deficit) and our national budget deficit are often lumped as “twins.” They are anything but. They have different causes and different consequences.

A budget deficit in no way reduces the portion of the national pie that goes to Americans. As long as other countries and their citizens have no net ownership of the U.S., $100\%$ of our country's output belongs to our citizens under any budget scenario, even one involving a huge deficit.

As a rich “family” awash in goods, Americans will argue through their legislators as to how government should redistribute the national output – that is who pays taxes and who receives governmental benefits. If “entitlement” promises from an earlier day have to be reexamined, “family members” will angrily debate among themselves as to who feels the pain. Maybe taxes will go up; maybe promises will be modified; maybe more internal debt will be issued. But when the fight is finished, all of the family’s huge pie remains available for its members, however it is divided. No slice must be sent abroad.

Large and persisting current account deficits produce an entirely different result. As time passes, and as claims against us grow, we own less and less of what we produce. In effect, the rest of the world enjoys an ever-growing royalty on American output. Here, we are like a family that consistently overspends its income. As time passes, the family finds that it is working more and more for the “finance company” and less for itself.

Should we continue to run current account deficits comparable to those now prevailing, the net ownership of the U.S. by other countries and their citizens a decade from now will amount to roughly \$11 trillion. And, if foreign investors were to earn only 5% on that net holding, we would need to send a net of \$.55 trillion of goods and services abroad every year merely to service the U.S. investments then held by foreigners. At that date, a decade out, our GDP would probably total about \$18 trillion (assuming low inflation, which is far from a sure thing). Therefore, our U.S. “family” would then be delivering 3% of its annual output to the rest of the world simply as tribute for the overindulgences of the past. In this case, unlike that involving budget deficits, the sons would truly pay for the sins of their fathers.

This annual royalty paid the world – which would not disappear unless the U.S. massively underconsumed and began to run consistent and large trade surpluses – would undoubtedly produce significant political unrest in the U.S. Americans would still be living very well, indeed better than now because of the growth in our economy. But they would chafe at the idea of perpetually paying tribute to their creditors and owners abroad. A country that is now aspiring to an “Ownership Society” will not find happiness in – and I’ll use hyperbole here for emphasis – a “Sharecropper’s Society.” But that’s precisely where our trade policies, supported by Republicans and Democrats alike, are taking us.

Many prominent U.S. financial figures, both in and out of government, have stated that our current-account deficits cannot persist. For instance, the minutes of the Federal Reserve Open Market Committee of June 29-30, 2004 say: “The staff noted that outsized external deficits could not be sustained indefinitely.” But, despite the constant handwringing by luminaries, they offer no substantive suggestions to tame the burgeoning imbalance.

In the article I wrote for Fortune 16 months ago, I warned that “a gently declining dollar would not provide the answer.” And so far it hasn’t. Yet policymakers continue to hope for a “soft landing,” meanwhile counseling other countries to stimulate (read “inflate”) their economies and Americans to save more. In my view these admonitions miss the mark: There are deep-rooted structural problems that will cause America to continue to run a huge current-account deficit unless trade policies either change materially or the dollar declines by a degree that could prove unsettling to financial markets.

Proponents of the trade status quo are fond of quoting Adam Smith: “What is prudence in the conduct of every family can scarce be folly in that of a great kingdom. If a foreign country can supply us with a commodity cheaper than we ourselves can make it, better buy it of them with some part of the produce of our own industry, employed in a way in which we have some advantage.”

I agree. Note, however, that Mr. Smith's statement refers to trade of product for product, not of wealth for product as our country is doing to the tune of \$.6 trillion annually. Moreover, I am sure that he would never have suggested that “prudence” consisted of his “family” selling off part of its farm every day in order to finance its overconsumption. Yet that is just what the “great kingdom” called the United States is doing.

If the U.S. was running a \$.6 trillion current-account surplus, commentators worldwide would violently condemn our policy, viewing it as an extreme form of “mercantilism” – a long-discredited economic strategy under which countries fostered exports, discouraged imports, and piled up treasure. I would condemn such a policy as well. But, in effect if not in intent, the rest of the world is practicing mercantilism in respect to the U.S., an act made possible by our vast store of assets and our pristine credit history. Indeed, the world would never let any other country use a credit card denominated in its own currency to the insatiable extent we are employing ours. Presently, most foreign investors are sanguine: they may view us as spending junkies, but they know we are rich junkies as well.

Our spendthrift behavior won't, however, be tolerated indefinitely. And though it's impossible to forecast just when and how the trade problem will be resolved, it's improbable that the resolution will foster an increase in the value of our currency relative to that of our trading partners.

We hope the U.S. adopts policies that will quickly and substantially reduce the current-account deficit. True, a prompt solution would likely cause Berkshire to record losses on its foreign-exchange contracts. But Berkshire's resources remain heavily concentrated in dollar-based assets, and both a strong dollar and a low-inflation environment are very much in our interest.

If you wish to keep abreast of trade and currency matters, read The Financial Times. This London-based paper has long been the leading source for daily international financial news and now has an excellent American edition. Both its reporting and commentary on trade are first-class.

* * * * * * * * * * * *

And, again, our usual caveat: macro-economics is a tough game in which few people, Charlie and I included, have demonstrated skill. We may well turn out to be wrong in our currency judgments. (Indeed, the fact that so many pundits now predict weakness for the dollar makes us uneasy.) If so, our mistake will be very public. The irony is that if we chose the opposite course, leaving all of Berkshire's assets in dollars even as they declined significantly in value, no one would notice our mistake.

John Maynard Keynes said in his masterful The General Theory: “Worldly wisdom teaches that it is better for reputation to fail conventionally than to succeed unconventionally.” (Or, to put it in less elegant terms, lemmings as a class may be derided but never does an individual lemming get criticized.) From a reputational standpoint, Charlie and I run a clear risk with our foreign-exchange commitment. But we believe in managing Berkshire as if we owned 100% of it ourselves. And, were that the case, we would not be following a dollar-only policy.

Miscellaneous

- Last year I told you about a group of University of Tennessee finance students who played a key role in our \$1.7 billion acquisition of Clayton Homes. Earlier, they had been brought to Omaha by their professor, Al Auxier – he brings a class every year – to tour Nebraska Furniture Mart and Borsheim's, eat at Gorat's and have a Q&A session with me at Kiewit Plaza. These visitors, like those who come for our annual meeting, leave impressed by both the city and its friendly residents.

Other colleges and universities have now come calling. This school year we will have visiting classes, ranging in size from 30 to 100 students, from Chicago, Dartmouth (Tuck), Delaware State, Florida State, Indiana, Iowa, Iowa State, Maryland, Nebraska, Northwest Nazarene, Pennsylvania (Wharton), Stanford, Tennessee, Texas, Texas A&M, Toronto (Rotman), Union and Utah. Most of the students are MBA candidates, and I've been impressed by their quality. They are keenly interested in business and investments, but their questions indicate that they also have more on their minds than simply making money. I always feel good after meeting them.

At our sessions, I tell the newcomers the story of the Tennessee group and its spotting of Clayton Homes. I do this in the spirit of the farmer who enters his hen house with an ostrich egg and admonishes the flock: “I don’t like to complain, girls, but this is just a small sample of what the competition is doing.” To date, our new scouts have not brought us deals. But their mission in life has been made clear to them.

- You should be aware of an accounting rule that mildly distorts our financial statements in a pain-today, gain-tomorrow manner. Berkshire purchases life insurance policies from individuals and corporations who would otherwise surrender them for cash. As the new holder of the policies, we pay any premiums that become due and ultimately – when the original holder dies – collect the face value of the policies.

The original policyholder is usually in good health when we purchase the policy. Still, the price we pay for it is always well above its cash surrender value (“CSV”). Sometimes the original policyholder has borrowed against the CSV to make premium payments. In that case, the remaining CSV will be tiny and our purchase price will be a large multiple of what the original policyholder would have received, had he cashed out by surrendering it.

Under accounting rules, we must immediately charge as a realized capital loss the excess over CSV that we pay upon purchasing the policy. We also must make additional charges each year for the amount by which the premium we pay to keep the policy in force exceeds the increase in CSV. But obviously, we don't think these bookkeeping charges represent economic losses. If we did, we wouldn't buy the policies.

During 2004, we recorded net “losses” from the purchase of policies (and from the premium payments required to maintain them) totaling \$207 million, which was charged against realized investment gains in our earnings statement (included in “other” in the table on page 17). When the proceeds from these policies are received in the future, we will record as realized investment gain the excess over the then-CSV.

- Two post-bubble governance reforms have been particularly useful at Berkshire, and I fault myself for not putting them in place many years ago. The first involves regular meetings of directors without the CEO present. I’ve sat on 19 boards, and on many occasions this process would have led to dubious plans being examined more thoroughly. In a few cases, CEO changes that were needed would also have been made more promptly. There is no downside to this process, and there are many possible benefits.

The second reform concerns the “whistleblower line,” an arrangement through which employees can send information to me and the board’s audit committee without fear of reprisal. Berkshire’s extreme decentralization makes this system particularly valuable both to me and the committee. (In a sprawling “city” of 180,000 – Berkshire’s current employee count – not every sparrow that falls will be noticed at headquarters.) Most of the complaints we have received are of “the guy next to me has bad breath” variety, but on occasion I have learned of important problems at our subsidiaries that I otherwise would have missed. The issues raised are usually not of a type discoverable by audit, but relate instead to personnel and business practices. Berkshire would be more valuable today if I had put in a whistleblower line decades ago.

- Charlie and I love the idea of shareholders thinking and behaving like owners. Sometimes that requires them to be pro-active. And in this arena large institutional owners should lead the way.

So far, however, the moves made by institutions have been less than awe-inspiring. Usually, they've focused on minutiae and ignored the three questions that truly count. First, does the company have the right CEO? Second, is he/she overreaching in terms of compensation? Third, are proposed acquisitions more likely to create or destroy per-share value?

On such questions, the interests of the CEO may well differ from those of the shareholders. Directors, moreover, sometimes lack the knowledge or gumption to overrule the CEO. Therefore, it’s vital that large owners focus on these three questions and speak up when necessary.

Instead many simply follow a “checklist” approach to the issue du jour. Last year I was on the receiving end of a judgment reached in that manner. Several institutional shareholders and their advisors decided I lacked “independence” in my role as a director of Coca-Cola. One group wanted me removed from the board and another simply wanted me booted from the audit committee.

My first impulse was to secretly fund the group behind the second idea. Why anyone would wish to be on an audit committee is beyond me. But since directors must be assigned to one committee or another, and since no CEO wants me on his compensation committee, it's often been my lot to get an audit committee assignment. As it turned out, the institutions that opposed me failed and I was re-elected to the audit job. (I fought off the urge to ask for a recount.)

Some institutions questioned my “independence” because, among other things, McLane and Dairy Queen buy lots of Coke products. (Do they want us to favor Pepsi?) But independence is defined in Webster’s as “not subject to control by others.” I’m puzzled how anyone could conclude that our Coke purchases would “control” my decision-making when the counterweight is the well-being of \$8 billion of Coke stock held by Berkshire. Assuming I’m even marginally rational, elementary arithmetic should make it clear that my heart and mind belong to the owners of Coke, not to its management.

I can't resist mentioning that Jesus understood the calibration of independence far more clearly than do the protesting institutions. In Matthew 6:21 He observed: “For where your treasure is, there will your heart be also.” Even to an institutional investor, \$8 billion should qualify as “treasure” that dwarfs any profits Berkshire might earn on its routine transactions with Coke.

Measured by the biblical standard, the Berkshire board is a model: (a) every director is a member of a family owning at least \$4 million of stock; (b) none of these shares were acquired from Berkshire via options or grants; (c) no directors receive committee, consulting or board fees from the company that are more than a tiny portion of their annual income; and (d) although we have a standard corporate indemnity arrangement, we carry no liability insurance for directors.

At Berkshire, board members travel the same road as shareholders.

* * * * * * * * * * * *

Charlie and I have seen much behavior confirming the Bible's “treasure” point. In our view, based on our considerable boardroom experience, the least independent directors are likely to be those who receive an important fraction of their annual income from the fees they receive for board service (and who hope as well to be recommended for election to other boards and thereby to boost their income further). Yet these are the very board members most often classed as “independent.”

Most directors of this type are decent people and do a first-class job. But they wouldn't be human if they weren't tempted to thwart actions that would threaten their livelihood. Some may go on to succumb to such temptations.

