伯克希尔·哈撒韦 1984 年股东信

标题：1984年致股东信

致伯克希尔·哈撒韦公司的股东：

我们1984年的净资产增加了1.526亿美元，即每股133美元。这听起来相当不错，但实际上表现平平。经济收益必须与产生这些收益的资本相比较来评估。我们过去二十年账面价值的年复合增长率为22.1%（从1964年的每股19.46美元增长到1984年的每股1108.77美元），但1984年的增长率仅为13.6%。

正如我们去年所讨论的，每股内在商业价值的增长才是真正重要的经济衡量标准。但内在商业价值的计算是主观的。在我们的案例中，账面价值是一个有用但略有低估的替代指标。依我判断，1984年内在商业价值与账面价值大致以相同的速度增长。

用我的学术口吻，我过去曾告诉过你们，不断膨胀的资本基础对回报率造成的拖累。不幸的是，我的学术口吻现在正让位于报告口吻。我们历史性的22%的增长率只是历史而已。要想在未来十年内每年获得哪怕15%的收益（假设我们继续执行目前的股息政策，关于这一点稍后将在信中详细说明），我们需要总计约39亿美元的利润。要实现这一目标，需要一些大想法——小打小闹是行不通的。我的总经理合伙人查理·芒格和我目前还没有任何这样的想法，但我们的经验是，它们偶尔会冒出来。（这算不算一个战略计划？）

报告收益来源

下一页的表格显示了伯克希尔报告收益的来源。伯克希尔在许多组成业务中的净所有权权益在1983年年中蓝筹印花公司合并时发生了变化。由于这些变化，表格的前两列提供了衡量基础业务表现的最佳指标。

所有由任何商业实体进行的重大资产出售产生的损益，均与证券交易一起汇总在表格底部附近的栏目中，不计入营业利润。（我们认为任何年度的已实现资本利得或损失数字都是没有意义的，但我们认为多年期间已实现和未实现的资本利得总额非常重要。）

此外，商誉摊销不记入特定业务，而是作为单独项目列出，原因在我1983年年报信件的附录中有概述。

(1) 1983 figures are those for October through December.
(2) 1984 and 1983 are not comparable; major assets were transferred in the mid-year 1983 merger of Blue Chip Stamps.

Sharp-eyed shareholders will notice that the amount of the special GEICO distribution and its location in the table have been changed from the presentation of last year. Though they reclassify and reduce “accounting” earnings, the changes are entirely of form, not of substance. The story behind the changes, however, is interesting.

As reported last year: (1) in mid-1983 GEICO made a tender offer to buy its own shares; (2) at the same time, we agreed by written contract to sell GEICO an amount of its shares that would be proportionately related to the aggregate number of shares GEICO repurchased via the tender from all other shareholders; (3) at completion of the tender, we delivered 350,000 shares to GEICO, received $21 million cash, and were left owning exactly the same percentage of GEICO that we owned before the tender; (4) GEICO’s transaction with us amounted to a proportionate redemption, an opinion rendered us, without qualification, by a leading law firm; (5) the Tax Code logically regards such proportionate redemptions as substantially equivalent to dividends and, therefore, the $21 million we received was taxed at only the 6.9% inter-corporate dividend rate; (6) importantly, that $21 million was far less than the previously-undistributed earnings that had inured to our ownership in GEICO and, thus, from the standpoint of economic substance, was in our view equivalent to a dividend.

Because it was material and unusual, we highlighted the GEICO distribution last year to you, both in the applicable quarterly report and in this section of the annual report. Additionally, we emphasized the transaction to our auditors, Peat, Marwick, Mitchell & Co. Both the Omaha office of Peat Marwick and the reviewing Chicago partner, without objection, concurred with our dividend presentation.

In 1984, we had a virtually identical transaction with General Foods. The only difference was that General Foods repurchased its stock over a period of time in the open market, whereas GEICO had made a “one-shot” tender offer. In the General Foods case we sold to the company, on each day that it repurchased shares, a quantity of shares that left our ownership percentage precisely unchanged. Again our transaction was pursuant to a written contract executed before repurchases began. And again the money we received was far less than the retained earnings that had inured to our ownership interest since our purchase. Overall we received $21,843,601 in cash from General Foods, and our ownership remained at exactly 8.75%.

At this point the New York office of Peat Marwick came into the picture. Late in 1984 it indicated that it disagreed with the conclusions of the firm’s Omaha office and Chicago reviewing partner. The New York view was that the GEICO and General Foods transactions should be treated as sales of stock by Berkshire rather than as the receipt of dividends. Under this accounting approach, a portion of the cost of our investment in the stock of each company would be charged against the redemption payment and any gain would be shown as a capital gain, not as dividend income. This is an accounting approach only, having no bearing on taxes: Peat Marwick agrees that the transactions were dividends for IRS purposes.

We disagree with the New York position from both the viewpoint of economic substance and proper accounting. But, to avoid a qualified auditor’s opinion, we have adopted herein Peat Marwick’s 1984 view and restated 1983 accordingly. None of this, however, has any effect on intrinsic business value: our ownership interests in GEICO and General Foods, our cash, our taxes, and the market value and tax basis of our holdings all remain the same.

This year we have again entered into a contract with General Foods whereby we will sell them shares concurrently with open market purchases that they make. The arrangement provides that our ownership interest will remain unchanged at all times. By keeping it so, we will insure ourselves dividend treatment for tax purposes. In our view also, the economic substance of this transaction again is the creation of dividend income. However, we will account for the redemptions as sales of stock rather than dividend income unless accounting rules are adopted that speak directly to this point. We will continue to prominently identify any such special transactions in our reports to you.

While we enjoy a low tax charge on these proportionate redemptions, and have participated in several of them, we view such repurchases as at least equally favorable for shareholders who do not sell. When companies with outstanding businesses and comfortable financial positions find their shares selling far below intrinsic value in the marketplace, no alternative action can benefit shareholders as surely as repurchases.

(Our endorsement of repurchases is limited to those dictated by price/value relationships and does not extend to the “greenmail” repurchase—a practice we find odious and repugnant. In these transactions, two parties achieve their personal ends by exploitation of an innocent and unconsulted third party. The players are: (1) the “shareholder” extortionist who, even before the ink on his stock certificate dries, delivers his “your-money-or-your-life” message to managers; (2) the corporate insiders who quickly seek peace at any price—as long as the price is paid by someone else; and (3) the shareholders whose money is used by (2) to make (1) go away. As the dust settles, the mugging, transient shareholder gives his speech on “free enterprise”, the muggee management gives its speech on “the best interests of the company”, and the innocent shareholder standing by mutely funds the payoff.)

The companies in which we have our largest investments have all engaged in significant stock repurchases at times when wide discrepancies existed between price and value. As shareholders, we find this encouraging and rewarding for two important reasons—one that is obvious, and one that is subtle and not always understood. The obvious point involves basic arithmetic: major repurchases at prices well below per-share intrinsic business value immediately increase, in a highly significant way, that value. When companies purchase their own stock, they often find it easy to get $2 of present value for $1. Corporate acquisition programs almost never do as well and, in a discouragingly large number of cases, fail to get anything close to $1 of value for each $1 expended.

