BERKSHIRE HATHAWAY INC.
To the Shareholders of Berkshire Hathaway Inc.:
This past year our registered shareholders increased from
about 1900 to about 2900. Most of this growth resulted from our
merger with Blue Chip Stamps, but there also was an acceleration
in the pace of “natural” increase that has raised us from the
1000 level a few years ago.
With so many new shareholders, it’s appropriate to summarize
the major business principles we follow that pertain to the
manager-owner relationship:
o Although our form is corporate, our attitude is
partnership. Charlie Munger and I think of our shareholders as
owner-partners, and of ourselves as managing partners. (Because
of the size of our shareholdings we also are, for better or
worse, controlling partners.) We do not view the company itself
as the ultimate owner of our business assets but, instead, view
the company as a conduit through which our shareholders own the
assets.
o In line with this owner-orientation, our directors are all
major shareholders of Berkshire Hathaway. In the case of at
least four of the five, over 50% of family net worth is
represented by holdings of Berkshire. We eat our own cooking.
o Our long-term economic goal (subject to some qualifications
mentioned later) is to maximize the average annual rate of gain
in intrinsic business value on a per-share basis. We do not
measure the economic significance or performance of Berkshire by
its size; we measure by per-share progress. We are certain that
the rate of per-share progress will diminish in the future - a
greatly enlarged capital base will see to that. But we will be
disappointed if our rate does not exceed that of the average
large American corporation.
o Our preference would be to reach this goal by directly
owning a diversified group of businesses that generate cash and
consistently earn above-average returns on capital. Our second
choice is to own parts of similar businesses, attained primarily
through purchases of marketable common stocks by our insurance
subsidiaries. The price and availability of businesses and the
need for insurance capital determine any given year’s capital
allocation.
o Because of this two-pronged approach to business ownership
and because of the limitations of conventional accounting,
consolidated reported earnings may reveal relatively little about
our true economic performance. Charlie and I, both as owners and
managers, virtually ignore such consolidated numbers. However,
we will also report to you the earnings of each major business we
control, numbers we consider of great importance. These figures,
along with other information we will supply about the individual
businesses, should generally aid you in making judgments about
them.
o Accounting consequences do not influence our operating or
capital-allocation decisions. When acquisition costs are
similar, we much prefer to purchase $2 of earnings that is not
reportable by us under standard accounting principles than to
purchase $1 of earnings that is reportable. This is precisely
the choice that often faces us since entire businesses (whose
earnings will be fully reportable) frequently sell for double the
pro-rata price of small portions (whose earnings will be largely
unreportable). In aggregate and over time, we expect the
unreported earnings to be fully reflected in our intrinsic
business value through capital gains.
o We rarely use much debt and, when we do, we attempt to
structure it on a long-term fixed rate basis. We will reject
interesting opportunities rather than over-leverage our balance
sheet. This conservatism has penalized our results but it is the
only behavior that leaves us comfortable, considering our
fiduciary obligations to policyholders, depositors, lenders and
the many equity holders who have committed unusually large
portions of their net worth to our care.
o A managerial “wish list” will not be filled at shareholder
expense. We will not diversify by purchasing entire businesses
at control prices that ignore long-term economic consequences to
our shareholders. We will only do with your money what we would
do with our own, weighing fully the values you can obtain by
diversifying your own portfolios through direct purchases in the
stock market.
o We feel noble intentions should be checked periodically
against results. We test the wisdom of retaining earnings by
assessing whether retention, over time, delivers shareholders at
least $1 of market value for each $1 retained. To date, this
test has been met. We will continue to apply it on a five-year
rolling basis. As our net worth grows, it is more difficult to
use retained earnings wisely.
o We will issue common stock only when we receive as much in
business value as we give. This rule applies to all forms of
issuance - not only mergers or public stock offerings, but stock
for-debt swaps, stock options, and convertible securities as
well. We will not sell small portions of your company - and that
is what the issuance of shares amounts to - on a basis
inconsistent with the value of the entire enterprise.
o You should be fully aware of one attitude Charlie and I
share that hurts our financial performance: regardless of price,
we have no interest at all in selling any good businesses that
Berkshire owns, and are very reluctant to sell sub-par businesses
as long as we expect them to generate at least some cash and as
long as we feel good about their managers and labor relations.