Let's look at an example based upon circumstantial evidence. I have first-hand knowledge of a recent acquisition proposal (not from Berkshire) that was favored by management, blessed by the company's investment banker and slated to go forward at a price above the level at which the stock had sold for some years (or now sells for). In addition, a number of directors favored the transaction and wanted it proposed to shareholders.

Several of their brethren, however, each of whom received board and committee fees totaling about \$100,000 annually, scuttled the proposal, which meant that shareholders never learned of this multi-billion offer. Non-management directors owned little stock except for shares they had received from the company. Their open-market purchases in recent years had meanwhile been nominal, even though the stock had sold far below the acquisition price proposed. In other words, these directors didn't want the shareholders to be offered X even though they had consistently declined the opportunity to buy stock for their own account at a fraction of X.

I don't know which directors opposed letting shareholders see the offer. But I do know that \$100,000 is an important portion of the annual income of some of those deemed “independent,” clearly meeting the Matthew 6:21 definition of “treasure.” If the deal had gone through, these fees would have ended.

Neither the shareholders nor I will ever know what motivated the dissenters. Indeed they themselves will not likely know, given that self-interest inevitably blurs introspection. We do know one thing, though: At the same meeting at which the deal was rejected, the board voted itself a significant increase in directors' fees.

- While we are on the subject of self-interest, let's turn again to the most important accounting mechanism still available to CEOs who wish to overstate earnings: the non-expensing of stock options. The accomplices in perpetuating this absurdity have been many members of Congress who have defied the arguments put forth by all Big Four auditors, all members of the Financial Accounting Standards Board and virtually all investment professionals.

I'm enclosing an op-ed piece I wrote for The Washington Post describing a truly breathtaking bill that was passed 312-111 by the House last summer. Thanks to Senator Richard Shelby, the Senate didn't ratify the House's foolishness. And, to his great credit, Bill Donaldson, the investor-minded Chairman of the SEC, has stood firm against massive political pressure, generated by the check-waving CEOs who first muscled Congress in 1993 about the issue of option accounting and then repeated the tactic last year.

Because the attempts to obfuscate the stock-option issue continue, it's worth pointing out that no one – neither the FASB, nor investors generally, nor I – are talking about restricting the use of options in any way. Indeed, my successor at Berkshire may well receive much of his pay via options, albeit logically-structured ones in respect to 1) an appropriate strike price, 2) an escalation in price that reflects the retention of earnings, and 3) a ban on his quickly disposing of any shares purchased through options. We cheer arrangements that motivate managers, whether these be cash bonuses or options. And if a company is truly receiving value for the options it issues, we see no reason why recording their cost should cut down on their use.

The simple fact is that certain CEOs know their own compensation would be far more rationally determined if options were expensed. They also suspect that their stock would sell at a lower price if realistic accounting were employed, meaning that they would reap less in the market when they unloaded their personal holdings. To these CEOs such unpleasant prospects are a fate to be fought with all the resources they have at hand – even though the funds they use in that fight normally don’t belong to them, but are instead put up by their shareholders.

Option-expensing is scheduled to become mandatory on June 15 $^{th}$ . You can therefore expect intensified efforts to stall or emasculate this rule between now and then. Let your Congressman and Senators know what you think on this issue.

The Annual Meeting

There are two changes this year concerning the annual meeting. First, we have scheduled the meeting for the last Saturday in April (the $30^{\text{th}}$ ), rather than the usual first Saturday in May. This year Mother's Day falls on May 8, and it would be unfair to ask the employees of Borsheim's and Gorat's to take care of us at that special time – so we've moved everything up a week. Next year we'll return to our regular timing, holding the meeting on May 6, 2006.

Additionally, we are changing the sequence of events on meeting day, April 30. Just as always, the doors will open at the Qwest Center at 7 a.m. and the movie will be shown at 8:30. At 9:30, however, we will go directly to the question and answer period, which (allowing for lunch at the Qwest's stands) will last until 3:00. Then, after a short recess, Charlie and I will convene the annual meeting at 3:15.

We have made this change because a number of shareholders complained last year about the time consumed by two speakers who advocated proposals of limited interest to the majority of the audience – and who were no doubt relishing their chance to talk to a captive group of about 19,500. With our new procedure, those shareholders who wish to hear it all can stick around for the formal meeting and those who don’t can leave – or better yet shop.

There will be plenty of opportunity for that pastime in the vast exhibition hall that adjoins the meeting area. Kelly Muchemore, the Flo Ziegfeld of Berkshire, put on a magnificent shopping extravaganza last year, and she says that was just a warm-up for this year. (Kelly, I am delighted to report, is getting married in October. I'm giving her away and suggested that she make a little history by holding the wedding at the annual meeting. She balked, however, when Charlie insisted that he be the ringbearer.)

Again we will showcase a 2,100 square foot Clayton home (featuring Acme brick, Shaw carpet, Johns Manville insulation, MiTek fasteners, Carefree awnings and NFM furniture). Take a tour through the home. Better yet, buy it.

GEICO will have a booth staffed by a number of its top counselors from around the country, all of them ready to supply you with auto insurance quotes. In most cases, GEICO will be able to give you a special shareholder discount (usually 8%). This special offer is permitted by 45 of the 50 jurisdictions in which we operate. Bring the details of your existing insurance and check out whether we can save you money.

On Saturday, at the Omaha airport, we will have the usual array of aircraft from NetJets® available for your inspection. Stop by the NetJets booth at the Qwest to learn about viewing these planes. Come to Omaha by bus; leave in your new plane.

The Bookworm shop did a terrific business last year selling Berkshire-related books. Displaying 18 titles, they sold 2,920 copies for \$61,000. Since we charge the shop no rent (I must be getting soft), it gives shareholders a 20% discount. This year I’ve asked The Bookworm to add Graham Allison’s Nuclear Terrorism: The Ultimate Preventable Catastrophe, a must-read for those concerned with the safety of our country. In addition, the shop will premiere Poor Charlie’s Almanack, a book compiled by Peter Kaufman. Scholars have for too long debated whether Charlie is the reincarnation of Ben Franklin. This book should settle the question.

An attachment to the proxy material that is enclosed with this report explains how you can obtain the credential you will need for admission to the meeting and other events. As for plane, hotel and car reservations, we have again signed up American Express (800-799-6634) to give you special help. They do a terrific job for us each year, and I thank them for it.

At Nebraska Furniture Mart, located on a 77-acre site on $72^{\text{nd}}$ Street between Dodge and Pacific, we will again be having “Berkshire Weekend” pricing. We initiated this special event at NFM eight years ago, and sales during the “Weekend” grew from \$5.3 million in 1997 to \$25.1 million in 2004 (up 45% from a year earlier). Every year has set a new record, and on Saturday of last year, we had the largest single-day sales in NFM's history – \$6.1 million.

To get the discount, you must make your purchases between Thursday, April 28 and Monday, May 2 inclusive, and also present your meeting credential. The period's special pricing will even apply to the products of several prestigious manufacturers that normally have ironclad rules against discounting but that, in the spirit of our shareholder weekend, have made an exception for you. We appreciate their cooperation. NFM is open from 10 a.m. to 9 p.m. Monday through Saturday, and 10 a.m. to 6 p.m. on Sunday. On Saturday this year, from 5:30 p.m. to 8 p.m. we are having a special affair for shareholders only. I'll be there, eating barbeque and drinking Coke.

Borsheim's – the largest jewelry store in the country except for Tiffany's Manhattan store – will have two shareholder-only events. The first will be a cocktail reception from 6 p.m. to 10 p.m. on Friday, April 29. The second, the main gala, will be from 9 a.m. to 4 p.m. on Sunday, May 1. On Saturday, we will be open until 6 p.m.

We will have huge crowds at Borsheim's throughout the weekend. For your convenience, therefore, shareholder prices will be available from Monday, April 25 through Saturday, May 7. During that period, just identify yourself as a shareholder through your meeting credentials or a brokerage statement.

Borsheim's operates on a gross margin that is fully twenty percentage points below that of its major rivals, even before the shareholders' discount. Last year, business over the weekend increased $73\%$ from 2003, setting a record that will be tough to beat. Show me it can be done.

In a tent outside of Borsheim's, Patrick Wolff, twice U.S. chess champion, will take on all comers in groups of six – blindfolded. Additionally, we will have Bob Hamman and Sharon Osberg, two of the world's top bridge experts, available to play with our shareholders on Sunday afternoon. They plan to keep their eyes open – but Bob never sorts his cards, even when playing for a national championship.

Gorat's – my favorite steakhouse – will again be open exclusively for Berkshire shareholders on Sunday, May 1, and will be serving from 4 p.m. until 10 p.m. Please remember that to come to Gorat's on that day, you must have a reservation. To make one, call 402-551-3733 on April 1 (but not before). If Sunday is sold out, try Gorat's on one of the other evenings you will be in town. Enhance your reputation as an epicure by ordering, as I do, a rare T-bone with a double helping of hash browns.

We will again have a special reception from 4:00 to 5:30 on Saturday afternoon for shareholders who have come from outside of North America. Every year our meeting draws many people from around the globe, and Charlie and I want to be sure we personally greet those who have come so far. Last year we enjoyed meeting more than 400 of you including at least 100 from Australia. Any shareholder who comes from other than the U.S. or Canada will be given a special credential and instructions for attending this function.

* * * * * * * * * * * *

Charlie and I are lucky. We have jobs that we love and are helped every day in a myriad of ways by talented and cheerful associates. No wonder we tap-dance to work. But nothing is more fun for us than getting together with our shareholder-partners at Berkshire's annual meeting. So join us on April $30^{\text{th}}$ at the Qwest for our annual Woodstock for Capitalists.

February 28, 2005

Warren E. Buffett

Chairman of the Board

注：下表刊载于印刷版年报中董事长信函的对面页，信函内也有引用。
伯克希尔的业绩与标普500之比较

年份		年度百分比变化		相对业绩 (1)-(2)
		伯克希尔每股账面价值 (1)	标普500（含股息）(2)
1965		23.8	10.0	13.8
1966		20.3	(11.7)	32.0
1967		11.0	30.9	(19.9)
1968		19.0	11.0	8.0
1969		16.2	(8.4)	24.6
1970		12.0	3.9	8.1
1971		16.4	14.6	1.8
1972		21.7	18.9	2.8
1973		4.7	(14.8)	19.5
1974		5.5	(26.4)	31.9
1975		21.9	37.2	(15.3)
1976		59.3	23.6	35.7
1977		31.9	(7.4)	39.3
1978		24.0	6.4	17.6
1979		35.7	18.2	17.5
1980		19.3	32.3	(13.0)
1981		31.4	(5.0)	36.4
1982		40.0	21.4	18.6
1983		32.3	22.4	9.9
1984		13.6	6.1	7.5
1985		48.2	31.6	16.6
1986		26.1	18.6	7.5
1987		19.5	5.1	14.4
1988		20.1	16.6	3.5
1989		44.4	31.7	12.7
1990		7.4	(3.1)	10.5
1991		39.6	30.5	9.1
1992		20.3	7.6	12.7
1993		14.3	10.1	4.2
1994		13.9	1.3	12.6
1995		43.1	37.6	5.5
1996		31.8	23.0	8.8
1997		34.1	33.4	.7
1998		48.3	28.6	19.7
1999		.5	21.0	(20.5)
2000		6.5	(9.1)	15.6
2001		(6.2)	(11.9)	5.7
2002		10.0	(22.1)	32.1
2003		21.0	28.7	(7.7)
2004		10.5	10.9	(.4)
年均涨幅 — 1965-2004		21.9	10.4	11.5
总涨幅 — 1964-2004		286,865	5,318

说明：数据按日历年度统计，以下除外：1965和1966年截至9月30日；1967年为截至12月31日的15个月。

自1979年起，会计规则要求保险公司将其持有的权益证券按市价估值，而非此前的成本与市价孰低法。在本表中，伯克希尔1978年之前的业绩已按新规则重述。其他方面，业绩均按最初报告的数字计算。
标普500的数字是税前，而伯克希尔的数字是税后。如果有一家像伯克希尔这样的公司，仅仅是持有标普500并计提相应税款，那么在该指数正回报的年份，其业绩会落后于标普500；而在指数负回报的年份，其业绩则会超过标普500。多年来，税收成本会导致累计落后幅度相当大。