The other benefit of repurchases is less subject to precise measurement but can be fully as important over time. By making repurchases when a company’s market value is well below its business value, management clearly demonstrates that it is given to actions that enhance the wealth of shareholders, rather than to actions that expand management’s domain but that do nothing for (or even harm) shareholders. Seeing this, shareholders and potential shareholders increase their estimates of future returns from the business. This upward revision, in turn, produces market prices more in line with intrinsic business value. These prices are entirely rational. Investors should pay more for a business that is lodged in the hands of a manager with demonstrated pro-shareholder leanings than for one in the hands of a self-interested manager marching to a different drummer. (To make the point extreme, how much would you pay to be a minority shareholder of a company controlled by Robert Wesco?)

The key word is “demonstrated”. A manager who consistently turns his back on repurchases, when these clearly are in the interests of owners, reveals more than he knows of his motivations. No matter how often or how eloquently he mouths some public relations-inspired phrase such as “maximizing shareholder wealth” (this season’s favorite), the market correctly discounts assets lodged with him. His heart is not listening to his mouth—and, after a while, neither will the market.

We have prospered in a very major way—as have other shareholders—by the large share repurchases of GEICO, Washington Post, and General Foods, our three largest holdings. (Exxon, in which we have our fourth largest holding, has also wisely and aggressively repurchased shares but, in this case, we have only recently established our position.) In each of these companies, shareholders have had their interests in outstanding businesses materially enhanced by repurchases made at bargain prices. We feel very comfortable owning interests in businesses such as these that offer excellent economics combined with shareholder-conscious managements.

The following table shows our 1984 yearend net holdings in marketable equities. All numbers exclude the interests attributable to minority shareholders of Wesco and Nebraska Furniture Mart.

It’s been over ten years since it has been as difficult as now to find equity investments that meet both our qualitative standards and our quantitative standards of value versus price. We try to avoid compromise of these standards, although we find doing nothing the most difficult task of all. (One English statesman attributed his country’s greatness in the nineteenth century to a policy of “masterly inactivity”. This is a strategy that is far easier for historians to commend than for participants to follow.)

In addition to the figures supplied at the beginning of this section, information regarding the businesses we own appears in Management’s Discussion on pages 42-47. An amplified discussion of Wesco’s businesses appears in Charlie Munger’s report on pages 50-59. You will find particularly interesting his comments about conditions in the thrift industry. Our other major controlled businesses are Nebraska Furniture Mart, See’s, Buffalo Evening News, and the Insurance Group, to which we will give some special attention here.

Nebraska Furniture Mart

Last year I introduced you to Mrs. B (Rose Blumkin) and her family. I told you they were terrific, and I understated the case. After another year of observing their remarkable talents and character, I can honestly say that I never have seen a managerial group that either functions or behaves better than the Blumkin family.

Mrs. B, Chairman of the Board, is now 91, and recently was quoted in the local newspaper as saying, “I come home to eat and sleep, and that’s about it. I can’t wait until it gets daylight so I can get back to the business”. Mrs. B is at the store seven days a week, from opening to close, and probably makes more decisions in a day than most CEOs do in a year (better ones, too).

In May Mrs. B was granted an Honorary Doctorate in Commercial Science by New York University. (She’s a “fast track” student: not one day in her life was spent in a school room prior to her receipt of the doctorate.) Previous recipients of honorary degrees in business from NYU include Clifton Garvin, Jr., CEO of Exxon Corp.; Walter Wriston, then CEO of Citicorp; Frank Cary, then CEO of IBM; Tom Murphy, then CEO of General Motors; and, most recently, Paul Volcker. (They are in good company.)

The Blumkin blood did not run thin. Louie, Mrs. B’s son, and his three boys, Ron, Irv, and Steve, all contribute in full measure to NFM’s amazing success. The younger generation has attended the best business school of them all—that conducted by Mrs. B and Louie—and their training is evident in their performance.

Last year NFM’s net sales increased by $14.3 million, bringing the total to $115 million, all from the one store in Omaha. That is by far the largest volume produced by a single home furnishings store in the United States. In fact, the gain in sales last year was itself greater than the annual volume of many good-sized successful stores. The business achieves this success because it deserves this success. A few figures will tell you why.

In its fiscal 1984 10-K, the largest independent specialty retailer of home furnishings in the country, Levitz Furniture, described its prices as “generally lower than the prices charged by conventional furniture stores in its trading area”. Levitz, in that year, operated at a gross margin of 44.4% (that is, on average, customers paid it $100 for merchandise that had cost it $55.60 to buy). The gross margin at NFM is not much more than half of that. NFM’s low mark-ups are possible because of its exceptional efficiency: operating expenses (payroll, occupancy, advertising, etc.) are about 16.5% of sales versus 35.6% at Levitz.

None of this is in criticism of Levitz, which has a well-managed operation. But the NFM operation is simply extraordinary (and, remember, it all comes from a $500 investment by Mrs. B in 1937). By unparalleled efficiency and astute volume purchasing, NFM is able to earn excellent returns on capital while saving its customers at least $30 million annually from what, on average, it would cost them to buy the same merchandise at stores maintaining typical mark-ups. Such savings enable NFM to constantly widen its geographical reach and thus to enjoy growth well beyond the natural growth of the Omaha market.

I have been asked by a number of people just what secrets the Blumkins bring to their business. These are not very esoteric. All members of the family: (1) apply themselves with an enthusiasm and energy that would make Ben Franklin and Horatio Alger look like dropouts; (2) define with extraordinary realism their area of special competence and act decisively on all matters within it; (3) ignore even the most enticing propositions failing outside of that area of special competence; and, (4) unfailingly behave in a high-grade manner with everyone they deal with. (Mrs. B boils it down to “sell cheap and tell the truth”.)

Our evaluation of the integrity of Mrs. B and her family was demonstrated when we purchased 90% of the business: NFM had never had an audit and we did not request one; we did not take an inventory nor verify the receivables; we did not check property titles. We gave Mrs. B a check for $55 million and she gave us her word. That made for an even exchange.

You and I are fortunate to be in partnership with the Blumkin family.

See’s Candy Shops, Inc.

Below is our usual recap of See’s performance since the time of purchase by Blue Chip Stamps:

This performance has not been produced by a generally rising tide. To the contrary, many well-known participants in the boxed-chocolate industry either have lost money in this same period or have been marginally profitable. To our knowledge, only one good-sized competitor has achieved high profitability. The success of See’s reflects the combination of an exceptional product and an exceptional manager, Chuck Huggins.

During 1984 we increased prices considerably less than has been our practice in recent years: per-pound realization was $5.49, up only 1.4% from 1983. Fortunately, we made good progress on cost control, an area that has caused us problems in recent years. Per-pound costs—other than those for raw materials, a segment of expense largely outside of our control—increased by only 2.2% last year.

Our cost-control problem has been exacerbated by the problem of modestly declining volume (measured by pounds, not dollars) on a same-store basis. Total pounds sold through shops in recent years has been maintained at a roughly constant level only by the net addition of a few shops annually. This more-shops-to-get-the-same-volume situation naturally puts heavy pressure on per-pound selling costs.