We hope not to repeat the capital-allocation mistakes that led us
into such sub-par businesses. And we react with great caution to
suggestions that our poor businesses can be restored to
satisfactory profitability by major capital expenditures. (The
projections will be dazzling - the advocates will be sincere -
but, in the end, major additional investment in a terrible
industry usually is about as rewarding as struggling in
quicksand.) Nevertheless, gin rummy managerial behavior (discard
your least promising business at each turn) is not our style. We
would rather have our overall results penalized a bit than engage
in it.
o We will be candid in our reporting to you, emphasizing the
pluses and minuses important in appraising business value. Our
guideline is to tell you the business facts that we would want to
know if our positions were reversed. We owe you no less.
Moreover, as a company with a major communications business, it
would be inexcusable for us to apply lesser standards of
accuracy, balance and incisiveness when reporting on ourselves
than we would expect our news people to apply when reporting on
others. We also believe candor benefits us as managers: the CEO
who misleads others in public may eventually mislead himself in
private.
o Despite our policy of candor, we will discuss our
activities in marketable securities only to the extent legally
required. Good investment ideas are rare, valuable and subject
to competitive appropriation just as good product or business
acquisition ideas are. Therefore, we normally will not talk
about our investment ideas. This ban extends even to securities
we have sold (because we may purchase them again) and to stocks
we are incorrectly rumored to be buying. If we deny those
reports but say “no comment” on other occasions, the no-comments
become confirmation.
That completes the catechism, and we can now move on to the
high point of 1983 - the acquisition of a majority interest in
Nebraska Furniture Mart and our association with Rose Blumkin and
her family.
Nebraska Furniture Mart
Last year, in discussing how managers with bright, but
adrenalin-soaked minds scramble after foolish acquisitions, I
quoted Pascal: “It has struck me that all the misfortunes of men
spring from the single cause that they are unable to stay quietly
in one room.”
Even Pascal would have left the room for Mrs. Blumkin.
About 67 years ago Mrs. Blumkin, then 23, talked her way
past a border guard to leave Russia for America. She had no
formal education, not even at the grammar school level, and knew
no English. After some years in this country, she learned the
language when her older daughter taught her, every evening, the
words she had learned in school during the day.
In 1937, after many years of selling used clothing, Mrs.
Blumkin had saved $500 with which to realize her dream of opening
a furniture store. Upon seeing the American Furniture Mart in
Chicago - then the center of the nation’s wholesale furniture
activity - she decided to christen her dream Nebraska Furniture
Mart.
She met every obstacle you would expect (and a few you
wouldn’t) when a business endowed with only $500 and no
locational or product advantage goes up against rich, long-
entrenched competition. At one early point, when her tiny
resources ran out, “Mrs. B” (a personal trademark now as well
recognized in Greater Omaha as Coca-Cola or Sanka) coped in a way
not taught at business schools: she simply sold the furniture and
appliances from her home in order to pay creditors precisely as
promised.
Omaha retailers began to recognize that Mrs. B would offer
customers far better deals than they had been giving, and they
pressured furniture and carpet manufacturers not to sell to her.
But by various strategies she obtained merchandise and cut prices
sharply. Mrs. B was then hauled into court for violation of Fair
Trade laws. She not only won all the cases, but received
invaluable publicity. At the end of one case, after
demonstrating to the court that she could profitably sell carpet
at a huge discount from the prevailing price, she sold the judge
$1400 worth of carpet.