伯克希尔·哈撒韦公司

致伯克希尔·哈撒韦公司股东：

2004年，我们的净资产增加了83亿美元，使得A类股和B类股的每股账面价值均增长了10.5%。在过去40年（即现任管理层接手以来），账面价值已从19美元增长至55,824美元，年复合增长率为21.9%。*

然而，真正重要的是每股内在价值。在这方面，好消息是：1964年至2004年间，伯克希尔从一家挣扎求生、内在价值低于账面价值的北方纺织企业，蜕变为一个内在价值远超账面价值的多元化企业。因此，我们40年内在价值的增长幅度略微超过了账面价值21.9%的增长率。（关于内在价值，以及指导查理·芒格——我的合伙人兼伯克希尔副董事长——和我在经营伯克希尔时所遵循的经济原则，请阅读第73页开始的《所有者手册》。）

尽管存在缺陷，但每年计算账面价值在伯克希尔仍然有用，它可以作为衡量我们内在价值长期增长率的一个略微保守的指标。不过，与过去相比，用它来评估任何单一年度相对于标普500指数的表现（我们在对页展示了这一比较）时，其相关性已有所下降。我们的股权投资（包括可转换优先股）占净资产的比例已大幅下降，例如从20世纪80年代平均114%降至近年来的不足50%。因此，股市的年度波动现在对我们净资产的影响比过去小得多，这一事实通常会在股市大幅上涨的年份导致我们表现落后，而在股市下跌的年份表现领先。

无论年度比较结果如何，伯克希尔相对于标普的长期表现始终至关重要。我们的股东可以通过指数基金以极低的成本买入标普500。除非我们未来能实现每股内在价值的增长超过标普，否则查理和我并没有为你们自己能做到的事情增添任何价值。

去年，伯克希尔账面价值10.5%的增长低于指数10.9%的回报。我们平淡无奇的表现并非因为旗下经营企业CEO们的失误：和往常一样，他们超额完成了自己的那份重担。我给他们的信息很简单：经营你的企业，就像它是你家族未来一百年唯一拥有的资产一样。他们几乎总是这样做，在满足企业自身需求后，将多余的现金送到奥马哈让我来配置。

我去年没有很好地完成那份工作。我希望进行几笔数十亿美元的收购，为我们已有的众多收益流增添新的重要来源。但我三振出局了。此外，我几乎找不到什么有吸引力的证券可买。因此，伯克希尔在年末持有430亿美元的现金等价物，这可不是一个令人愉快的状况。查理和我会在2005年努力将这些囤积的一部分转化为更有趣的资产，尽管我们不能保证成功。
在某种意义上，2004年对股市而言是引人注目的一年——这一事实被埋藏在第2页那一堆数字迷宫之中。如果你回顾上世纪60年代末以来的35年，你会发现，投资者持有标普500的回报（含股息）平均每年为11.2%（远高于我们对未来回报的预期）。但如果你要找回报接近11.2%的年份——比方说在8%到14%之间——你会发现2004年之前只有一年符合。换句话说，去年的“正常”回报其实一点都不正常。

在这35年里，美国企业交出了卓越的成绩单。因此，投资者按理说应该很容易就能获得丰厚的回报：他们只需以低成本、分散化的方式搭上美国企业的顺风车即可。一只他们从不碰触的指数基金就能做到这一点。然而，许多投资者的经历却介于平庸与灾难之间。

原因主要有三个：第一，高成本，通常是因为投资者交易过于频繁，或在投资管理上花费太多；第二，投资决策基于小道消息和潮流，而非基于对企业深思熟虑、量化的评估；第三，进出市场的“走走停停”方式——不合时宜地入场（在涨势持续很久之后）和离场（在停滞或下跌之后）。投资者应该记住，兴奋和费用是他们的敌人。如果他们执意要试图择时参与股市，那么他们应当在别人贪婪时恐惧，只在别人恐惧时贪婪。

各板块业绩

作为管理者，查理和我希望，如果角色互换，我们也能获得自己期望得到的信息和评论。随着伯克希尔的版图不断扩大，既要清晰又要合理简洁地做到这一点变得越来越困难。我们的一些业务与其他业务的经济特征截然不同，这意味着我们的合并报表——上面堆满了各种数字——几乎不可能进行有用的分析。

因此，在接下来的几页中，我们将展示我们四大业务类别的部分资产负债表和盈利数据，并附上各自的评论。我们特别希望你们理解，鉴于我们通常回避债务，我们将在何种有限情况下使用债务。但我们不会用那些对估算伯克希尔内在价值没有实际价值的数据淹没你们。这样做往往会混淆真正重要的事实。

受监管的公用事业业务

我们持有中美能源控股（MidAmerican Energy Holdings）80.5%（完全摊薄）的权益，该公司拥有多种公用事业业务。其中最大的是：（1）约克郡电力与北方电力（Yorkshire Electricity and Northern Electric），其370万电力客户使其成为英国第三大电力分销商；（2）中美能源（MidAmerican Energy），主要服务于爱荷华州69.8万电力客户；以及（3）克恩河与北方天然气管道（Kern River and Northern Natural pipelines），输送美国天然气消费量的7.9%。

中美能源剩余的19.5%由我们的三位合伙人持有：戴夫·索科尔（Dave Sokol）和格雷格·阿贝尔（Greg Abel）——这些业务的杰出管理者，以及沃尔特·斯科特（Walter Scott）——我的老朋友，是他把我介绍给这家公司的。由于中美能源受《公用事业控股公司法》（PUHCA）约束，伯克希尔的投票权被限制在9.9%。投票控制权掌握在沃尔特手中。
我们有限的投票权迫使我们以一种简略的方式核算中美能源。我们没有将公司的资产、负债、收入和费用全部并入伯克希尔的报表，而是在资产负债表和利润表中只做单行登记。不过，PUHCA（公用事业控股公司法）有朝一日——也许很快——会被废除，或者会计准则会发生变化。届时伯克希尔的合并数据将包含中美能源的全部，包括其使用的大量债务（尽管这些债务现在不是，将来也永远不会成为伯克希尔的负债）。

年底时，中美能源有14.78亿美元的次级债务需偿还给伯克希尔。这笔债务使得其收购得以融资，而无需我们的合伙人增加他们已经相当可观的中美能源投资。伯克希尔收取11%的利息，为收购所需的资金获得了公平的补偿，同时我们的合伙人避免了股权摊薄。由于中美能源去年没有进行大规模收购，它偿还了欠我们的1亿美元债务。

中美能源还拥有一项重要的非公用事业业务——美国家园服务（HomeServices of America），这是美国第二大房地产经纪公司。与我们的公用事业运营不同，这项业务具有高度的周期性，但我们仍然非常看好它。我们有一位出色的经理人Ron Peltier，他凭借收购和运营技能，正在打造一个经纪业务帝国。

2004年，美国家园服务参与了598亿美元的房产交易，比2003年增加了112亿美元。其中约24%的增长来自于当年的六次收购。通过我们的17家经纪公司——所有这些公司都保留其本地品牌——我们在18个州雇佣了超过18,000名经纪人。随着我们继续收购领先的本地化运营机构，美国家园服务几乎肯定会在未来十年大幅增长。

去年，中美能源对一个锌回收项目进行了重大减值，该项目于1998年启动，2002年投入运营。我们加州地热运营产生的卤水中含有大量锌，我们曾认为可以从中获利提取。好几个月里，商业上可行的回收似乎近在眼前。但在采矿中，就像在石油勘探中一样，前景总是会“逗弄”开发人员，每次解决一个问题，另一个问题又冒出来。9月份，我们认输了。

这次失败说明了一条指导方针的重要性——坚持简单的命题——这条方针我们通常也应用于投资和运营。如果一个决策只有单一关键变量，且该变量有90%的概率对你有利，那么成功的结果显然是90%。但如果有十个独立变量都需要有利才能成功，且每个变量都有90%的成功概率，那么最终获胜的可能性只有35%。在我们的锌矿项目中，我们解决了大部分问题。但有一个问题被证明是无法解决的，而这一条就太多了。既然一条链子的强度取决于最薄弱的一环，那么寻找——请原谅我使用矛盾修辞法——单一环节的链条是有道理的。

以下是中美能源业绩的分项数据。2004年，“其他”类别包括出售安然应收款所得的7,220万美元利润，这笔应收款是我们在两年前收购北方天然气公司时附带获得的。沃特、戴夫和我作为奥马哈本地人，将这笔意外之财视为战争赔款——部分补偿我们城市在1986年遭受的损失，当时Ken Lay在承诺将公司留在这里之后，却将北方天然气公司迁到了休斯顿。（详情见伯克希尔2002年年报。）

以下是中美能源运营的一些关键数据：

	收益（单位：百万美元）
	2004年	2003年
英国公用事业	326	289
爱荷华州公用事业	268	269
管道	288	261
居家服务	130	113
其他（净额）	172	190
锌项目亏损	（579）	（46）
企业利息及税前收益	605	1,076
利息（伯克希尔以外）	（212）	（225）
伯克希尔次级债务利息	（170）	（184）
所得税	（53）	（251）
净收益	170	416
伯克希尔应占收益*	237	429
欠第三方债务	10,528	10,296
欠伯克希尔债务	1,478	1,578

*包括伯克希尔赚取的利息（扣除相关所得税后），2004年为1.1亿美元，2003年为1.18亿美元。

保险

自1967年伯克希尔收购National Indemnity（国民保险公司，简称NICO）以来，财产意外险一直是我们的核心业务，也是我们增长的助推器。保险为我们提供了源源不断的资金，我们用它收购了那些如今给我们带来日益多元收益流的证券和企业。所以在这一节，我会花点时间讲讲我们是如何走到今天的。

我们保险资金的来源是“浮存金”——这些钱不属于我们，但我们暂时持有。浮存金大部分来自：（1）保费通常提前支付，而我们提供的服务——保险保障——却是在通常为期一年的时间段内交付；（2）今天发生的事故，并不总是导致我们立即支付理赔款，因为有些损失从报告（比如石棉损失）、谈判到最终结清，有时需要很多年。1967年收购时带来的2,000万美元浮存金，如今已通过内部增长和收购，增加到了461亿美元。

浮存金是美妙的——前提是它代价不高。它的成本取决于承保结果，即我们最终要支付的费用和损失与所收保费之间的比较。当承保利润实现时——伯克希尔经营保险业务的38年中约有半数年份如此——浮存金比免费还划算。那些年份，我们实际上是在为持有别人的钱而获得报酬。然而，对大多数保险公司来说，日子要艰难得多：财产意外险行业整体上几乎总是处于承保亏损状态。亏损一大，浮存金就变得昂贵，有时甚至极其致命。

保险公司通常回报率很低，原因很简单：它们销售的是类似商品的产品。保单格式是标准的，产品可从许多供应商处获得，其中一些是互助公司（由保单持有人“拥有”而非股东），其盈利目标有限。而且，大多数投保人并不在意从谁那里购买。成百上千万的顾客会说“我要买点吉列刀片”或“来瓶可口可乐”，但我们等不到有人说“请给我一份国民保险公司的保单吧”。因此，保险业的价格竞争通常十分激烈——就像航空公司的座位一样。

那么你可能会问，伯克希尔的保险业务是如何克服行业糟糕的经济状况，在一定程度上实现持久竞争优势的呢？我们通过多种方式应对这个问题。先来看看NICO的策略。
当我们收购这家公司时——它专营商业车险和一般责任险——它似乎并不具备任何能克服行业长期困境的特质。它名气不大，没有信息优势（该公司从未有过精算师），也不是低成本运营者，而且通过总代理销售——许多人认为这种方式已经过时。然而，在过去几乎38年的时间里，NICO一直是明星级表现者。确实，如果我们没有进行这次收购，伯克希尔能有今天一半的市值就算幸运了。

我们拥有的优势是一种管理心态，大多数保险公司认为无法复制。请看对页。你能想象任何一家上市公司会拥抱一种导致收入下滑——正如我们从1986年到1999年所经历的那样——的商业模式吗？需要强调的是，那次巨大的下滑并非因为业务拿不到。如果我们愿意降价，NICO本可以轻易获得数十亿美元的保费收入。但我们始终坚持定价以盈利为目的，而不是为了匹配我们最乐观的竞争对手。我们从未离开客户——是客户离开了我们。