In 1984, same-store volume declined 1.1%. Total shop volume, however, grew 0.6% because of an increase in stores. (Both percentages are adjusted to compensate for a 53-week fiscal year in 1983.)

See’s business tends to get a bit more seasonal each year. In the four weeks prior to Christmas, we do 40% of the year’s volume and earn about 75% of the year’s profits. We also earn significant sums in the Easter and Valentine’s Day periods, but pretty much tread water the rest of the year. In recent years, shop volume at Christmas has grown in relative importance, and so have quantity orders and mail orders. The increased concentration of business in the Christmas period produces a multitude of managerial problems, all of which have been handled by Chuck and his associates with exceptional skill and grace.

Their solutions have in no way involved compromises in either quality of service or quality of product. Most of our larger competitors could not say the same. Though faced with somewhat less extreme peaks and valleys in demand than we, they add preservatives or freeze the finished product in order to smooth the production cycle and thereby lower unit costs. We reject such techniques, opting, in effect, for production headaches rather than product modification.

Our mall stores face a host of new food and snack vendors that provide particularly strong competition at non-holiday periods. We need new products to fight back and during 1984 we introduced six candy bars that, overall, met with a good reception. Further product introductions are planned.

In 1985 we will intensify our efforts to keep per-pound cost increases below the rate of inflation. Continued success in these efforts, however, will require gains in same-store poundage. Prices in 1985 should average 6%—7% above those of 1984. Assuming no change in same-store volume, profits should show a moderate gain.

Buffalo Evening News

Profits at the News in 1984 were considerably greater than we expected. As at See’s, excellent progress was made in controlling costs. Excluding hours worked in the newsroom, total hours worked decreased by about 2.8%. With this productivity improvement, overall costs increased only 4.9%. This performance by Stan Lipsey and his management team was one of the best in the industry.

However, we now face an acceleration in costs. In mid-1984 we entered into new multi-year union contracts that provided for a large “catch-up” wage increase. This catch-up is entirely appropriate: the cooperative spirit of our unions during the unprofitable 1977-1982 period was an important factor in our success in remaining cost competitive with The Courier-Express. Had we not kept costs down, the outcome of that struggle might well have been different.

Because our new union contracts took effect at varying dates, little of the catch-up increase was reflected in our 1984 costs. But the increase will be almost totally effective in 1985 and, therefore, our unit labor costs will rise this year at a rate considerably greater than that of the industry. We expect to mitigate this increase by continued small gains in productivity, but we cannot avoid significantly higher wage costs this year. Newsprint price trends also are less favorable now than they were in 1984. Primarily because of these two factors, we expect at least a minor contraction in margins at the News.

Working in our favor at the News are two factors of major economic importance:

1. Our circulation is concentrated to an unusual degree in the area of maximum utility to our advertisers. “Regional” newspapers with wide-ranging circulation, on the other hand, have a significant portion of their circulation in areas that are of negligible utility to most advertisers. A subscriber several hundred miles away is not much of a prospect for the puppy you are offering to sell via a classified ad—nor for the grocer with stores only in the metropolitan area. “Wasted” circulation—as the advertisers call it -hurts profitability: expenses of a newspaper are determined largely by gross circulation while advertising revenues (usually 70%—80% of total revenues) are responsive only to useful circulation;

2. Our penetration of the Buffalo retail market is exceptional; advertisers can reach almost all of their potential customers using only the News.

Last year I told you about this unusual reader acceptance: among the 100 largest newspapers in the country, we were then number one, daily, and number three, Sunday, in penetration. The most recent figures show us number one in penetration on weekdays and number two on Sunday. (Even so, the number of households in Buffalo has declined, so our current weekday circulation is down slightly; on Sundays it is unchanged.)

I told you also that one of the major reasons for this unusual acceptance by readers was the unusual quantity of news that we delivered to them: a greater percentage of our paper is devoted to news than is the case at any other dominant paper in our size range. In 1984 our “news hole” ratio was 50.9%, (versus 50.4% in 1983), a level far above the typical 35%-40%. We will continue to maintain this ratio in the 50% area. Also, though we last year reduced total hours worked in other departments, we maintained the level of employment in the newsroom and, again, will continue to do so. Newsroom costs advanced 9.1% in 1984, a rise far exceeding our overall cost increase of 4.9%.

Our news hole policy costs us significant extra money for newsprint. As a result, our news costs (newsprint for the news hole plus payroll and expenses of the newsroom) as a percentage of revenue run higher than those of most dominant papers of our size. There is adequate room, however, for our paper or any other dominant paper to sustain these costs: the difference between “high” and “low” news costs at papers of comparable size runs perhaps three percentage points while pre-tax profit margins are often ten times that amount.

The economics of a dominant newspaper are excellent, among the very best in the business world. Owners, naturally, would like to believe that their wonderful profitability is achieved only because they unfailingly turn out a wonderful product. That comfortable theory wilts before an uncomfortable fact. While first-class newspapers make excellent profits, the profits of third-rate papers are as good or better—as long as either class of paper is dominant within its community. Of course, product quality may have been crucial to the paper in achieving dominance. We believe this was the case at the News, in very large part because of people such as Alfred Kirchhofer who preceded us.

Once dominant, the newspaper itself, not the marketplace, determines just how good or how bad the paper will be. Good or bad, it will prosper. That is not true of most businesses: inferior quality generally produces inferior economics. But even a poor newspaper is a bargain to most citizens simply because of its “bulletin board” value. Other things being equal, a poor product will not achieve quite the level of readership achieved by a first-class product. A poor product, however, will still remain essential to most citizens, and what commands their attention will command the attention of advertisers.

Since high standards are not imposed by the marketplace, management must impose its own. Our commitment to an above-average expenditure for news represents an important quantitative standard. We have confidence that Stan Lipsey and Murray Light will continue to apply the far-more important qualitative standards. Charlie and I believe that newspapers are very special institutions in society. We are proud of the News, and intend an even greater pride to be justified in the years ahead.

Insurance Operations

Shown below is an updated version of our usual table listing two key figures for the insurance industry:

Source: Best’s Aggregates and Averages.

Best’s data reflect the experience of practically the entire industry, including stock, mutual, and reciprocal companies. The combined ratio represents total insurance costs (losses incurred plus expenses) compared to revenue from premiums; a ratio below 100 indicates an underwriting profit, and one above 100 indicates a loss.

For a number of years, we have told you that an annual increase by the industry of about 10% per year in premiums written is necessary for the combined ratio to remain roughly unchanged. We assumed in making that assertion that expenses as a percentage of premium volume would stay relatively stable and that losses would grow at about 10% annually because of the combined influence of unit volume increases, inflation, and judicial rulings that expand what is covered by the insurance policy.

Our opinion is proving dismayingly accurate: a premium increase of 10% per year since 1979 would have produced an aggregate increase through 1984 of 61% and a combined ratio in 1984 almost identical to the 100.6 of 1979. Instead, the industry had only a 30% increase in premiums and a 1984 combined ratio of 117.7. Today, we continue to believe that the key index to the trend of underwriting profitability is the year-to-year percentage change in industry premium volume.