Today Nebraska Furniture Mart generates over $100 million of
sales annually out of one 200,000 square-foot store. No other
home furnishings store in the country comes close to that volume.
That single store also sells more furniture, carpets, and
appliances than do all Omaha competitors combined.
One question I always ask myself in appraising a business is
how I would like, assuming I had ample capital and skilled
personnel, to compete with it. I’d rather wrestle grizzlies than
compete with Mrs. B and her progeny. They buy brilliantly, they
operate at expense ratios competitors don’t even dream about, and
they then pass on to their customers much of the savings. It’s
the ideal business - one built upon exceptional value to the
customer that in turn translates into exceptional economics for
its owners.
Mrs. B is wise as well as smart and, for far-sighted family
reasons, was willing to sell the business last year. I had
admired both the family and the business for decades, and a deal
was quickly made. But Mrs. B, now 90, is not one to go home and
risk, as she puts it, “losing her marbles”. She remains Chairman
and is on the sales floor seven days a week. Carpet sales are
her specialty. She personally sells quantities that would be a
good departmental total for other carpet retailers.
We purchased 90% of the business - leaving 10% with members
of the family who are involved in management - and have optioned
10% to certain key young family managers.
And what managers they are. Geneticists should do
handsprings over the Blumkin family. Louie Blumkin, Mrs. B’s
son, has been President of Nebraska Furniture Mart for many years
and is widely regarded as the shrewdest buyer of furniture and
appliances in the country. Louie says he had the best teacher,
and Mrs. B says she had the best student. They’re both right.
Louie and his three sons all have the Blumkin business ability,
work ethic, and, most important, character. On top of that, they
are really nice people. We are delighted to be in partnership
with them.
Corporate Performance
During 1983 our book value increased from $737.43 per share
to $975.83 per share, or by 32%. We never take the one-year
figure very seriously. After all, why should the time required
for a planet to circle the sun synchronize precisely with the
time required for business actions to pay off? Instead, we
recommend not less than a five-year test as a rough yardstick of
economic performance. Red lights should start flashing if the
five-year average annual gain falls much below the return on
equity earned over the period by American industry in aggregate.
(Watch out for our explanation if that occurs as Goethe observed,
“When ideas fail, words come in very handy.”)
During the 19-year tenure of present management, book value
has grown from $19.46 per share to $975.83, or 22.6% compounded
annually. Considering our present size, nothing close to this
rate of return can be sustained. Those who believe otherwise
should pursue a career in sales, but avoid one in mathematics.
We report our progress in terms of book value because in our
case (though not, by any means, in all cases) it is a
conservative but reasonably adequate proxy for growth in
intrinsic business value - the measurement that really counts.
Book value’s virtue as a score-keeping measure is that it is easy
to calculate and doesn’t involve the subjective (but important)
judgments employed in calculation of intrinsic business value.
It is important to understand, however, that the two terms - book
value and intrinsic business value - have very different
meanings.
Book value is an accounting concept, recording the
accumulated financial input from both contributed capital and
retained earnings. Intrinsic business value is an economic
concept, estimating future cash output discounted to present
value. Book value tells you what has been put in; intrinsic
business value estimates what can be taken out.
An analogy will suggest the difference. Assume you spend
identical amounts putting each of two children through college.
The book value (measured by financial input) of each child’s
education would be the same. But the present value of the future
payoff (the intrinsic business value) might vary enormously -
from zero to many times the cost of the education. So, also, do
businesses having equal financial input end up with wide
variations in value.
At Berkshire, at the beginning of fiscal 1965 when the
present management took over, the $19.46 per share book value
considerably overstated intrinsic business value. All of that
book value consisted of textile assets that could not earn, on
average, anything close to an appropriate rate of return. In the
terms of our analogy, the investment in textile assets resembled
investment in a largely-wasted education.
Now, however, our intrinsic business value considerably
exceeds book value. There are two major reasons:
(1) Standard accounting principles require that common
stocks held by our insurance subsidiaries be stated on
our books at market value, but that other stocks we own
be carried at the lower of aggregate cost or market.