大多数美国企业都怀有一种"机构惯性"，拒绝业务量的持续下降。哪个CEO愿意向股东报告说，去年业务不仅萎缩，而且还会继续下滑？在保险业，维持承保业务的冲动还会加剧，因为愚蠢定价保单的后果可能在一段时间内不会显现。如果保险公司在准备金上过于乐观，报告利润就会被夸大，可能多年后真实损失成本才会暴露（这种自欺欺人的做法在1970年代初期几乎毁了GEICO）。

一位严守纪律的承保人的写照
国民赔偿公司

年份		承保保费（百万美元）	年末员工数	运营费用与保费之比	承保利润（亏损）占保费百分比（按2004年底测算）*
1980	......	79.6	372	32.3%	8.2%
1981	......	59.9	353	36.1%	(0.8%)
1982	......	52.5	323	36.7%	(15.3%)
1983	......	58.2	308	35.6%	(18.7%)
1984	......	62.2	342	35.5%	(17.0%)
1985	......	160.7	380	28.0%	1.9%
1986	......	366.2	403	25.9%	30.7%
1987	......	232.3	368	29.5%	27.3%
1988	......	139.9	347	31.7%	24.8%
1989	......	98.4	320	35.9%	14.8%
1990	......	87.8	289	37.4%	7.0%
1991	......	88.3	284	35.7%	13.0%
1992	......	82.7	277	37.9%	5.2%
1993	......	86.8	279	36.1%	11.3%
1994	......	85.9	263	34.6%	4.6%
1995	......	78.0	258	36.6%	9.2%
1996	......	74.0	243	36.5%	6.8%
1997	......	65.3	240	40.4%	6.2%
1998	......	56.8	231	40.4%	9.4%
1999	......	54.5	222	41.2%	4.5%
2000	......	68.1	230	38.4%	2.9%
2001	......	161.3	254	28.8%	(11.6%)
2002	......	343.5	313	24.0%	16.8%
2003	......	594.5	337	22.2%	18.1%
2004	......	605.6	340	22.5%	5.1%

*要摸清某一年真正的盈利水平需要很长时间。首先，许多理赔申请在年底后才收到，我们必须估计会有多少以及需要多少费用。（保险术语中，这些理赔被称为IBNR——已发生未报告。）其次，理赔往往需要多年甚至数十年才能结案，这意味着过程中可能意外不断。

基于这些原因，这一列的结果只是我们在2004年底对过往年份表现的最佳估计。1999年及以前年份的利润率应该相当准确，因为这些年份已经"成熟"，基本没有未结清的理赔。年份越近，猜测的成分就越大。特别是2003年和2004年显示的结果，未来很可能发生显著变化。

最后，这里还存在一个恐惧因素：业务萎缩通常会导致裁员。为了避免收到解雇通知，员工会为定价不足找理由，告诉自己必须忍受低价业务，才能维持团队完整、让分销渠道满意。他们会辩称，如果不这么做，公司就赶不上那场他们总觉得即将到来的复苏。
为了对付员工保住自己饭碗的天性，我们一直向 NICO 的员工承诺：无论业务收缩有多严重，绝不会因为业务量下降而解雇任何人。（这里不是唐纳德·特朗普那种地方。）NICO 不是劳动密集型公司，而且如下表所示，它能承受冗余的开销。但它无法承受定价过低的业务以及随之而来的承销纪律崩溃。一个今年都不在乎承保利润的保险机构，明年也不太可能在乎。

自然，一家奉行不裁员政策的企业，在景气时期必须格外小心，避免人员过剩。三十年前，当时 Cap Cities 的 CEO Tom Murphy 用一个假设的故事向我阐明了这一点。一个员工请求老板批准他雇一名助理。该员工以为每年增加 2 万美元的薪酬无关紧要。但老板告诉他，这个提议应当被视作 300 万美元的决策——考虑到加薪、福利和其他开支（人多了，厕纸也用得多），一个额外员工在其职业生涯中很可能至少花费这么多钱。除非公司陷入极度困境，否则无论这个员工对业务的贡献多么微不足道，他都不太可能被解雇。

需要真正的坚韧不拔——深深植根于公司文化之中——才能像 NICO 那样运作。任何人查看表格都能飞快扫过 1986 年到 1999 年。但眼睁睁看着业务量一天天萎缩，同时竞争对手却在夸耀增长、收获华尔街的掌声——这种经历很少有管理者能忍受。然而，NICO 自 1940 年成立以来有过四位 CEO，没有一个人屈服。（应当指出，四位中只有一位是大学毕业。我们的经验告诉我们，非凡的商业能力很大程度上是天生的。）

NICO 现任管理者——不，是超级巨星——是 Don Wurster（没错，他就是那位“毕业生”），他自 1989 年起负责运营。他的本垒打率和巴里·邦兹一样高，因为像巴里一样，Don 宁愿选四坏球保送，也不挥棒去击打坏球。如今 Don 在 NICO 积累了 9.5 亿美元的浮存金，经过时间的考验，几乎肯定属于负成本类型。由于保险价格正在下跌，Don 的业务量很快就会大幅下降，而随着这种情况发生，查理和我将更加大声地为他鼓掌。

* * * * * * * * * * * * * * * * * * * * * *

在商品型业务中繁荣的另一种方式是成为低成本运营商。在广泛经营的汽车保险公司中，GEICO 拥有那个梦寐以求的称号。如我们所见，对 NICO 而言，潮汐式业务模式是合理的。但拥有低成本优势的公司必须执行一种不松懈的、踩足油门的策略。这正是我们在 GEICO 所做的。

一个世纪前，汽车首次出现时，财产意外险行业以卡特尔形式运作。主要公司（大部分总部设在东北部）制定了“同业公会”费率，仅此而已。没有人降价来吸引业务。相反，保险公司争夺实力雄厚、口碑好的代理人，这种重点导致代理人佣金高企，消费者价格高昂。
1922年，来自伊利诺伊州默纳的农民乔治·麦克尔勒（George Mecherle）创立了State Farm（州立农业保险），其目标是利用行业内高成本巨头搭建的价格保护伞来获取优势。State Farm采用“专属”代理人体系，这一系统使其获客成本低于公会保险公司（后者的“独立”代理人成功地在各家保险公司之间挑拨离间）。凭借低成本结构，State Farm最终拿下了个人险种（车险和房屋险）约25%的市场份额，远远甩开了昔日强大的竞争对手。Allstate（好事达保险）于1931年成立，建立了类似的销售体系，很快成为个人险领域中的亚军，仅次于State Farm。资本主义发挥了它的魔力，这些低成本企业看起来势不可挡。

但一位名叫利奥·古德温（Leo Goodwin）的人构想了效率更高的车险公司，1936年，他用区区20万美元起步，创办了GEICO（政府雇员保险公司）。古德温的计划是彻底取消代理人，直接与车主打交道。他自问：当车险这种产品既是强制性的又价格高昂时，为什么分销机制中要存在任何不必要的、昂贵的环节呢？他推断，企业保险的买家或许需要专业建议，但大多数消费者都知道自己需要什么样的车险保单。这是个极为深刻的洞见。

最初，GEICO将其低价信息邮寄给有限的政府雇员群体。后来，它拓宽了视野，将营销重点转向电话，处理来自广播和平面广告的咨询。而如今，互联网正强势崛起。

从1936年到1975年，GEICO从零起步，市场份额增长到4%，成为全美第四大车险公司。在这段时期的大部分时间里，公司管理极为出色，既实现了可观的业务量增长，又获得了高额利润。它看起来势不可挡。然而，1970年，我的朋友兼英雄洛里默·戴维森（Lorimer Davidson）退休后，他的继任者很快犯了一个大错：损失准备金计提不足。这导致了错误的成本信息，进而引发了定价不足。到1976年，GEICO濒临破产。

随后杰克·伯恩（Jack Byrne）出任GEICO首席执行官，几乎凭一己之力，通过包括大幅提价在内的英勇举措拯救了公司。尽管GEICO的生存需要这些涨价，但保单持有者纷纷逃离，到1980年，其市场份额已降至1.8%。此后，公司又进行了一些不明智的多元化扩张。这种偏离其卓越核心业务的转型阻碍了GEICO的增长，到1993年，其市场份额仅小幅增长至1.9%。随后托尼·奈斯利（Tony Nicely）接手了公司。

他的到来带来了天壤之别：2005年，GEICO很可能拿下6%的市场份额。更棒的是，托尼在增长的同时也兼顾了盈利。事实上，GEICO为其所有相关方都带来了重大利益：2004年，其客户相比他们原本可能需要支付的保费节省了约10亿美元；其员工获得了1.91亿美元的利润分享奖金，平均占工资的24.3%；而其所有者——也就是我们——则获得了丰厚的财务回报。

还有更多好消息。1976年杰克·伯恩拯救公司时，新泽西州拒绝批准他维持盈利所需的费率。他因此立即——且恰当地——退出了该州。此后，GEICO避开了新泽西和马萨诸塞州，认识到这两个司法管辖区是保险公司注定难以生存的地方。

然而在2003年，新泽西州重新审视了其长期存在的车险问题，并颁布了旨在遏制欺诈、为保险公司提供公平竞争环境的立法。即便如此，人们本可能会预期该州的官僚机构会让变革变得缓慢而艰难。
但实际情况却恰恰相反。新泽西州保险专员 Holly Bakke——她无论做什么工作都会取得成功——决心将法律意图变为现实。在她的团队配合下，GEICO 敲定了重新进入该州的细节，并于去年 8 月获批。自那以后，新泽西州驾车者的反响远超我的预期。

我们目前服务 14 万名保单持有人——约占新泽西市场的 4%——为他们节省了大笔开支（就像我们为其他州的驾车者所做的一样）。州内的口口相传让咨询电话络绎不绝。而一旦我们收到新泽西潜在客户的咨询，我们的成交率——即咨询转化为保单的比例——远远高于全国平均水平。

当然，我们并不声称能为所有人省钱。有些公司采用与我们不同的评级体系，会向某些特定类别的驾车者提供比我们更低的费率。但我们相信，在服务于所有公众群体的全国性公司中，GEICO 提供最低价格的频率最高。此外，在包括新泽西在内的多数州，伯克希尔股东还可享受 8% 的折扣。所以，花上十五分钟去 GEICO.com 看看——或者拨打 800-847-7536——看看你是否能省下一大笔钱（你当然可以用这笔钱买其他伯克希尔的产品）。

* * * * * * * * * * * *

再保险——即向希望转嫁部分承保风险的其他保险公司出售的保险——不应当是一种商品。说到底，任何保单不过是一纸承诺，而众所周知，承诺的质量千差万别。

然而在直接保险层面，作出承诺的主体通常并不那么重要。例如在个人险种中，各州会向有偿付能力的公司征收费用，以赔付破产公司的保单持有人。在商业保险领域，工伤保险也适用同样的安排。这些"受保护"的保单约占财产-意外险行业总量的 60%。经营稳健的保险公司对这种补贴糟糕或鲁莽同行的要求感到恼火，但现实就是这样。

直接保险层面的其他商业保险则涉及对投保人而言风险更大的承诺。例如，当 Reliance Insurance 和 Home Insurance 破产时，他们的承诺便一文不值了。因此，许多购买其商业保单的客户（工伤保险除外）遭受了惨痛损失。

然而，直接保单中的偿付能力风险，与再保险保单中潜藏的风险相比，简直是小巫见大巫。当一家再保险公司破产时，与之打交道的直接保险公司几乎总会蒙受巨额损失。这种风险绝非小事：GEICO 在 80 年代初因不慎选择再保险公司就损失了数千万美元。

如果在未来一二十年里发生一场真正的超级巨灾——这完全有可能——一些再保险公司将无法存活。迄今为止最大的保险损失是世贸中心灾难，保险业估计损失了 350 亿美元。1992 年的飓风"安德鲁"给保险公司造成了约 155 亿美元的损失（按今天的美元计算则更高）。这两次事件都震撼了保险和再保险行业。但如果一场特别严重的地震或飓风恰好袭击了错误的地点，那么 1000 亿美元的事件，甚至更大的灾难，仍有可能发生。2004 年，四场重大飓风袭击了佛罗里达州，造成总计约 250 亿美元的保险损失。其中两场——"查理"和"伊万"——如果在距实际登陆点不远的地方进入美国，造成的损失至少会是实际的三倍。
许多保险公司将1,000亿美元的行业损失视为"想都不敢想"，甚至根本不做相关预案。但在伯克希尔，我们对此做好了充分准备。我们的份额可能在3%到5%之间，而我们投资及其他业务的收益将轻松覆盖这部分成本。当"大灾之后的日子"来临时，伯克希尔的支票一定能兑现。