It now appears that premium volume in 1985 will grow well over 10%. Therefore, assuming that catastrophes are at a “normal” level, we would expect the combined ratio to begin easing downward toward the end of the year. However, under our industrywide loss assumptions (i.e., increases of 10% annually), five years of 15%-per-year increases in premiums would be required to get the combined ratio back to 100. This would mean a doubling of industry volume by 1989, an outcome that seems highly unlikely to us. Instead, we expect several years of premium gains somewhat above the 10% level, followed by highly-competitive pricing that generally will produce combined ratios in the 108-113 range.

Our own combined ratio in 1984 was a humbling 134. (Here, as throughout this report, we exclude structured settlements and the assumption of loss reserves in reporting this ratio. Much additional detail, including the effect of discontinued operations on the ratio, appears on pages 42-43). This is the third year in a row that our underwriting performance has been far poorer than that of the industry. We expect an improvement in the combined ratio in 1985, and also expect our improvement to be substantially greater than that of the industry. Mike Goldberg has corrected many of the mistakes I made before he took over insurance operations. Moreover, our business is concentrated in lines that have experienced poorer-than-average results during the past several years, and that circumstance has begun to subdue many of our competitors and even eliminate some. With the competition shaken, we were able during the last half of 1984 to raise prices significantly in certain important lines with little loss of business.

For some years I have told you that there could be a day coming when our premier financial strength would make a real difference in the competitive position of our insurance operation. That day may have arrived. We are almost without question the strongest property/casualty insurance operation in the country, with a capital position far superior to that of well-known companies of much greater size.

Equally important, our corporate policy is to retain that superiority. The buyer of insurance receives only a promise in exchange for his cash. The value of that promise should be appraised against the possibility of adversity, not prosperity. At a minimum, the promise should appear able to withstand a prolonged combination of depressed financial markets and exceptionally unfavorable underwriting results. Our insurance subsidiaries are both willing and able to keep their promises in any such environment—and not too many other companies clearly are.

Our financial strength is a particular asset in the business of structured settlements and loss reserve assumptions that we reported on last year. The claimant in a structured settlement and the insurance company that has reinsured loss reserves need to be completely confident that payments will be forthcoming for decades to come. Very few companies in the property/casualty field can meet this test of unquestioned long-term strength. (In fact, only a handful of companies exists with which we will reinsure our own liabilities.)

We have grown in these new lines of business: funds that we hold to offset assumed liabilities grew from $16.2 million to $30.6 million during the year. We expect growth to continue and perhaps to greatly accelerate. To support this projected growth we have added substantially to the capital of Columbia Insurance Company, our reinsurance unit specializing in structured settlements and loss reserve assumptions. While these businesses are very competitive, returns should be satisfactory.

At GEICO the news, as usual, is mostly good. That company achieved excellent unit growth in its primary insurance business during 1984, and the performance of its investment portfolio continued to be extraordinary. Though underwriting results deteriorated late in the year, they still remain far better than those of the industry. Our ownership in GEICO at yearend amounted to 36% and thus our interest in their direct property/casualty volume of $885 million amounted to $320 million, or well over double our own premium volume.

I have reported to you in the past few years that the performance of GEICO’s stock has considerably exceeded that company’s business performance, brilliant as the latter has been. In those years, the carrying value of our GEICO investment on our balance sheet grew at a rate greater than the growth in GEICO’s intrinsic business value. I warned you that over performance by the stock relative to the performance of the business obviously could not occur every year, and that in some years the stock must under perform the business. In 1984 that occurred and the carrying value of our interest in GEICO changed hardly at all, while the intrinsic business value of that interest increased substantially.

Since 27% of Berkshire’s net worth at the beginning of 1984 was represented by GEICO, its static market value had a significant impact upon our rate of gain for the year. We are not at all unhappy with such a result: we would far rather have the business value of GEICO increase by X during the year, while market value decreases, than have the intrinsic value increase by only 1/2 X with market value soaring. In GEICO’s case, as in all of our investments, we look to business performance, not market performance. If we are correct in expectations regarding the business, the market eventually will follow along.

You, as shareholders of Berkshire, have benefited in enormous measure from the talents of GEICO’s Jack Byrne, Bill Snyder, and Lou Simpson. In its core business—low-cost auto and homeowners insurance—GEICO has a major, sustainable competitive advantage. That is a rare asset in business generally, and it’s almost non-existent in the field of financial services. (GEICO, itself, illustrates this point: despite the company’s excellent management, superior profitability has eluded GEICO in all endeavors other than its core business.) In a large industry, a competitive advantage such as GEICO’s provides the potential for unusual economic rewards, and Jack and Bill continue to exhibit great skill in realizing that potential.

Most of the funds generated by GEICO’s core insurance operation are made available to Lou for investment. Lou has the rare combination of temperamental and intellectual characteristics that produce outstanding long-term investment performance. Operating with below-average risk, he has generated returns that have been by far the best in the insurance industry. I applaud and appreciate the efforts and talents of these three outstanding managers.

Errors in Loss Reserving

Any shareholder in a company with important interests in the property/casualty insurance business should have some understanding of the weaknesses inherent in the reporting of current earnings in that industry. Phil Graham, when publisher of the Washington Post, described the daily newspaper as “a first rough draft of history”. Unfortunately, the financial statements of a property/casualty insurer provide, at best, only a first rough draft of earnings and financial condition.

The determination of costs is the main problem. Most of an insurer’s costs result from losses on claims, and many of the losses that should be charged against the current year’s revenue are exceptionally difficult to estimate. Sometimes the extent of these losses, or even their existence, is not known for decades.

The loss expense charged in a property/casualty company’s current income statement represents: (1) losses that occurred and were paid during the year; (2) estimates for losses that occurred and were reported to the insurer during the year, but which have yet to be settled; (3) estimates of ultimate dollar costs for losses that occurred during the year but of which the insurer is unaware (termed “IBNR”: incurred but not reported); and (4) the net effect of revisions this year of similar estimates for (2) and (3) made in past years.

Such revisions may be long delayed, but eventually any estimate of losses that causes the income for year X to be misstated must be corrected, whether it is in year X + 1, or X + 10. This, perforce, means that earnings in the year of correction also are misstated. For example, assume a claimant was injured by one of our insureds in 1979 and we thought a settlement was likely to be made for $10,000. That year we would have charged $10,000 to our earnings statement for the estimated cost of the loss and, correspondingly, set up a liability reserve on the balance sheet for that amount. If we settled the claim in 1984 for $100,000, we would charge earnings with a loss cost of $90,000 in 1984, although that cost was truly an expense of 1979. And if that piece of business was our only activity in 1979, we would have badly misled ourselves as to costs, and you as to earnings.

The necessarily-extensive use of estimates in assembling the figures that appear in such deceptively precise form in the income statement of property/casualty companies means that some error must seep in, no matter how proper the intentions of management. In an attempt to minimize error, most insurers use various statistical techniques to adjust the thousands of individual loss evaluations (called case reserves) that comprise the raw data for estimation of aggregate liabilities. The extra reserves created by these adjustments are variously labeled “bulk”, “development”, or “supplemental” reserves. The goal of the adjustments should be a loss-reserve total that has a 50-50 chance of being proved either slightly too high or slightly too low when all losses that occurred prior to the date of the financial statement are ultimately paid.