At the end of 1983, the market value of this latter
group exceeded carrying value by $70 million pre-tax,
or about $50 million after tax. This excess belongs in
our intrinsic business value, but is not included in
the calculation of book value;
(2) More important, we own several businesses that possess
economic Goodwill (which is properly includable in
intrinsic business value) far larger than the
accounting Goodwill that is carried on our balance
sheet and reflected in book value.
Goodwill, both economic and accounting, is an arcane subject
and requires more explanation than is appropriate here. The
appendix that follows this letter - “Goodwill and its
Amortization: The Rules and The Realities” - explains why
economic and accounting Goodwill can, and usually do, differ
enormously.
You can live a full and rewarding life without ever thinking
about Goodwill and its amortization. But students of investment
and management should understand the nuances of the subject. My
own thinking has changed drastically from 35 years ago when I was
taught to favor tangible assets and to shun businesses whose
value depended largely upon economic Goodwill. This bias caused
me to make many important business mistakes of omission, although
relatively few of commission.
Keynes identified my problem: “The difficulty lies not in
the new ideas but in escaping from the old ones.” My escape was
long delayed, in part because most of what I had been taught by
the same teacher had been (and continues to be) so
extraordinarily valuable. Ultimately, business experience,
direct and vicarious, produced my present strong preference for
businesses that possess large amounts of enduring Goodwill and
that utilize a minimum of tangible assets.
I recommend the Appendix to those who are comfortable with
accounting terminology and who have an interest in understanding
the business aspects of Goodwill. Whether or not you wish to
tackle the Appendix, you should be aware that Charlie and I
believe that Berkshire possesses very significant economic
Goodwill value above that reflected in our book value.
Sources of Reported Earnings
The table below shows the sources of Berkshire’s reported
earnings. In 1982, Berkshire owned about 60% of Blue Chip Stamps
whereas, in 1983, our ownership was 60% throughout the first six
months and 100% thereafter. In turn, Berkshire’s net interest in
Wesco was 48% during 1982 and the first six months of 1983, and
80% for the balance of 1983. Because of these changed ownership
percentages, the first two columns of the table provide the best
measure of underlying business performance.
All of the significant gains and losses attributable to
unusual sales of assets by any of the business entities are
aggregated with securities transactions on the line near the
bottom of the table, and are not included in operating earnings.
(We regard any annual figure for realized capital gains or losses
as meaningless, but we regard the aggregate realized and
unrealized capital gains over a period of years as very
important.) Furthermore, amortization of Goodwill is not charged
against the specific businesses but, for reasons outlined in the
Appendix, is set forth as a separate item.
Net Earnings
Earnings Before Income Taxes After Tax
-------------------------------------- ------------------
Total Berkshire Share Berkshire Share
------------------ ------------------ ------------------
1983 1982 1983 1982 1983 1982
-------- -------- -------- -------- -------- --------
(000s omitted)
Operating Earnings:
Insurance Group:
Underwriting ............ $(33,872) $(21,558) $(33,872) $(21,558) $(18,400) $(11,345)
Net Investment Income ... 43,810 41,620 43,810 41,620 39,114 35,270
Berkshire-Waumbec Textiles (100) (1,545) (100) (1,545) (63) (862)
Associated Retail Stores .. 697 914 697 914 355 446
Nebraska Furniture Mart(1) 3,812 -- 3,049 -- 1,521 --
See’s Candies ............. 27,411 23,884 24,526 14,235 12,212 6,914
Buffalo Evening News ...... 19,352 (1,215) 16,547 (724) 8,832 (226)
Blue Chip Stamps(2) ....... (1,422) 4,182 (1,876) 2,492 (353) 2,472
Wesco Financial - Parent .. 7,493 6,156 4,844 2,937 3,448 2,210
Mutual Savings and Loan ... (798) (6) (467) (2) 1,917 1,524
Precision Steel ........... 3,241 1,035 2,102 493 1,136 265
Interest on Debt .......... (15,104) (14,996) (13,844) (12,977) (7,346) (6,951)
Special GEICO Distribution 21,000 -- 21,000 -- 19,551 --
Shareholder-Designated
Contributions .......... (3,066) (891) (3,066) (891) (1,656) (481)
Amortization of Goodwill .. (532) 151 (563) 90 (563) 90
Other ..................... 10,121 3,371 9,623 2,658 8,490 2,171
-------- -------- -------- -------- -------- --------
Operating Earnings .......... 82,043 41,102 72,410 27,742 68,195 31,497
Sales of securities and
unusual sales of assets .. 67,260 36,651 65,089 21,875 45,298 14,877
-------- -------- -------- -------- -------- --------
Total Earnings .............. $149,303 $ 77,753 $137,499 $ 49,617 $113,493 $ 46,374
======== ======== ======== ======== ======== ========
(1) October through December
(2) 1982 and 1983 are not comparable; major assets were
transferred in the merger.