尽管飓风给我们造成了12.5亿美元的损失，但我们的再保险业务去年表现依然不错。在通用再保险，Joe Brandon重建了备受推崇的承销纪律文化——这种文化曾一度迷失方向。然而，他在任内对当前业务取得的出色成果，却被其上任前遗留问题所造成的不利影响抵消了。在国家赔款公司的再保险业务中，Ajit Jain继续成功承保着其他再保险公司不愿或无力承接的巨大风险。Ajit对伯克希尔的价值不可估量。

---

我们的保险经理人充分发挥了我在本部分提到的竞争优势，去年再次交出了优秀的承销成绩。正因如此，我们的浮存金成本甚至低于零。以下是成绩单：

	(单位：百万美元)
	承销利润	年末浮存金
保险业务	2004	2004	2003
通用再保险	$3	$23,120	$23,654
伯克希尔·哈撒韦再保险	417	15,278	13,948
GEICO	970	5,960	5,287
其他原保险*	161	1,736	1,331
合计	$1,551	$46,094	$44,220

*除国家赔款公司外，还包括由Rod Eldred、John Kizer、Tom Nerney和Don Towle管理的一系列其他卓越的保险业务。

2004年，尽管部分被保人选择了"回溯"（即解除）某些再保险合同，伯克希尔的浮存金仍增加了19亿美元。我们同意此类回溯，仅仅是因为我们认为这对我们在经济上有利（充分考虑了归还资金后其未来可能产生的收益）。

总而言之，去年我们被支付超过15亿美元，用于持有平均约452亿美元的浮存金。2005年的定价将不如以往有吸引力。然而，只要不发生特大巨灾，我们今年仍有不错的概率再次实现零成本浮存金。

金融与金融产品

去年在这一部分，我们讨论了五花八门的业务。在本次报告中，我们将跳过几个现在重要性已降低的业务：Berkadia已进入收尾阶段；Value Capital引入了其他投资者，使得我们此前预计需要将其财报并入我方的预期落空；而我负责的交易业务也在持续收缩。

• 去年，伯克希尔的两项租赁业务均出现反弹。在CORT（办公家具）方面，盈利仍不尽人意，但趋势向好。XTRA出售了其集装箱和多式联运业务，以便专注于其长期强项——拖车租赁。在新任CEO Bill Franz的领导下，管理费用得以下降，资产利用率提高，并已实现可观的利润。
• 通用再保险证券的收尾工作仍在继续。我们三年前就决定退出这项衍生品业务，但退出说来容易做来难。尽管衍生品工具号称具有极高的流动性——而且我们在清算过程中还受益于有利的市场环境——但我们到年底仍有2,890份未结清合约，低于峰值时的23,218份。如同地狱，衍生品交易进去容易，出来难。（还有别的相似之处，就请各位自行脑补了。）
  Gen Re 的衍生品合约始终要求按市价计价，我相信公司管理层恪尽职守地设定了现实的价格。然而，在交易结算有时要等到几十年后、且往往涉及多个变量的世界里，衍生品的市场价格可能非常模糊。在此期间，这些市价会影响每年支付的管理层和交易奖金。难怪账面利润常常是虚幻的。

投资者应该明白，在所有类型的金融机构中，快速增长有时会掩盖重大的潜在问题（偶尔也包括欺诈）。对一家衍生品业务盈利能力的真正考验，是在无增长模式下长期运营后所取得的成果。只有退潮时才知道谁在裸泳。

—— 40 年后，伯克希尔终于产生了一点协同效应：Clayton Homes（克莱顿家园）表现不错，这在一定程度上得益于它与伯克希尔的关联。活动房屋行业仍然身处企业美国的重症监护室，去年新房销量不到 13.5 万套，与 2003 年大致相当。这几年的销量是 1962 年以来的最低水平，也仅有 1995-1999 年期间年销量的 40% 左右。那个以不负责任的融资和天真的出资人为特征的时代，对行业来说是一个傻瓜的天堂。

由于大型贷款机构一个接一个地逃离这个领域，融资问题仍然困扰着活动房屋的制造商、零售商和购买者。在这方面，伯克希尔的支持对克莱顿来说弥足珍贵。我们随时准备为任何合理的项目提供资金，去年克莱顿的管理层找到了很多符合条件的机会。

正如我们在 2003 年的报告中所述，我们相信利用借来的资金支持盈利的、带息的应收账款。去年年初，我们借了 20 亿美元再贷给克莱顿（加价一个百分点），到 2005 年 1 月，总额达到 73.5 亿美元。其中大部分新增资金是我们于 2005 年 1 月 4 日借入的，用于为克莱顿在 2004 年 12 月 30 日从一家退出该业务的银行购买的一个成熟资产组合提供资金。

我们现在还有两笔额外的资产组合收购正在进行中，总额约为 16 亿美元，但再获得其他重大交易的可能性相当低。因此，克莱顿的应收账款（新发放贷款大致抵消还款）可能在一段时间内徘徊在 90 亿美元左右，并应带来稳定的收益。这种模式将与过去截然不同——过去克莱顿和该行业所有主要参与者一样，将其应收账款“证券化”，导致收益被提前确认。在过去两年中，证券化市场已经枯竭。如今可获得的有限资金不仅成本更高，而且条款苛刻。如果克莱顿在此期间保持独立，它将在融资困难中挣扎，收益平平。

4 月份，克莱顿完成了对 Oakwood Homes（奥克伍德家园）的收购，现在它是行业最大的活动房屋生产商和零售商。我们喜欢把更多资产交到公司 CEO Kevin Clayton（凯文·克莱顿）手中。他是伯克希尔管理者的典范。如今，克莱顿拥有 11,837 名员工，高于我们收购时的 7,136 名，查理和我很高兴伯克希尔为促进这一增长发挥了作用。

为简单起见，我们将克莱顿的所有收益都计入该板块，尽管其中有相当一部分来自消费金融以外的领域。

	（单位：百万美元）
	税前利润		生息负债
	2004	2003	2004	2003
交易业务——普通收益	$264	$355	$5,751	$7,826
通用再保险证券	(44)	(99)	5,437*	8,041*
寿险与年金业务	(57)	85	2,467	2,331
价值资本	30	31	不适用	不适用
伯卡迪亚	1	101	—	525
租赁业务	92	34	391	482
活动房融资（克雷顿）	220	37**	3,636	2,032
其他	78	75	不适用	不适用
资本利得前收益	584	619
交易业务——资本利得	1,750	1,215
合计	$2,334	$1,834

* 包括所有负债
** 自收购日起计算，2003年8月7日

制造、服务及零售业务

我们在这一类别的活动覆盖了方方面面。但让我们先看看整个集团的汇总资产负债表和利润表。

资产负债表 2004年12月31日（单位：百万美元）

资产		负债与权益
现金及现金等价物	$899	应付票据	$1,143
应收款项及应收票据	3,074	其他流动负债	4,685
存货	3,842	流动负债合计	5,828
其他流动资产	254
流动资产合计	8,069
商誉及其他无形资产	8,362	递延所得税	248
固定资产	6,161	长期债务及其他负债	1,965
其他资产	1,044	权益	15,595
	$23,636		$23,636

利润表（单位：百万美元）

	2004	2003
收入	$44,142	$32,106
营业费用（含折旧：2004年$676，2003年$605）	41,604	29,885
利息费用（净额）	57	64
税前利润	2,481	2,157
所得税	941	813
净利润	$1,540	$1,344

这个五花八门的业务群，销售的产品从Dilly Bar甜筒到波音737的分时权益，去年平均有形净资产收益率高达21.7%，而2003年为20.7%。值得注意的是，这些业务在实现这一回报时仅使用了极低的财务杠杆。显然，我们拥有一些非常优秀的企业。不过，我们购买其中许多企业时支付了远高于净资产的溢价——这一点体现在资产负债表上的商誉项目中——这个事实将我们平均账面价值的收益率拉低至9.9%。

以下是较大类别或业务的税前利润：

	税前利润（单位：百万美元）
	2004	2003
建筑产品	$643	$559
Shaw Industries	466	436
服装与鞋类	325	289
珠宝、家居饰品与糖果零售	215	224
航空服务	191	72
McLane	228	150*
其他业务	413	427
	$2,481	$2,157

*自2003年5月23日收购之日起。

• 在建筑产品板块和Shaw，我们经历了原材料和能源成本的惊人上涨。例如到12月，MiTek（主要业务是屋顶桁架连接件）的钢材成本比一年前涨了100%。而MiTek每年要用掉6.65亿磅钢材。尽管如此，这家公司的表现依然出色。

自2001年我们收购MiTek以来，其CEO Gene Toombs完成了几笔漂亮的"补强式"收购，正在打造一个迷你版的伯克希尔。

Shaw当年在其主要纤维材料上遭遇了一连串涨价，这一打击使其成本增加了超过3亿美元。（你踩在地毯上，其实就是在踩加工过的石油。）尽管我们随后也提高了售价，但不可避免地存在滞后。因此，随着时间推移利润率持续收窄，至今仍面临压力。尽管有这些障碍，在Bob Shaw和Julian Saul的领导下，Shaw在2004年仍然获得了惊人的25.6%有形净资产收益率。这是一家实力雄厚的公司，前景光明。

• 在服装领域，Fruit of the Loom的销量增加了1000万打，即14%，其中女性和少女内衣的出货量增长了31%。查理对这个话题比我懂行得多，他信誓旦旦地告诉我，女人并没有多穿内衣。有了他的专业意见，我只能得出结论：我们在女性内衣市场的份额一定在快速增长。多亏了John Holland，Fruit正在前进。

较小规模的Garan也表现出色。在Seymour Lichtenstein和Jerry Kamiel的带领下，该公司生产了广受欢迎的儿童产品线Garanimals。下次你去沃尔玛，记得看看这个富有想象力的产品。

• 在我们的零售业务中，Ben Bridge（珠宝）和R. C. Willey（家居饰品）去年尤其突出。

Ben Bridge的同店销售额增长了11.4%，是我所看到的已上市珠宝商报告中最好的增幅。此外，公司的利润率也在扩大。去年并非侥幸：过去十年，该公司的同店销售平均增长率为8.8%。

Ed和Jon Bridge是第四代管理者，他们经营这家公司完全像经营自己的产业一样——除了股票证书上印的是伯克希尔的名字，其他方面确实如此。Bridge家族通过选对地段，更重要的是用热情且专业的员工充实这些门店，成功实现了扩张。我们今年将进入明尼阿波利斯-圣保罗市场。

在犹他州的R. C. Willey，扩张的成效更为显著：2004年41.9%的销售额来自1999年之前不存在的州外门店。受益于拉斯维加斯的两家新店，该公司在2004年的利润率也有所提高。
我本想告诉你们，这些门店是我的主意。但说实话，我当时觉得它们是个错误。我当然知道 Bill Child 把犹他州的 R. C. Willey（R.C. Willey 家居）经营得有多出色——那里的市场份额长期巨大。但我认为，我们的周日闭店政策到了外地会是一场灾难。即便是我们在博伊西开的第一家州外门店大获成功，我也没被说服。我一直追问：拉斯维加斯的居民，习惯了七天营业的零售商，会适应我们吗？2001年我们开的第一家拉斯维加斯店，用响亮的成绩回答了这个问题——它立刻成了我们销售额最高的门店。

Bill 和他的继任 CEO Scott Hymas 接着提议再开一家拉斯维加斯店，距离第一家仅约20分钟车程。我当时觉得这次扩张会蚕食第一家店的生意，增加大量成本却只带来微薄销售额。结果呢？现在每家店的单店销售额都比连锁中其他任何门店高出约26%，而且持续录得大幅同比增长。

R. C. Willey 很快将在里诺开店。在做出这个决定之前，Bill 和 Scott 又征求了我的意见。一开始，他们来请教我，我还挺得意。但后来我恍然大悟：一个总是出错之人的意见，对决策者来说反而有独特的价值。

• 航空服务业务利润有所改善。在飞行员培训领域全球领先的 FlightSafety（飞行安全国际），随着企业航空业复苏以及我们与支线航空公司的业务增长，利润有所上升。我们现在运营着283台模拟器，原始成本为12亿美元。飞行员在这套昂贵设备上是一对一培训的。这意味着，要产生1美元的年收入，需要投入高达3.50美元的资本。在这种资本密集度下，FlightSafety 需要极高的运营利润率才能获得合理的资本回报，因此设备利用率至关重要。去年，FlightSafety 的有形净资产收益率从2003年的8.4%提升到了15.1%。