At Berkshire, we have added what we thought were appropriate supplemental reserves but in recent years they have not been adequate. It is important that you understand the magnitude of the errors that have been involved in our reserving. You can thus see for yourselves just how imprecise the process is, and also judge whether we may have some systemic bias that should make you wary of our current and future figures.

The following table shows the results from insurance underwriting as we have reported them to you in recent years, and also gives you calculations a year later on an “if-we-knew-then-what-we think-we-know-now” basis. I say “what we think we know now” because the adjusted figures still include a great many estimates for losses that occurred in the earlier years. However, many claims from the earlier years have been settled so that our one-year-later estimate contains less guess work than our earlier estimate:

Our structured settlement and loss-reserve assumption businesses are not included in this table. Important additional information on loss reserve experience appears on pages 43-45.

To help you understand this table, here is an explanation of the most recent figures: 1984’s reported pre-tax underwriting loss of $45.4 million consists of $27.6 million we estimate that we lost on 1984’s business, plus the increased loss of $17.8 million reflected in the corrected figure for 1983.

As you can see from reviewing the table, my errors in reporting to you have been substantial and recently have always presented a better underwriting picture than was truly the case. This is a source of particular chagrin to me because: (1) I like for you to be able to count on what I say; (2) our insurance managers and I undoubtedly acted with less urgency than we would have had we understood the full extent of our losses; and (3) we paid income taxes calculated on overstated earnings and thereby gave the government money that we didn’t need to. (These overpayments eventually correct themselves, but the delay is long and we don’t receive interest on the amounts we overpaid.)

Because our business is weighted toward casualty and reinsurance lines, we have more problems in estimating loss costs than companies that specialize in property insurance. (When a building that you have insured burns down, you get a much faster fix on your costs than you do when an employer you have insured finds out that one of his retirees has contracted a disease attributable to work he did decades earlier.) But I still find our errors embarrassing. In our direct business, we have far underestimated the mushrooming tendency of juries and courts to make the “deep pocket” pay, regardless of the factual situation and the past precedents for establishment of liability. We also have underestimated the contagious effect that publicity regarding giant awards has on juries. In the reinsurance area, where we have had our worst experience in under reserving, our customer insurance companies have made the same mistakes. Since we set reserves based on information they supply us, their mistakes have become our mistakes.

I heard a story recently that is applicable to our insurance accounting problems: a man was traveling abroad when he received a call from his sister informing him that their father had died unexpectedly. It was physically impossible for the brother to get back home for the funeral, but he told his sister to take care of the funeral arrangements and to send the bill to him. After returning home he received a bill for several thousand dollars, which he promptly paid. The following month another bill came along for $15, and he paid that too. Another month followed, with a similar bill. When, in the next month, a third bill for $15 was presented, he called his sister to ask what was going on. “Oh”, she said. “I forgot to tell you. We buried Dad in a rented suit.”

If you’ve been in the insurance business in recent years—particularly the reinsurance business—this story hurts. We have tried to include all of our “rented suit” liabilities in our current financial statement, but our record of past error should make us humble, and you suspicious. I will continue to report to you the errors, plus or minus, that surface each year.

Not all reserving errors in the industry have been of the innocent-but-dumb variety. With underwriting results as bad as they have been in recent years—and with managements having as much discretion as they do in the presentation of financial statements—some unattractive aspects of human nature have manifested themselves. Companies that would be out of business if they realistically appraised their loss costs have, in some cases, simply preferred to take an extraordinarily optimistic view about these yet-to-be-paid sums. Others have engaged in various transactions to hide true current loss costs.

Both of these approaches can “work” for a considerable time: external auditors cannot effectively police the financial statements of property/casualty insurers. If liabilities of an insurer, correctly stated, would exceed assets, it falls to the insurer to volunteer this morbid information. In other words, the corpse is supposed to file the death certificate. Under this “honor system” of mortality, the corpse sometimes gives itself the benefit of the doubt.

In most businesses, of course, insolvent companies run out of cash. Insurance is different: you can be broke but flush. Since cash comes in at the inception of an insurance policy and losses are paid much later, insolvent insurers don’t run out of cash until long after they have run out of net worth. In fact, these “walking dead” often redouble their efforts to write business, accepting almost any price or risk, simply to keep the cash flowing in. With an attitude like that of an embezzler who has gambled away his purloined funds, these companies hope that somehow they can get lucky on the next batch of business and thereby cover up earlier shortfalls. Even if they don’t get lucky, the penalty to managers is usually no greater for a $100 million shortfall than one of $10 million; in the meantime, while the losses mount, the managers keep their jobs and perquisites.

The loss-reserving errors of other property/casualty companies are of more than academic interest to Berkshire. Not only does Berkshire suffer from sell-at-any-price competition by the “walking dead”, but we also suffer when their insolvency is finally acknowledged. Through various state guarantee funds that levy assessments, Berkshire ends up paying a portion of the insolvent insurers’ asset deficiencies, swollen as they usually are by the delayed detection that results from wrong reporting. There is even some potential for cascading trouble. The insolvency of a few large insurers and the assessments by state guarantee funds that would follow could imperil weak-but-previously-solvent insurers. Such dangers can be mitigated if state regulators become better at prompt identification and termination of insolvent insurers, but progress on that front has been slow.

Washington Public Power Supply System

标题：1984年致股东的信

第二部分

从1983年10月到1984年6月，伯克希尔的保险子公司持续大量购买华盛顿公共电力供应系统（WPPSS）项目1、2和3的债券。正是同一个主体，在1983年7月1日，对用于资助现已废弃的项目4和5部分建设的22亿美元债券违约。虽然这两类债券在债务人、承诺和基础资产上存在重大差异，但项目4和5的问题给项目1、2和3蒙上了巨大阴影，并可能对后者造成严重问题。此外，还存在许多直接与项目1、2和3相关的问题，这些问题可能会削弱或摧毁由邦纳维尔电力管理局担保所带来的原本强劲的信用地位。

尽管存在这些重要的负面因素，查理和我判断，在我们购买债券时以及以伯克希尔支付的价格（远低于当前价格），风险被获利前景充分补偿，甚至绰绰有余。

如你所知，我们为保险公司购买可流通股票，基于我们在购买整个企业时会采用的标准。这种企业估值方法在专业资金管理者中并不普遍，且受到许多学术界的嘲笑。然而，它确实为遵循者带来了良好服务（学术界对此似乎是说：“嗯，它在实践中可能没问题，但在理论上永远行不通。”）简而言之，我们觉得，如果我们能够以整个企业每股价值的一小部分，购买具有满意基础经济特征的小块业务，那么好事很可能会发生在我们身上——特别是如果我们拥有这样一组证券的话。