For a discussion of the businesses owned by Wesco, please
read Charlie Munger’s report on pages 46-51. Charlie replaced
Louie Vincenti as Chairman of Wesco late in 1983 when health
forced Louie’s retirement at age 77. In some instances, “health”
is a euphemism, but in Louie’s case nothing but health would
cause us to consider his retirement. Louie is a marvelous man
and has been a marvelous manager.
The special GEICO distribution reported in the table arose
when that company made a tender offer for a portion of its stock,
buying both from us and other shareholders. At GEICO’s request,
we tendered a quantity of shares that kept our ownership
percentage the same after the transaction as before. The
proportional nature of our sale permitted us to treat the
proceeds as a dividend. Unlike individuals, corporations net
considerably more when earnings are derived from dividends rather
than from capital gains, since the effective Federal income tax
rate on dividends is 6.9% versus 28% on capital gains.
Even with this special item added in, our total dividends
from GEICO in 1983 were considerably less than our share of
GEICO’s earnings. Thus it is perfectly appropriate, from both an
accounting and economic standpoint, to include the redemption
proceeds in our reported earnings. It is because the item is
large and unusual that we call your attention to it.
The table showing you our sources of earnings includes
dividends from those non-controlled companies whose marketable
equity securities we own. But the table does not include
earnings those companies have retained that are applicable to our
ownership. In aggregate and over time we expect those
undistributed earnings to be reflected in market prices and to
increase our intrinsic business value on a dollar-for-dollar
basis, just as if those earnings had been under our control and
reported as part of our profits. That does not mean we expect
all of our holdings to behave uniformly; some will disappoint us,
others will deliver pleasant surprises. To date our experience
has been better than we originally anticipated, In aggregate, we
have received far more than a dollar of market value gain for
every dollar of earnings retained.
The following table shows our 1983 yearend net holdings in
marketable equities. All numbers represent 100% of Berkshire’s
holdings, and 80% of Wesco’s holdings. The portion attributable
to minority shareholders of Wesco has been excluded.
No. of Shares Cost Market
------------- ---------- ----------
(000s omitted)
690,975 Affiliated Publications, Inc. .... $ 3,516 $ 26,603
4,451,544 General Foods Corporation(a) ..... 163,786 228,698
6,850,000 GEICO Corporation ................ 47,138 398,156
2,379,200 Handy & Harman ................... 27,318 42,231
636,310 Interpublic Group of Companies, Inc. 4,056 33,088
197,200 Media General .................... 3,191 11,191
250,400 Ogilvy & Mather International .... 2,580 12,833
5,618,661 R. J. Reynolds Industries, Inc.(a) 268,918 341,334
901,788 Time, Inc. ....................... 27,732 56,860
1,868,600 The Washington Post Company ...... 10,628 136,875
---------- ----------
$558,863 $1,287,869
All Other Common Stockholdings ... 7,485 18,044
---------- ----------
Total Common Stocks .............. $566,348 $1,305,913
========== ==========
(a) WESCO owns shares in these companies.