  2004年还有一件事：1951年以1万美元创立 FlightSafety 的 Al Ueltschi，把 CEO 职位交给了在公司服务43年的 Bruce Whitman。（但 Al 哪儿也不会去；我不会让他走的。）Bruce 认同 Al 的信念：驾驶飞机是一项特权，只应授予那些定期接受最高质量培训且毋庸置疑具备能力的人。几年前，有人请 Charlie 替他一位富商朋友向 Al 说情——这位朋友在 FlightSafety 的考试中没通过。Al 对 Charlie 的回答是：“告诉你朋友，他该坐飞机后排，而不是驾驶舱。”

  FlightSafety 最大的客户是 NetJets（耐特杰），我们的飞机分时所有权子公司。其2100名飞行员平均每年有18天在培训中度过。此外，这些飞行员只飞一种机型，而很多飞行运营部门会让飞行员在多种机型之间切换。NetJets 在这两方面的高标准，正是我在伯克希尔收购它之前多年就选择加入该公司的原因之一。

  然而，在我决定使用并收购 NetJets 的过程中，同样重要的是：这家公司由 Rich Santulli 管理——他是飞机分时所有权行业的开创者，对安全和服务近乎狂热。我把选择飞行服务提供商比作挑选脑外科医生：你就是想要最好的。（让别人去拿低价竞标者做实验吧。）

  去年，NetJets 再次赢得了主导行业的四家公司中约70%的新增业务净额（按美元价值计算）。我们的一部分增长来自 Marquis Jet Partners 提供的25小时卡。Marquis 并非 NetJets 所有，而是一个客户——它把从我们这里购买的服务重新包装成更小的套餐，通过它的卡片销售。Marquis 只与 NetJets 合作，在营销中借助我们声誉的力量。
  我们的美国合同数（包括Marquis客户）从2004年的3,877份增至4,967份（而伯克希尔1998年收购NetJets时约为1,200份）。有些客户（包括我本人）签了多份合同，因为他们想使用多款机型，根据每次出行任务挑选最合适的飞机。

NetJets去年在美国赚了一点钱。但是我们在国内赚的钱很大程度上被欧洲的亏损抵消了。不过，我们目前在海外正积聚真正的动能。合同数（包括我们在欧洲自行销售的25小时卡）年内从364份增至693份。2005年我们在欧洲仍会有巨额亏损，但国内业务的利润很可能让我们整体实现盈利。

欧洲对NetJets来说代价高昂——比我预想的要昂贵得多——但对于打造一个永远出类拔萃的飞行运营公司来说，这是必不可少的。我们的美国业主无论去哪里旅行都希望享受优质服务，未来几十年他们对海外飞行时数的需求必将大幅增长。去年，美国业主在欧洲完成了2,003次飞行，比前一年增长22%，比2000年增长137%。同样重要的是，我们的欧洲业主在美国完成了1,067次飞行，比2003年增长65%，比2000年增长239%。

投资

下面列出我们的普通股投资。截至2004年底市值超过6亿美元的股票单独列出。

持股数	公司	持股比例	2004年12月31日
			成本*
151,610,700	美国运通公司	12.1%	$1,470
200,000,000	可口可乐公司	8.3%	1,299
96,000,000	吉列公司	9.7%	600
14,350,600	H&R布洛克公司	8.7%	223
6,708,760	M&T银行公司	5.8%	103
24,000,000	穆迪公司	16.2%	499
2,338,961,000	中石油"H"股（或等价物）	1.3%	488
1,727,765	华盛顿邮报公司	18.1%	11
56,448,380	富国银行	3.3%	463
1,724,200	白山保险集团	16.0%	369
	其他		3,531
	普通股合计		$9,056

*这是我们实际的买入价格，也是我们的计税基础；美国通用会计准则下的"成本"在少数情况下因需要计提减值或增值而有所不同。

有些人看到这张表，可能会把它当成一份股票清单，根据图形走势、券商观点或短期盈利预测来买卖。查理和我忽略这些干扰，而是把我们持有的股票视为企业的一部分所有权。这是一个重要的区别。事实上，这种思维方式从我19岁起就成为我投资行为的基石。那时我读了本·格雷厄姆的《聪明的投资者》，顿时豁然开朗。（在此之前，我对股市心醉神迷，但对如何投资毫无头绪。）
让我们来看看“四大金刚”——美国运通、可口可乐、吉列和富国银行——自我们买入以来它们的生意表现如何。下表显示，从1988年5月到2003年10月，我们通过多笔交易在这四家公司上投资了38.3亿美元。按综合加权计算，我们的购入时点是1992年7月。因此，截至2004年底，我们持有这些“商业权益”的加权时间约为12.5年。

2004年，伯克希尔在这四家公司中所占的净利润份额达到12亿美元。这些利润可以合理地视为“正常”水平。没错，吉列和富国银行在披露利润时未计入期权成本，这夸大了它们的利润；但另一方面，可口可乐因一次性冲销而减少了利润。

我们在这四家公司中的持股利润几乎每年都在增长，目前已达到我们成本的约31.3%。它们分配给我们（伯克希尔）的现金也持续增长，2004年总计4.34亿美元，约占成本的11.3%。总而言之，四大金刚为我们带来了令人满意、尽管远非惊艳的商业回报。

我们在市场上的经历也是如此。自最初买入以来，估值增值略超利润增长，原因是市盈率有所提高。不过，按年度看，商业表现和市场表现常常背离，有时甚至极度背离。在大泡沫期间，市值增长远远超过了商业表现。而在泡沫破灭后，情况正好相反。

显然，如果我抓住这只钟摆的摆动，伯克希尔的成绩会好得多。当你透过永远擦得锃亮的后视镜看时，这似乎很容易做到。但不幸的是，投资者必须透过挡风玻璃向前看，而那块玻璃总是雾气蒙蒙。我们庞大的持股规模，更增加了我们在估值摆动时灵活地进进出出的难度。

不过，在泡沫期间我只是对高得离谱的估值嘀咕几句，而没有把想法付诸行动，这确实该挨批评。尽管我当时说过我们持有的某些股票定价超前，但我低估了估值过高有多么严重。我本该动手却只动了嘴。

查理和我现在想要的是有点动作。我们并不乐见于坐拥430亿美元的现金等价物却只赚着微薄的收益。我们更渴望买入更多类似于我们现有持股的少量权益——或者更棒的是，直接买下更多大公司。不过，我们只会在买入价格能提供合理投资回报前景时，才会采取行动。

* * * * * * * * * * * *

我们一再强调，我们按季度或年度报告的“已实现”收益，对分析而言毫无意义。我们的账面上有巨额未实现收益，而我们何时（以及是否）兑现这些收益，完全不以我们希望在某特定时间报告利润为出发点。我们报告收益的另一个复杂因素在于，美国通用会计准则要求外汇合同必须按市值计价，这一规定导致这些持有的未实现损益会透过我们公布的利润表反映出来，就好像我们已经卖掉头寸一样。

尽管有上述种种问题，你们可能仍想了解一下我们在2003年和2004年报告收益的细分情况。这些数据反映的是实际卖出情况，但外汇收益例外——它由卖出和逐市计价的综合结果组成。

类别	税前收益（百万美元）
	2004年	2003年
普通股	$ 870	$ 448
美国政府债券	104	1,485
垃圾债券	730	1,138
外汇合约	1,839	825
其他	(47)	233
合计	$3,496	$4,129

垃圾债券的利润中包含外汇成分。我们在2001年和2002年买入这些债券时，首先关注的自然都是发行方的信用质量——它们全都是美国公司。不过，其中一些公司发行的债券是以外币计价的。基于我们对美元的看法，只要有机会，我们就优先买入了这些外币债券。

举个例子，2001年我们以面值51.7%的价格买入了2.54亿欧元的Level 3债券（2008年到期，票息10 $\frac{3}{4}$%），并于2004年12月以面值85%的价格卖出。这笔债券以欧元交易，买入时欧元兑美元汇率为0.88美元，卖出时却变成了1.29美元。因此，在我们1.63亿美元的总收益中，约8500万美元来自市场对Level 3信用评价的改观，其余7800万美元则来自欧元升值。（此外，持有期间我们还收到了现金利息，按我们的美元成本计算，年息约为25%。）

* * * * * * * * * * * *

媒体仍在报道“巴菲特买入”某只股票。这类报道几乎总是基于伯克希尔提交给SEC的文件，因此都是错的。我以前说过，应该写成“伯克希尔买入”。

一位纪律严明的投资者画像

Lou Simpson

年份		GEICO股票投资回报	标普回报	相对收益
1980	....	23.7%	32.3%	(8.6%)
1981	....	5.4%	(5.0%)	10.4%
1982	....	45.8%	21.4%	24.4%
1983	....	36.0%	22.4%	13.6%
1984	....	21.8%	6.1%	15.7%
1985	....	45.8%	31.6%	14.2%
1986	....	38.7%	18.6%	20.1%
1987	....	(10.0%)	5.1%	(15.1%)
1988	....	30.0%	16.6%	13.4%
1989	....	36.1%	31.7%	4.4%
1990	....	(9.9%)	(3.1%)	(6.8%)
1991	....	56.5%	30.5%	26.0%
1992	....	10.8%	7.6%	3.2%
1993	....	4.6%	10.1%	(5.5%)
1994	....	13.4%	1.3%	12.1%
1995	....	39.8%	37.6%	2.2%
1996	....	29.2%	23.0%	6.2%
1997	....	24.6%	33.4%	(8.8%)
1998	....	18.6%	28.6%	(10.0%)
1999	....	7.2%	21.0%	(13.8%)
2000	....	20.9%	(9.1%)	30.0%
2001	....	5.2%	(11.9%)	17.1%
2002	....	(8.1%)	(22.1%)	14.0%
2003	....	38.3%	28.7%	9.6%
2004	....	16.9%	10.9%	6.0%
1980-2004年平均年回报		20.3%	13.5%	6.8%

即便如此，通常情况下拍板买入的人也不是我。卢·辛普森管理者GEICO持有的约25亿美元股票，伯克希尔上报的通常都是他的交易。他买入的规模通常在2-3亿美元之间，且偏向于比我关注的公司更小的标的。看看对面那页，你就明白为什么卢肯定能入选投资名人堂了。

你可能会惊讶地发现，卢并不一定会把他在做什么告诉我。当查理和我分配职责时，我们是真正地交出接力棒——就像对待旗下运营经理一样，我们也把接力棒交给了卢。因此，我通常要等到每月结束后的十天左右，才知道卢的交易情况。顺便说一句，有时我会暗自不赞同他的决定。但通常他是对的。

外汇

年末，伯克希尔持有约214亿美元的外汇合约，分散在12种货币中。正如我去年提到的，这种持仓对我们来说是一个明显的转变。在2002年3月之前，无论是伯克希尔还是我本人，都从未涉足过外汇交易。但越来越多的证据表明，我们的贸易政策将在未来多年给美元带来持续压力——因此自2002年以来，我们在制定投资路线时一直留意这个警告。（正如W.C.菲尔兹曾经在被人讨钱时说的：“抱歉，孩子，我的钱都套在货币里了。”）

请明确一点：我们对汇率的思考绝不源于对美国本身的怀疑。我们生活在一个极其富饶的国家，这是一个珍视市场经济、法治和机会均等的体制的产物。我们的经济毫无疑问是世界上最强大的，并且将继续如此。能生活在这里是我们的幸运。

但正如我在2003年11月10日发表于《财富》杂志（可在berkshirehathaway.com上查阅）的文章中所论述的，我们国家的贸易做法正在拖累美元。美元的价值已经大幅下跌，但很可能还会继续。如果不改变政策，外汇市场甚至可能陷入混乱，并产生政治和金融两方面的溢出效应。没人知道这些问题是否会成真。但这种可能性绝不算遥远，政策制定者现在就应该考虑。然而，他们倾向于采取一种不那么良性的忽视态度：2000年11月，国会曾发布了一份长达318页、关于持续贸易赤字后果的研究报告，但此后便无人问津。这份报告是在1999年贸易赤字达到当时令人震惊的2630亿美元后要求撰写的；到去年，赤字已升至6180亿美元。