我们将这种企业估值方法甚至扩展到债券购买，例如WPPSS。我们将年底在WPPSS投资的1.39亿美元成本与类似的1.39亿美元投资于运营企业进行比较。就WPPSS而言，该“企业”按合同每年赚取2270万美元税后利润（通过债券支付的利息），这些收益目前以现金形式供我们使用。我们无法购买具有类似经济特征的运营企业。只有相对较少的企业能够在无杠杆资本上获得16.3%的税后收益，就像我们的WPPSS投资那样，而这些企业在可供购买时，其售价相对于该资本有大幅溢价。在平均的商业交易谈判中，2270万美元的税后无杠杆企业收益（相当于约4500万美元税前）可能价值2.5亿至3亿美元（有时甚至更高）。对于一家我们非常了解且强烈喜欢的企业，我们乐意支付那么多。但这是我们从WPPSS债券获取相同收益所需支付价格的两倍。

然而，对于WPPSS，存在我们认为非常轻微的风险，即该“企业”可能在一两年内一文不值。还存在利息支付可能中断相当长一段时间的风险。此外，该“企业”最多能值我们拥有的债券面值约2.05亿美元，这仅比我们支付的价格高出48%。

这个上涨潜力上限是一个重要的缺点。然而，应该认识到，绝大多数运营企业的上涨潜力也有限，除非持续投入更多资本。这是因为大多数企业无法显著提高其平均净资产收益率——即使在通货膨胀条件下也是如此，尽管人们曾认为通货膨胀会自动提高回报率。

（让我们进一步推进债券即企业的例子：如果你选择通过用票息收益购买更多债券来“留存”12%债券的年度收益，那么该债券“企业”的收益将以与大多数类似留存所有收益的运营企业相当的速度增长。第一种情况，今天以1000万美元购买的30年期、零息票、12%债券，到2015年将价值3亿美元。第二种情况，一家1000万美元的企业，定期获得12%的净资产收益率并留存所有收益用于增长，到2015年也将拥有3亿美元的资本。企业和债券在最后一年都将赚取超过3200万美元的收益。）

我们处理债券投资的方法——将其视为一种具有特殊优势和劣势的独特“企业”——可能会让你觉得有点古怪。然而，我们相信，如果投资者以企业家的视角看待债券投资，许多令人震惊的错误本可以避免。例如，在1946年，20年期AAA免税债券的交易收益率略低于1%。实际上，当时购买这些债券的投资者购买了一个“企业”，该企业以“账面价值”赚取约1%的收益（而且，它永远无法赚取超过账面价值1%的收益），并以100美分兑1美元的价格购买了这个糟糕的企业。

如果投资者有足够商业头脑，以这种方式思考——而这正是所达成交易的精确现实——他会对这个提议嗤之以鼻然后走开。因为，与此同时，拥有良好未来前景的企业可以按账面价值或接近账面价值购买，同时赚取账面价值的10%、12%或15%税后收益。1946年，可能没有一家美国企业在账面价值易手时，买家认为其缺乏赚取超过账面价值1%收益的能力。但具有债券购买习惯的投资者全年都热切地基于此做出经济承诺。类似的情况（尽管不那么极端）在接下来的二十年中持续存在，债券投资者欣然签署二十年或三十年的条款，其条件按商业标准看是极其不充分的。（在我看来，有史以来最好的投资书籍——本·格雷厄姆的《聪明的投资者》——最后一章的最后一部分开头是：“当投资最具有商业性时，它才是最聪明的。”这部分被称为“最后的话”，这标题很恰当。）

我们再次强调，WPPSS承诺无疑存在一些风险。这也是那种难以评估的风险。如果查理和我在一生中要处理50个类似的评估，我们会期望我们的判断能证明是相当令人满意的。但我们没有机会在一年内做出50个甚至5个这样的决定。即使我们的长期结果可能最终不错，在任何给定年份，我们都面临着看起来极其愚蠢的风险。（这就是为什么所有这些句子都说“查理和我”，或“我们”。）

大多数经理人几乎没有动力去做出聪明但有一定概率看起来像傻瓜的决定。他们的个人得失比太明显了：如果一个非常规的决策结果良好，他们会得到拍拍背的鼓励；如果结果糟糕，他们会得到解雇通知。（按常规失败才是出路；作为群体，旅鼠可能形象不佳，但从未有单个旅鼠收到过负面报道。）

我们的情况不同。查理和我持有伯克希尔47%的股份，我们不担心被解雇，我们以所有者的身份，而非管理者的身份获得回报。因此，我们对待伯克希尔的资金就像对待自己的资金一样。这经常导致我们在投资和一般企业管理中采取非常规行为。

我们在集中投资保险公司的资产（包括WPPSS债券）方面仍然保持非常规。这种集中只有在我们的保险业务建立在异常财务实力的基础上才有意义。对于几乎所有其他保险公司来说，类似的集中程度（或任何接近它的程度）是完全不合适的。它们的资本状况不足以承受一个大错误，无论从概率角度分析看起来多么有吸引力的投资机会。

凭借我们的财务实力，我们可以拥有少数几只我们深思熟虑并以有吸引力的价格购买的证券的大块头。（比利·罗斯描述了过度分散化的问题：“如果你有一个四十个女人的后宫，你永远无法深入了解她们中的任何一个。”）随着时间的推移，我们的集中政策应该会产生优越的结果，尽管这些结果会受到我们庞大规模的抑制。当这项政策带来一个真正糟糕的年份时（正如它必然会发生的），至少你会知道我们的资金是在同样的基础上投入的。

我们在不同的价格和与当前存在一定不同的事实情况下进行了WPPSS投资的大部分。如果我们决定改变我们的立场，我们不会在变更完成很久之前通知股东。（在你阅读本文时，我们可能正在买入或卖出。）买卖证券是一项竞争性业务，即使是任何一方增加一点竞争，都可能会让我们损失大量资金。我们的WPPSS购买就说明了这个原则。从1983年10月到1984年6月，我们试图购买尽可能多的项目1、2和3的债券。然而，我们只购买了不到3%的流通债券。如果我们面临少数几个额外的资金充足的投资者，他们因为我们买入而受到刺激也买入，我们最终可能获得数量少得多的债券，并且以高得多的价格购买。（几个跟风者很容易让我们花费500万美元。）出于这个原因，我们不会就我们的证券活动发表评论——无论是向媒体、股东还是其他任何人——除非法律要求我们这样做。

关于我们WPPSS购买的一个最终观察：在大多数情况下，我们不喜欢购买大多数长期债券，并且在最近几年很少购买。那是因为债券和美元一样稳健——而我们看待美元的长期前景是黯淡的。我们相信未来存在严重的通货膨胀，尽管我们不知道平均利率会是多少。此外，我们认为存在很小但并非微不足道的恶性通货膨胀的可能性。

考虑到通货膨胀率已经下降的程度，这种可能性似乎荒谬。但我们认为当前的财政政策——以巨额赤字为特征——既极其危险又难以扭转。（到目前为止，两党的大多数政治家都遵循了查理·布朗的建议：“没有哪个问题是如此之大以至于不能逃避的。”）没有逆转，高通胀率可能被推迟（也许是长时间），但不会被避免。如果高通胀率出现，它们会带来恶性上涨螺旋的可能性。