Based upon present holdings and present dividend rates -
excluding any special items such as the GEICO proportional
redemption last year - we would expect reported dividends from
this group to be approximately $39 million in 1984. We can also
make a very rough guess about the earnings this group will retain
that will be attributable to our ownership: these may total about
$65 million for the year. These retained earnings could well
have no immediate effect on market prices of the securities.
Over time, however, we feel they will have real meaning.
In addition to the figures already supplied, information
regarding the businesses we control appears in Management’s
Discussion on pages 40-44. The most significant of these are
Buffalo Evening News, See’s, and the Insurance Group, to which we
will give some special attention here.
Buffalo Evening News
First, a clarification: our corporate name is Buffalo
Evening News, Inc. but the name of the newspaper, since we began
a morning edition a little over a year ago, is Buffalo News.
In 1983 the News somewhat exceeded its targeted profit
margin of 10% after tax. Two factors were responsible: (1) a
state income tax cost that was subnormal because of a large loss
carry-forward, now fully utilized, and (2) a large drop in the
per-ton cost of newsprint (an unanticipated fluke that will be
reversed in 1984).
Although our profit margins in 1983 were about average for
newspapers such as the News, the paper’s performance,
nevertheless, was a significant achievement considering the
economic and retailing environment in Buffalo.
Buffalo has a concentration of heavy industry, a segment of
the economy that was hit particularly hard by the recent
recession and that has lagged the recovery. As Buffalo consumers
have suffered, so also have the paper’s retailing customers.
Their numbers have shrunk over the past few years and many of
those surviving have cut their linage.
Within this environment the News has one exceptional
strength: its acceptance by the public, a matter measured by the
paper’s “penetration ratio” - the percentage of households within
the community purchasing the paper each day. Our ratio is
superb: for the six months ended September 30, 1983 the News
stood number one in weekday penetration among the 100 largest
papers in the United States (the ranking is based on “city zone”
numbers compiled by the Audit Bureau of Circulations).
In interpreting the standings, it is important to note that
many large cities have two papers, and that in such cases the
penetration of either paper is necessarily lower than if there
were a single paper, as in Buffalo. Nevertheless, the list of
the 100 largest papers includes many that have a city to
themselves. Among these, the News is at the top nationally, far
ahead of many of the country’s best-known dailies.
Among Sunday editions of these same large dailies, the News
ranks number three in penetration - ten to twenty percentage
points ahead of many well-known papers. It was not always this
way in Buffalo. Below we show Sunday circulation in Buffalo in
the years prior to 1977 compared with the present period. In
that earlier period the Sunday paper was the Courier-Express (the
News was not then publishing a Sunday paper). Now, of course, it
is the News.
Average Sunday Circulation
--------------------------
Year Circulation
---- -----------
1970 314,000
1971 306,000
1972 302,000
1973 290,000
1974 278,000
1975 269,000
1976 270,000
1984 (Current) 376,000
We believe a paper’s penetration ratio to be the best
measure of the strength of its franchise. Papers with unusually
high penetration in the geographical area that is of prime
interest to major local retailers, and with relatively little
circulation elsewhere, are exceptionally efficient buys for those
retailers. Low-penetration papers have a far less compelling
message to present to advertisers.
In our opinion, three factors largely account for the
unusual acceptance of the News in the community. Among these,
points 2 and 3 also may explain the popularity of the Sunday News
compared to that of the Sunday Courier-Express when it was the
sole Sunday paper:
(1) The first point has nothing to do with merits of the
News. Both emigration and immigration are relatively
low in Buffalo. A stable population is more interested
and involved in the activities of its community than is
a shifting population - and, as a result, is more
interested in the content of the local daily paper.