需要强调的是，查理和我认为真正的贸易——即与其他国家交换商品和服务——对双方都极其有益。去年我们进行了1.15万亿美元这样货真价实的贸易，这种贸易越多越好。但如前所述，我们国家还额外从世界其他地区购买了6180亿美元的商品和服务，而没有给予对等的回报。这是一个惊人的数字，并且带来了重要后果。

这种单向的伪贸易的平衡项——经济学中总有某个对等项——是财富从美国向世界其他地区的转移。这种转移可能以我们私人或政府机构向外国人提供借据的形式实现，也可能通过外国人取得我们资产（如股票和房地产）所有权的形式实现。无论哪种情况，结果都是美国人对本国财富的占有一部分减少，而非美国人占有的部分增加。这种向世界其他地区强行输送美国财富的行为，目前正以每天18亿美元的速度进行，比我去年写信给您时增长了20%。因此，其他国家及其公民现在净持有约3万亿美元的美国资产。十年前，他们的净持有量几乎可以忽略不计。
提到“万亿”这个数字，大多数人的大脑就会发懵。更让人混淆的是，经常账户赤字（包含三个项目，其中最重要的是贸易赤字）和我们的国家预算赤字常常被混为一谈，称作“双胞胎”。但它们根本不是一回事。两者起因不同，后果也不同。

预算赤字绝不会减少美国人分得的国民蛋糕份额。只要其他国家及其公民对美国没有净所有权，在任何预算情景下——哪怕赤字巨大——我们国家产出的100%都归美国公民所有。

作为一个物品充裕的富裕“家庭”，美国人会通过他们的立法者来争论政府应如何重新分配国民产出——也就是谁纳税、谁享受政府福利。如果早先“福利承诺”需要重新审视，“家庭成员”之间就会愤怒地争论谁来承担痛苦。也许税率会提高；也许承诺会被修改；也许国内债务会进一步发行。但无论争论如何收场，这个家庭的巨大蛋糕仍然全部由家庭成员享用，无论怎样分割。没有一块需要送出国门。

大规模且持续的经常账户赤字则会产生完全不同的结果。随着时间的推移，随着外国人对我们的债权不断增加，我们对自己产出的所有权越来越少。实际上，世界其他国家对美国产出享有的“特许权使用费”日益增长。在这里，我们就像一个持续超支的家庭。日子久了，这个家庭会发现，自己越来越多地是在为“金融公司”打工，而为自己干的越来越少。

如果我们继续维持目前水平的经常账户赤字，十年后其他国家及其公民对美国的净所有权将达到约11万亿美元。而且，如果外国投资者仅凭这些净持有资产赚取5%的收益，我们每年就需要向国外净输送5500亿美元的商品和服务，仅仅用于支付当时外国人持有的美国投资。到那时（十年后），我们的GDP大概总计约18万亿美元（假设低通胀，这远非确定之事）。因此，我们的美国“家庭”将每年拿出产出的大约3%上交给世界其他地方，作为过去过度挥霍的贡赋。在这种情况下——与预算赤字不同——子孙真要偿还父辈的罪孽了。

这种每年向世界支付的“特许权”——除非美国大幅减少消费并开始持续保持大规模贸易顺差，否则不会消失——无疑会在美国引发严重的政治动荡。美国人仍然会生活得很好，甚至比现在更好，因为经济在增长。但他们会对自己永远要向国外的债权人和所有者进贡这个念头感到愤怒。如今一个渴望成为“所有者社会”的国家，在——请允许我用夸张的说法来强调——“佃农社会”里是找不到幸福的。而这恰恰是我们的贸易政策——两党一致支持的政策——正把我们带向的方向。

许多美国著名金融界人士，无论身在政府内外，都曾表示我们的经常账户赤字不可持续。例如，2004年6月29-30日美联储公开市场委员会的会议纪要写道：“工作人员指出，过大的外部赤字不可能无限期持续下去。”然而，尽管精英们不断唉声叹气，他们却没有提出任何实质性的建议来遏制这一日益严重的不平衡。
16个月前我为《财富》杂志撰文时曾警告过："美元温和贬值并非解决之道。"至今仍未应验。然而政策制定者依然寄希望于"软着陆"，一边劝他国刺激（说白了就是"注水"）经济，一边让美国人多储蓄。依我看，这些告诫全没点中要害——美国存在根深蒂固的结构性问题，除非贸易政策彻底改弦更张，或者美元贬值到足以让金融市场不安的程度，否则我们的经常项目赤字仍会居高不下。

贸易现状的拥护者喜欢引用亚当·斯密的话："对一个家庭而言的审慎之举，对一个大国而言几乎不可能是荒谬之行。若外国能比我们自己制造更便宜的货物，不如用我们自身擅长的产业产出的一部分去向他们购买。"

我赞同。但请注意，斯密先生说的是货物换货物，而不是像我国这样以财富换货物——我国每年这么做高达6000亿美元之巨。而且我敢肯定，他绝不会认为"审慎"意味着他的"家庭"每天卖掉一部分农场来贴补过度消费。而这正是这个名为美利坚的"大国"正在做的事。

如果美国拥有6000亿美元的经常项目顺差，全世界的评论员都会猛烈抨击我们的政策，视其为极端形式的"重商主义"——一种早已被唾弃的经济策略，即国家鼓励出口、限制进口、囤积财富。我也会谴责这样的政策。但无论意图如何，实际效果上，世界其他地区对美实行的正是重商主义——这之所以可能，是因为我国拥有巨量资产和完美的信用记录。事实上，全世界绝不会允许任何其他国家像我们这样无度地使用以本国货币计价的信用卡。目前多数外国投资者还很乐观：他们也许把我们当成挥霍成瘾的败家子，但他们知道我们也是有钱的败家子。

然而，这种挥霍无度的行为不可能永远被容忍。尽管无法精确预测贸易问题会在何时以何种方式解决，但解决结果不太可能有利于美元相对于贸易伙伴货币的升值。

我们希望美国采取能够迅速、大幅减少经常项目赤字的政策。诚然，立竿见影的解决方案可能会让伯克希尔的外汇合约出现亏损。但伯克希尔的资源仍高度集中于美元资产，强势美元和低通胀环境非常符合我们的利益。

如果你想跟踪贸易和货币问题，请读《金融时报》。这家伦敦报纸长期以来一直是国际金融新闻的每日首要来源，现在还有了出色的美国版。它的贸易报道和评论都是一流的。

---

再次附上我们的常规告诫：宏观经济学是门难啃的学问，包括我和查理在内，很少有人证明过自己在这方面有什么本事。我们对汇率的判断很可能出错。（事实上，那么多权威人士如今都预测美元走弱，这反而让我们不安。）如果那样，我们的错误将暴露在光天化日之下。讽刺的是，假如我们选择相反方向——在美元大幅贬值时仍将所有伯克希尔资产留在美元上——没有人会注意到我们的错误。
约翰·梅纳德·凯恩斯在其杰作《就业、利息和货币通论》中写道：“世俗的智慧教导我们：墨守成规地失败，比打破常规地成功，更有益于名声。”（或者，用不那么优雅的话说：旅鼠这一类生物可能会被嘲笑，但从来没有哪一只单独的旅鼠受到过批评。）从声誉角度看，查理和我对外汇承诺承担了明显的风险。但我们相信，管理伯克希尔就像我们自己拥有它100%的股权那样行事。而且，如果真是那样，我们不会奉行只持有美元的政策。

杂项

• 去年我告诉过你们，一群田纳西大学金融系的学生在我们17亿美元收购Clayton Homes的过程中发挥了关键作用。早些时候，他们被教授Al Auxier带到了奥马哈——他每年都带一个班来——参观内布拉斯加家居城和博希姆珠宝店，在戈拉特's吃牛排，并在基维特广场与我进行问答交流。这些访客，就像那些来参加我们年会的嘉宾一样，对这座城市及其友好的居民印象深刻。

现在其他大学也纷纷前来拜访。本学年，我们将迎来来自芝加哥、达特茅斯（塔克商学院）、特拉华州立、佛罗里达州立、印第安纳、爱荷华、爱荷华州立、马里兰、内布拉斯加、西北拿撒勒、宾夕法尼亚（沃顿）、斯坦福、田纳西、德克萨斯、德克萨斯农工、多伦多（罗特曼）、联合学院和犹他等大学的访问班级，人数从30到100不等。大多数学生是MBA候选人，他们的素质给我留下了深刻印象。他们对商业和投资充满热情，但他们的提问表明，他们考虑的不仅仅是赚钱。每次与他们见面后，我都感觉很好。

在我们的交流课上，我向新来的客人讲述了田纳西大学团队发现Clayton Homes的故事。我这样做，就像那个拿着鸵鸟蛋走进鸡舍并告诫鸡群的农夫：“姑娘们，我不想抱怨，但这只是竞争对手在做的一点点小小示范。”到目前为止，我们的新侦察兵还没有给我们带来交易。但他们的人生使命已经明确。

• 你们应该了解一项会计准则，它以一种“今天痛苦，明天收获”的方式轻微扭曲了我们的财务报表。伯克希尔会从个人和公司那里购买人寿保险单，而这些保单的持有人原本会选择退保换取现金。作为保单的新持有人，我们支付到期应付的保费，最终——当原持有人去世时——我们收取保单的面值。

当我们购买保单时，原投保人通常健康状况良好。尽管如此，我们支付的价格总是远高于其现金退保价值（“CSV”）。有时原投保人已经以CSV为抵押借款来支付保费。在这种情况下，剩余CSV会很少，我们的购买价格将是原投保人如果退保所能获得金额的很多倍。

根据会计准则，我们必须在购买保单时，将支付金额超过CSV的部分立即作为已实现资本损失入账。此外，我们每年还必须支付保费超过CSV增加额的部分，作为额外费用。但显然，我们不认为这些账面费用代表经济损失。如果真是亏损，我们不会购买这些保单。

2004年，我们因购买保单（以及维持保单所需的保费支付）而录得净“亏损”总计2.07亿美元，这笔金额在损益表中冲抵了已实现投资收益（计入第17页表格中的“其他”项）。未来当这些保单的收益兑现时，我们将把超出当时CSV的部分记为已实现投资收益。
两项泡沫后的治理改革对伯克希尔尤其有益，我自责没有在多年前就实施它们。第一项是定期召开无CEO在场的董事会会议。我曾担任19家公司的董事，在许多场合，这一流程本会让可疑的计划得到更彻底的审视。在少数情况下，原本需要的CEO更换也会更及时地完成。这一流程没有坏处，而且可能带来许多好处。

第二项改革涉及“举报热线”，通过这一安排，员工可以在不怕报复的情况下向我及董事会审计委员会发送信息。伯克希尔极度分权的结构使这一系统对我和委员会都特别有价值。（在一个拥有18万人的庞大“城市”——伯克希尔当前的员工数——并非每只掉下来的麻雀都会被总部注意到。）我们收到的大多数投诉都是“我旁边的人有口臭”这类鸡毛蒜皮的事，但偶尔我会了解到子公司的重要问题，否则它们就会被忽视。所提出的问题通常不是审计能发现的那种，而是与人事和商业惯例相关。如果我在几十年前就建立了举报热线，今天的伯克希尔会更有价值。

查理和我喜欢股东像所有者那样思考和行动的理念。有时这要求他们积极主动。在这个领域，大型机构股东应该带头。

然而到目前为止，机构采取的行动并不令人敬畏。通常，他们关注细枝末节，却忽略了真正重要的三个问题。第一，公司是否有合适的CEO？第二，他/她的薪酬是否过高？第三，提议的收购更有可能创造还是摧毁每股价值？

在这些问题上，CEO的利益很可能与股东的利益不同。此外，董事有时缺乏知识或勇气来否决CEO。因此，大型股东聚焦这三个问题并在必要时发声至关重要。

相反，许多机构只是对当前热门议题采取“清单”式方法。去年，我就成了这种判断方式的受体。几家机构股东及其顾问认为，我在担任可口可乐董事时缺乏“独立性”。一个团体想让我离开董事会，另一个则只想让我从审计委员会被赶出去。

我的第一反应是秘密资助第二个想法背后的那个团体。为什么有人会想加入审计委员会，这让我无法理解。但由于董事必须被分配到某个委员会，而且没有CEO想让我加入其薪酬委员会，所以审计委员会的差事常常落到我头上。结果，那些反对我的机构失败了，我再次当选审计委员。（我抑制住了要求重新计票的冲动。）