虽然当年度通胀率在5%-10%范围内时，债券和股票（作为类别）之间没有太多可选择的，但恶性通货膨胀则是另一回事。在这种情况下，多样化股票投资组合几乎肯定会遭受实际价值的巨大损失。但已发行的债券将遭受更大的损失。因此，我们认为全债券投资组合具有很小但不可接受的“清零”风险，我们要求任何长期债券的购买都要通过一个特殊的门槛。只有当债券购买显得明显优于其他商业机会时，我们才会参与其中。这些机会很可能很少且相隔甚远。

股息政策

股息政策经常向股东报告，但很少得到解释。一家公司可能会这样说：“我们的目标是将40%至50%的收益用于派息，并以至少等于CPI涨幅的比率增加股息。”然后就这样了——不会提供任何分析来说明为什么该特定政策对企业的所有者最有利。然而，资本配置对商业和投资管理至关重要。正因为如此，我们相信管理者和所有者应该认真思考在什么情况下收益应该留存，以及在什么情况下应该分配。

首先要理解的是，所有收益并非生而平等。在许多企业中，特别是那些资产/利润比率高的企业，通货膨胀会导致部分或全部报告收益变成虚假的。虚假部分——让我们称这些收益为“受限”收益——如果企业要保持其经济地位，就不能作为股息分配。如果这些收益被发放出去，企业将在以下一个或多个领域失去优势：维持单位销售量的能力、长期竞争地位、财务实力。无论其派息率多么保守，一家持续分配受限收益的公司注定要灭亡，除非股权资本另有注入。

受限收益对所有者来说很少毫无价值，但它们通常必须被大幅折现。实际上，它们被企业征用了，无论其经济潜力多么差。（这种无论回报多么无吸引力都要留存的情况，在十年前由联合爱迪生公司以一种绝妙的讽刺方式无意中传达出来。当时，惩罚性的监管政策是导致该公司股票价格低至账面价值四分之一的主要因素；也就是说，每留存一美元收益用于企业再投资，这一美元就变成了仅25美分的市场价值。但是，尽管有这个点金成铅的过程，大部分收益还是再投资于企业，而不是支付给所有者。与此同时，在纽约各地的建设和维护工地上，标语自豪地宣告着公司的口号：“我们必须挖掘”。）

在这个股息讨论中，我们无需进一步关注受限收益。让我们转向更有价值的非受限收益。这些收益既可以留存也可以分配，两者可行性相同。我们认为，管理层应该选择对企业的所有者更有意义的那个方向。

这一原则并未被普遍接受。出于多种原因，管理者喜欢将非受限、易于分配的收益截留于股东之外——以扩大他们统治的企业帝国，以在异常舒适的财务条件下运营等。但我们相信留存只有一个有效的理由。只有当有合理的前景——最好由历史证据支持，或在适当情况下，经过对未来深思熟虑的分析——表明公司每留存一美元，至少为所有者创造一美元的市场价值时，才应该留存非受限收益。只有当留存的资本产生的增量收益等于或高于投资者普遍可获得的收益时，这种情况才会发生。

为了说明，让我们假设一位投资者持有一份无风险的10%永续债券，具有一个非常不寻常的特征。每年，投资者可以选择以现金形式领取其10%的票息，或者将票息再投资于更多相同条款的10%债券；即永续期限和提供相同现金或再投资选择的票息。如果在任何给定年份，长期无风险债券的现行利率为5%，那么投资者以现金形式领取票息将是愚蠢的，因为他可以选择获得的10%债券的价值将远高于100美分兑1美元。在这种情况下，想要获得现金的投资者应该以额外债券的形式领取票息，然后立即出售它们。通过这样做，他将比直接以现金形式领取票息获得更多现金。假设所有债券都由理性投资者持有，在5%利率的时代，没有人会选择现金，即使是需要现金用于生活目的的债券持有人。

然而，如果利率是15%，那么没有理性投资者会希望他的钱以10%的利率被投资。相反，投资者会选择以现金形式领取票息，即使他个人对现金的需求为零。相反的做法——再投资票息——会使投资者获得额外债券，其市场价值远低于他本可以选择获得的现金。如果他想要10%的债券，他可以直接收取现金后在市场上购买，这些债券在那里将以大幅折扣出售。

与我们假设的债券持有人所做的类似分析，对于所有者在思考公司的非受限收益是应该留存还是分配时是合适的。当然，分析要困难得多且容易出错，因为再投资收益的回报率不是像我们债券案例中的合同数字，而是一个波动的数字。所有者必须猜测该利率在中期未来的平均值。然而，一旦做出了有根据的猜测，剩下的分析就很简单了：如果留存收益有望获得高回报，你希望收益再投资；如果再投资可能产生低回报，你希望收益支付给你。

许多公司管理者在决定子公司是否应向母公司分配收益时，很大程度上遵循这样的推理思路。在那个层面上，管理者能够毫不费力地像聪明的所有者那样思考。但母公司层面的派息决策通常是另一回事。在这里，管理者常常难以站在他们的股东-所有者的立场上思考。

采用这种分裂的方法，一家多部门公司的CEO会指示子公司A（其增量资本的预期平均收益率为5%）分配所有可用收益，以便它们可以投资于子公司B（其增量资本的预期收益率为15%）。CEO的商学院誓言不允许有更少的行为。但是，如果他自己在增量资本上的长期记录是5%——而市场利率是10%——他很可能对母公司股东施加的股息政策仅仅是遵循某种历史或行业范围的派息模式。此外，他会期望子公司的经理们向他充分说明为什么收益留在其运营中而非分配给母公司所有者是有道理的。但他很少会向他的所有者提供类似的关于整个公司的分析。

在判断管理者是否应该留存收益时，股东不应仅仅将最近几年的总增量收益与总增量资本进行比较，因为这种关系可能被核心业务中发生的事情所扭曲。在通货膨胀时期，拥有具有非凡经济特征的核心业务的公司可以在该业务中使用少量增量资本，获得非常高的回报率（如去年关于商誉的部分所述）。但是，除非它们经历巨大的单位增长，否则优秀的企业根据定义会产生大量过剩现金。如果一家公司将这些资金中的大部分投入其他低回报的业务，那么公司的整体留存资本回报率可能仍然看起来很好，因为增量投资于核心业务的那部分收益所获得的非凡回报。这种情况类似于职业与业余选手混合高尔夫赛事：即使所有业余选手都是无可救药的笨蛋，团队的最佳球成绩也会因职业选手的卓越技术而体面。

许多公司在股权回报率和整体增量资本上都持续显示出良好回报，但确实将其留存收益的很大一部分用于经济上不具吸引力，甚至是灾难性的方式。然而，它们绝妙的、收益逐年增长的核心业务，掩盖了其他地方资本配置的反复失败（通常涉及以高价收购具有固有平庸经济特征的业务）。有过失的管理者定期报告他们从最新失望中学到的教训。然后他们通常会寻找未来的教训。（失败似乎冲昏了他们的头脑。）