Increase the movement in and out of a city and
penetration ratios will fall.
(2) The News has a reputation for editorial quality and
integrity that was honed by our longtime editor, the
legendary Alfred Kirchhofer, and that has been preserved
and extended by Murray Light. This reputation was
enormously important to our success in establishing a
Sunday paper against entrenched competition. And without
a Sunday edition, the News would not have survived in the
long run.
(3) The News lives up to its name - it delivers a very
unusual amount of news. During 1983, our “news hole”
(editorial material - not ads) amounted to 50% of the
newspaper’s content (excluding preprinted inserts).
Among papers that dominate their markets and that are of
comparable or larger size, we know of only one whose news
hole percentage exceeds that of the News. Comprehensive
figures are not available, but a sampling indicates an
average percentage in the high 30s. In other words, page
for page, our mix gives readers over 25% more news than
the typical paper. This news-rich mixture is by intent.
Some publishers, pushing for higher profit margins, have
cut their news holes during the past decade. We have
maintained ours and will continue to do so. Properly
written and edited, a full serving of news makes our
paper more valuable to the reader and contributes to our
unusual penetration ratio.
Despite the strength of the News’ franchise, gains in ROP
linage (advertising printed within the newspaper pages as
contrasted to preprinted inserts) are going to be very difficult
to achieve. We had an enormous gain in preprints during 1983:
lines rose from 9.3 million to 16.4 million, revenues from $3.6
million to $8.1 million. These gains are consistent with
national trends, but exaggerated in our case by business we
picked up when the Courier-Express closed.
On balance, the shift from ROP to preprints has negative
economic implications for us. Profitability on preprints is less
and the business is more subject to competition from alternative
means of delivery. Furthermore, a reduction in ROP linage means
less absolute space devoted to news (since the news hole
percentage remains constant), thereby reducing the utility of the
paper to the reader.
Stan Lipsey became Publisher of the Buffalo News at midyear
upon the retirement of Henry Urban. Henry never flinched during
the dark days of litigation and losses following our introduction
of the Sunday paper - an introduction whose wisdom was questioned
by many in the newspaper business, including some within our own
building. Henry is admired by the Buffalo business community,
he’s admired by all who worked for him, and he is admired by
Charlie and me. Stan worked with Henry for several years, and
has worked for Berkshire Hathaway since 1969. He has been
personally involved in all nuts-and-bolts aspects of the
newspaper business from editorial to circulation. We couldn’t do
better.
See’s Candy Shops
The financial results at See’s continue to be exceptional.
The business possesses a valuable and solid consumer franchise
and a manager equally valuable and solid.
In recent years See’s has encountered two important
problems, at least one of which is well on its way toward
solution. That problem concerns costs, except those for raw
materials. We have enjoyed a break on raw material costs in
recent years though so, of course, have our competitors. One of
these days we will get a nasty surprise in the opposite
direction. In effect, raw material costs are largely beyond our
control since we will, as a matter of course, buy the finest
ingredients that we can, regardless of changes in their price
levels. We regard product quality as sacred.
But other kinds of costs are more controllable, and it is in
this area that we have had problems. On a per-pound basis, our
costs (not including those for raw materials) have increased in
the last few years at a rate significantly greater than the
increase in the general price level. It is vital to our
competitive position and profit potential that we reverse this
trend.