一些机构质疑我的“独立性”，原因之一是McLane（麦克莱恩）和Dairy Queen（冰雪皇后）购买了大量可口可乐产品。（他们是想让我们支持百事可乐吗？）但韦氏词典对“独立”的定义是“不受他人控制”。我困惑的是，当伯克希尔持有的80亿美元可口可乐股票的价值作为平衡物时，怎么会有人得出我们购买可口可乐产品会“控制”我的决策的结论？假设我还有一点点理性，基本算术就应该清楚表明，我的心和思想属于可口可乐的股东，而不是其管理层。

我忍不住要提一下，耶稣对独立校准的理解远比那些抗议的机构要清楚。在《马太福音》6:21中，他说：“因为你的财宝在哪里，你的心也在那里。”即使对机构投资者而言，80亿美元也应该算得上是“财宝”，它使伯克希尔从与可口可乐的日常交易中可能赚取的任何利润都相形见绌。
按照圣经的标准衡量，伯克希尔的董事会堪称典范：（a）每位董事都来自持有至少400万美元股票的家族；（b）这些股份均非通过期权或赠与从伯克希尔获得；（c）没有董事从公司领取的委员会、咨询或董事会酬金超过其年收入的极小部分；（d）尽管我们设有标准的公司赔偿安排，但我们不为董事购买任何责任保险。

在伯克希尔，董事会成员与股东走在同一条路上。

---

查理和我见过大量行为印证了圣经中关于"财宝"的观点。根据我们丰富的董事会经验，最不独立的董事往往是那些从董事会服务酬金中获得相当大比例年收入的人（并且他们还希望被推荐到其他董事会，以便进一步提高收入）。然而，正是这些董事最常被归类为"独立"。

这类董事大多数是正派人士，工作也做得一流。但如果他们不受到诱惑去阻挠可能威胁其生计的行动，那他们就不是正常人了。有些人可能会屈服于这些诱惑。

让我们看一个基于间接证据的例子。我亲身了解最近的一个收购提案（并非来自伯克希尔），该提案得到管理层青睐、公司投资银行支持，并计划以高于该股票数年（或当前）售价的价格推进。此外，多位董事支持该交易并希望提交给股东表决。

然而，他们中的几位同事——每人每年从董事会和委员会得到总计约10万美元的报酬——扼杀了这一提案，这意味着股东们从未得知这项数十亿美元的报价。非管理董事几乎不持有股票，除了公司赠与的股份。他们在近年来的公开市场购买也微不足道，尽管该股票的交易价格远低于提议的收购价。换句话说，这些董事不希望股东收到X价格的要约，尽管他们自己始终拒绝以X的一个零头价格买入股票。

我不知道具体是哪几位董事反对让股东看到这个报价。但我确实知道，10万美元对于一些被视为"独立"的董事而言是其年收入的重要部分，完全符合马太福音6章21节中"财宝"的定义。如果交易完成，这些酬金就会终止。

股东和我都永远不会知道反对者的动机。事实上，他们自己可能也未必清楚，因为私利难免模糊了自省。但有一件事我们确实知道：在同一场否决该交易的董事会上，董事会给自己大幅提高了董事酬金。

——既然谈到私利问题，我们再来看看CEO们如果想虚增利润，仍可使用的最后一项重要会计手段：不将股票期权列为费用。长期维持这种荒谬行径的同谋者包括许多国会议员，他们无视四大审计公司、财务会计准则委员会全体成员以及几乎全体投资专业人士提出的论据。

随信附上我为《华盛顿邮报》撰写的一篇专栏文章，描述去年夏天众议院以312比111通过的一项令人瞠目结舌的法案。多亏参议员Richard Shelby，参议院没有批准众议院的愚蠢举动。而更值得赞赏的是，具有投资者思维的美国证券交易委员会主席Bill Donaldson顶住了巨大的政治压力——这些压力来自那些挥着支票的CEO们，他们先是在1993年胁迫国会，去年又故伎重演。
由于试图遮掩股票期权问题的努力仍在继续，值得指出的是——无论是美国财务会计准则委员会（FASB）、一般投资者，还是我——都没有人主张以任何方式限制期权的使用。事实上，我在伯克希尔的继任者可能仍会通过期权获得大部分薪酬，尽管是逻辑合理的期权，条件是：1）适当的行权价，2）反映留存收益的价格递增，以及3）禁止其通过期权购得的股票迅速脱手。我们乐于看到激励管理层的安排，无论是现金奖金还是期权。如果一家公司真的因为发行期权而获得了价值，我们看不出记录其成本会减少其使用有何道理。

简单的事实是：某些CEO知道，如果期权被费用化，他们自己的薪酬将远更为理性地确定。他们还怀疑，如果采用现实主义的会计处理，他们的股票会以更低的价格出售，这意味着当他们抛售个人持股时，在市场上获得的收益会更少。对这些CEO而言，这种不愉快的前景是他们必欲倾其所有资源加以对抗的命运——尽管他们用来对抗的资金通常不属于他们，而是由他们的股东提供的。

期权费用化按计划将于6月15日成为强制规定。因此，从现在到那时，你可以预期会有加剧的努力来拖延或阉割这一规则。请就此事向你们的国会议员和参议员表达你们的看法。

年度会议

今年的年度会议有两项变化。首先，我们将会议安排在4月的最后一个星期六（30日），而不是通常的5月第一个星期六。今年母亲节是5月8日，在这样一个特殊的日子让波仙珠宝和戈拉特牛排馆的员工来招待我们是不公平的——所以我们把一切提前一周。明年我们将恢复常规时间，于2006年5月6日举行会议。

此外，我们改变了4月30日会议当天的活动顺序。和往常一样，奎斯特中心早上7点开门，8点30分放映电影。然而，9点30分我们将直接进入问答环节，该环节（允许在奎斯特中心的摊位吃午餐）将持续到下午3点。然后，在短暂休会后，查理和我在3点15分召开年度会议。

我们做出这一改变，是因为去年有几位股东抱怨，两位发言者占据了大量时间，他们所倡导的提议对大多数听众而言兴趣有限——而且无疑他们很享受有机会对大约19,500名无法脱身的听众发表演讲。有了新流程，那些想听完一切的股东可以留下来参加正式会议，不想听的则可以离开——或者更好的是去购物。

在毗邻会议区的大型展览厅里，这种消遣活动将有很多机会。凯利·穆彻摩尔，伯克希尔的弗洛·齐格菲尔德，去年上演了一场壮观的购物盛宴，她说那只是今年的一次热身。（我很高兴地报告，凯利将在10月结婚。我将把她交到新郎手中，并建议她通过把婚礼放在年度会议上举行来创造一点历史。然而，当查理坚持要当戒指男童时，她退缩了。）

我们将再次展示一座2,100平方英尺的克雷顿家园（采用阿克米砖、萧氏地毯、约翰斯·曼维尔保温材料、米泰克紧固件、凯尔富遮阳篷和内布拉斯加家具商城的家具）。请参观这座房子。更好的是，买下它。
GEICO（政府雇员保险公司）将在现场设立展位，由来自全国各地的多位顶尖顾问提供服务，随时准备为您提供汽车保险报价。大多数情况下，GEICO 能为您提供专属的股东折扣（通常为 8%）。在我们开展业务的 50 个司法管辖区中，有 45 个允许这项特殊优惠。请携带您现有保险的详细信息，看看我们能否为您省钱。

周六，在奥马哈机场，我们将一如既往地展示 NetJets® 的各类飞机，供您参观。请前往 Qwest 中心的 NetJets 展位，了解如何观赏这些飞机。坐大巴来奥马哈，开新飞机回家。

去年，"书虫"书店（The Bookworm）在销售伯克希尔相关书籍方面生意火爆。书店陈列了 18 种图书，共售出 2,920 册，销售额达 61,000 美元。由于我们不向书店收取租金（我一定是心太软了），书店为股东提供八折优惠。今年，我请"书虫"书店加上了格雷厄姆·艾利森（Graham Allison）的《核恐怖主义：最终可预防的灾难》（Nuclear Terrorism: The Ultimate Preventable Catastrophe），这是一本关心国家安危的人必读之作。此外，书店还将首发彼得·考夫曼（Peter Kaufman）编纂的《穷查理宝典》（Poor Charlie's Almanack）。学者们长久以来一直在争论查理是不是本·富兰克林（Ben Franklin）的转世。这本书应该能给出答案。

随本报告附上的委托书材料中有一份附件，说明了如何获取参会及其他活动所需的凭证。至于飞机、酒店和汽车预订，我们再次与 American Express（美国运通，800-799-6634）签约，为您提供特别协助。他们每年都为我们提供出色的服务，我对此深表感谢。

位于第 72 街（介于道奇街与太平洋街之间）占地 77 英亩的内布拉斯加家具城（Nebraska Furniture Mart，简称 NFM），将再次推出“伯克希尔周末”特价活动。八年前我们在 NFM 首次推出这一特别活动，“周末”期间的销售额从 1997 年的 530 万美元增长到 2004 年的 2,510 万美元（比上年增长 45%）。每年都刷新纪录，去年周六更是创下了 NFM 单日最高销售额——610 万美元。

要享受折扣，您必须在 4 月 28 日（星期四）至 5 月 2 日（星期一）期间（含首尾两天）购物，并出示您的参会凭证。这一时期的特价同样适用于多家知名制造商的产品——这些厂商通常有严格的不打折规定，但出于对我们股东周末活动的支持，它们为您破了例。我们感谢它们的合作。NFM 营业时间：周一至周六上午 10 点至晚上 9 点，周日上午 10 点至下午 6 点。今年周六下午 5:30 至晚上 8 点，我们将为股东举办专场活动。我会到场，吃烧烤、喝可乐。

博希姆珠宝（Borsheim's）——全美除蒂芙尼（Tiffany）曼哈顿店之外最大的珠宝店——将举办两场股东专场活动。第一场是 4 月 29 日（星期五）下午 6 点至晚上 10 点的鸡尾酒招待会。第二场是主要庆典，于 5 月 1 日（星期日）上午 9 点至下午 4 点举行。周六我们营业至下午 6 点。

整个周末，博希姆珠宝将人潮涌动。因此，为了方便您购物，股东优惠价将从 4 月 25 日（星期一）持续至 5 月 7 日（星期六）。在此期间，您只需出示参会凭证或券商账户对账单证明股东身份即可。

博希姆珠宝的毛利率比其主要竞争对手低整整 20 个百分点，这还是在股东折扣之前。去年周末的营业额比 2003 年增长了 73%，创下了难以超越的纪录。证明给我看，你们能行。
在博希姆珠宝店外的帐篷里，两次美国国际象棋冠军帕特里克·沃尔夫（Patrick Wolff）将蒙住双眼，同时迎战六人一组的挑战者。此外，周日下午，两位世界顶尖桥牌高手——鲍勃·哈曼（Bob Hamman）和莎伦·奥斯伯格（Sharon Osberg）也将与股东们对弈。他们打算睁着眼睛——不过鲍勃即使在争夺全国冠军时也从不理牌。

我最爱的牛排馆——戈拉特——将于5月1日（周日）再次专门为伯克希尔股东营业，下午4点至10点供应餐食。请记住，当天前往戈拉特必须提前预订。预订请于4月1日（但别提前）致电402-551-3733。如果周日客满，可以试试您在奥马哈的其他晚上去戈拉特。像我一样点一份三分熟的T骨牛排配双份土豆煎饼，能提升您作为美食家的声誉。

周六下午4点到5点30分，我们将再次为来自北美以外的股东举办一场特别招待会。每年我们的股东大会都会吸引全球各地的人，查理和我希望亲自问候那些远道而来的朋友。去年我们见到了400多位股东，其中至少100位来自澳大利亚。任何来自美国或加拿大以外的股东，都将获得特别凭证及参加此次活动的说明。

* * * * * * * * * * * *

查理和我很幸运。我们热爱自己的工作，每天都有才华横溢、乐观积极的同事从方方面面给予帮助。难怪我们跳着踢踏舞去上班。但对我们来说，最快乐的事莫过于在伯克希尔年度股东大会上与我们的股东合伙人相聚。所以，请于4月30日来Qwest中心参加我们的年度"资本家的伍德斯托克"吧。

2005年2月28日

沃伦·E·巴菲特

董事会主席