在这种情况下，如果收益仅用于扩展高回报业务，而余额以股息形式支付或用于回购股票（此举增加了所有者在卓越业务中的权益，同时使他们免于参与低劣业务），股东会好得多。高回报业务的管理者，如果持续将这些业务产生的大量现金中的大部分用于其他低回报业务，应该为这些配置决策负责，无论整个企业的盈利能力如何。

这个讨论中的任何内容都不是为了主张股息随着收益或投资机会的每一次波动而季度间跳动。上市公司的股东当然更希望股息保持一致和可预测。因此，支付应反映对收益和增量资本回报率的长期预期。由于长期公司前景变化不频繁，股息模式的变化也不应更频繁。但随着时间的推移，管理者留存的非受限收益应该证明其合理性。如果收益被不明智地留存，那么很可能管理者也被不明智地留用了。

现在让我们转向伯克希尔·哈撒韦，审视这些股息原则如何适用于它。从历史上看，伯克希尔的留存收益回报率远高于市场利率，从而每留存一美元创造了超过一美元的市场价值。在这种情况下，任何派息都将违背大小股东的利益。

事实上，在早期进行重大派息可能是灾难性的，回顾一下我们的起始位置就会向你展示这一点。查理和我当时控制和管理三家公司：伯克希尔·哈撒韦公司、多元零售公司和蓝筹印花公司（现已全部并入我们目前的运营中）。蓝筹只支付少量股息，伯克希尔和多元零售没有支付任何股息。相反，如果这些公司支付了全部收益，我们几乎肯定现在没有任何收益——也许连资本也没有。

这三家公司最初各自从单一业务盈利：（1）伯克希尔的纺织业；（2）多元零售的百货商店；（3）蓝筹的印花。这些基石业务（应特别注意，由你们的主席和副主席精心挑选）分别：（1）幸存但几乎不赚钱，（2）规模萎缩并遭受巨额亏损，（3）销售额缩减至我们进入时的大约5%。（谁说“你不能全输”？）只有通过将可用资金投入到更好的业务中，我们才能克服这些起源。（这就像克服被浪费的青春。）显然，多元化对我们有利。

我们期望继续多元化，同时也支持当前运营的增长，尽管正如我们指出的那样，我们这些努力的回报肯定会低于我们的历史回报。但只要预期回报高于每留存一美元产生一美元市场价值所需的比率，我们将继续留存所有收益。如果我们对未来回报的估计低于该点，我们将分配所有我们认为不能有效利用的非受限收益。在做出这一判断时，我们将审视我们的历史记录和前景。由于我们的逐年结果天生波动，我们认为五年滚动平均值适合判断历史记录。

我们目前的计划是使用我们的留存收益来进一步建立我们保险公司的资本。我们的大多数竞争对手财务状况疲弱，不愿大幅扩张。然而，该行业的大幅保费增长即将来临，1985年可能超过150亿美元，而1983年不到50亿美元。这些情况可能会为我们带来大量有利可图的业务。当然，这一结果并非确定无疑，但其前景比多年来要好得多。

杂项

每年我都在这里刊登我的小型“企业求购”广告。1984年，John Loomis，一位特别有见识且警觉的股东，为我们找到了一家满足我们所有测试的公司。我们立即跟进这个想法，仅仅一个偶然的复杂情况阻止了交易。既然我们的广告在起作用，我们将以去年完全相同的形式重复它：

我们偏好：

1. 大额购买（至少500万美元的税后收益），
2. 经过验证的持续盈利能力（未来预测对我们没什么兴趣，扭亏为盈的情况也不是），
3. 在不使用或很少使用债务的情况下获得良好净资产收益率的企业，
4. 现成管理层（我们无法提供），
5. 简单的业务（如果涉及大量技术，我们无法理解），
6. 报价（如果价格未知，我们不想浪费我们或卖方的时间去谈论交易，即使是初步的）。

我们不会进行敌意收购。我们可以保证完全保密和非常快速的答复——通常在五分钟内——告诉我们是否感兴趣。我们偏好现金购买，但如果我们获得与付出同等内在商业价值，也会考虑发行股票。我们邀请潜在卖家通过联系过去与我们有过业务来往的人来核实我们。对于合适的业务——和合适的人——我们可以提供一个好归宿。

 * * *

创纪录的97.2%的合格股份参与了伯克希尔1984年的股东指定捐款计划。通过该计划的总捐款为3,179,000美元，共有1,519家慈善机构受益。我们的年会委托材料将允许您投票表达您对该计划的看法——您是否认为我们应该继续它，如果应该，以何种每股水平继续。（您可能会感兴趣地了解到，我们未能找到管理层寻求股东对与所有者相关的公司政策意见的咨询投票先例。信任资本主义的管理者似乎并不急于信任资本家。）

我们敦促新股东阅读第60和61页关于我们股东指定捐款计划的描述。如果您希望参与未来的计划，我们强烈建议您立即确保您的股票以实际所有者的名义注册，而不是以“街道”名义或被提名人的名义。未在1985年9月30日按此方式注册的股份将没有资格参与1985年的计划。

 * * *

我们的年度会议将于1985年5月21日在奥马哈举行，我希望您能参加。许多年度会议是浪费时间，对股东和管理层都是如此。有时这是因为管理层不愿意就业务实质问题畅所欲言。更常见的是，非生产性的会议是股东参与者的过错，他们更关心自己登台亮相的时刻，而不是公司的事务。本应是商业讨论的论坛变成了戏剧、发泄情绪和倡导议题的论坛。（这笔交易不可抗拒：只需一股的代价，你就可以让一个被俘虏的听众听你谈论世界应该如何运行的想法。）在这种情况下，会议的质量常常逐年恶化，因为那些只关心自己的人的行为，劝阻了对业务真正感兴趣的人参加。

伯克希尔的会议则是另一回事。参加会议的股东人数每年都在增加，我们还没有遇到过一个愚蠢的问题或出于自我表达而来的评论。相反，我们收到了各种各样关于业务的深思熟虑的问题。因为年度会议是提出这些问题的时间和地点，查理和我乐意回答所有问题，无论需要多长时间。（然而，在一年中的其他时间，我们无法回复书面或电话询问的问题；在一家拥有3000名股东的公司中，逐一汇报是对管理层时间的低效利用。）年度会议上唯一不能谈论的业务事项是那些坦诚可能会给公司带来实际损失的事项。我们的证券活动就是主要例子。

我们总是在这些页面上对我们的股东合伙人的质量略加炫耀。来参加年度会议，你就会明白为什么。外地人应该安排去内布拉斯加家具城参观。如果你买些东西，节省的钱将远远超过支付你的旅行费用，而且你会享受这次体验。

1985年2月25日

沃伦·E·巴菲特 董事会主席

随后事件：3月18日，在本报告文稿交付打字员一周后，但在付印前不久，我们同意以每股172.50美元的价格购买三百万股资本城通信公司股票。我们的购买取决于资本城收购美国广播公司，并将在该交易完成时交割。最早也要到1985年底。我们对由汤姆·墨菲和丹·伯克领导的资本城管理层的钦佩，已在之前的年度报告中多次表达。简而言之，他们在能力和诚信方面都是顶尖的。我们将在明年的报告中对此投资进行更多说明。