In recent months much better control over costs has been
attained and we feel certain that our rate of growth in these
costs in 1984 will be below the rate of inflation. This
confidence arises out of our long experience with the managerial
talents of Chuck Huggins. We put Chuck in charge the day we took
over, and his record has been simply extraordinary, as shown by
the following table:
52-53 Week Year Operating Number of Number of
Ended About Sales Profits Pounds of Stores Open
December 31 Revenues After Taxes Candy Sold at Year End
------------------- ------------ ----------- ---------- -----------
1983 (53 weeks) ... $133,531,000 $13,699,000 24,651,000 207
1982 .............. 123,662,000 11,875,000 24,216,000 202
1981 .............. 112,578,000 10,779,000 24,052,000 199
1980 .............. 97,715,000 7,547,000 24,065,000 191
1979 .............. 87,314,000 6,330,000 23,985,000 188
1978 .............. 73,653,000 6,178,000 22,407,000 182
1977 .............. 62,886,000 6,154,000 20,921,000 179
1976 (53 weeks) ... 56,333,000 5,569,000 20,553,000 173
1975 .............. 50,492,000 5,132,000 19,134,000 172
1974 .............. 41,248,000 3,021,000 17,883,000 170
1973 .............. 35,050,000 1,940,000 17,813,000 169
1972 .............. 31,337,000 2,083,000 16,954,000 167
The other problem we face, as the table suggests, is our
recent inability to achieve meaningful gains in pounds sold. The
industry has the same problem. But for many years we
outperformed the industry in this respect and now we are not.
The poundage volume in our retail stores has been virtually
unchanged each year for the past four, despite small increases
every year in the number of shops (and in distribution expense as
well). Of course, dollar volume has increased because we have
raised prices significantly. But we regard the most important
measure of retail trends to be units sold per store rather than
dollar volume. On a same-store basis (counting only shops open
throughout both years) with all figures adjusted to a 52-week
year, poundage was down .8 of 1% during 1983. This small decline
was our best same-store performance since 1979; the cumulative
decline since then has been about 8%. Quantity-order volume,
about 25% of our total, has plateaued in recent years following
very large poundage gains throughout the 1970s.
We are not sure to what extent this flat volume - both in
the retail shop area and the quantity order area - is due to our
pricing policies and to what extent it is due to static industry
volume, the recession, and the extraordinary share of market we
already enjoy in our primary marketing area. Our price increase
for 1984 is much more modest than has been the case in the past
few years, and we hope that next year we can report better volume
figures to you. But we have no basis to forecast these.
Despite the volume problem, See’s strengths are many and
important. In our primary marketing area, the West, our candy is
preferred by an enormous margin to that of any competitor. In
fact, we believe most lovers of chocolate prefer it to candy
costing two or three times as much. (In candy, as in stocks,
price and value can differ; price is what you give, value is what
you get.) The quality of customer service in our shops - operated
throughout the country by us and not by franchisees is every bit
as good as the product. Cheerful, helpful personnel are as much
a trademark of See’s as is the logo on the box. That’s no small
achievement in a business that requires us to hire about 2000
seasonal workers. We know of no comparably-sized organization
that betters the quality of customer service delivered by Chuck
Huggins and his associates.
Because we have raised prices so modestly in 1984, we expect
See’s profits this year to be about the same as in 1983.
Insurance - Controlled Operations
We both operate insurance companies and have a large
economic interest in an insurance business we don’t operate,
GEICO. The results for all can be summed up easily: in
aggregate, the companies we operate and whose underwriting
results reflect the consequences of decisions that were my
responsibility a few years ago, had absolutely terrible results.
Fortunately, GEICO, whose policies I do not influence, simply
shot the lights out. The inference you draw from this summary is
the correct one. I made some serious mistakes a few years ago
that came home to roost.
The industry had its worst underwriting year in a long time,
as indicated by the table below:
Yearly Change Combined Ratio
in Premiums after Policy-
Written (%) holder Dividends
------------- ----------------
1972 .................... 10.2 96.2
1973 .................... 8.0 99.2
1974 .................... 6.2 105.4
1975 .................... 11.0 107.9
1976 .................... 21.9 102.4
1977 .................... 19.8 97.2
1978 .................... 12.8 97.5
1979 .................... 10.3 100.6
1980 .................... 6.0 103.1
1981 .................... 3.9 106.0
1982 (Revised) .......... 4.4 109.7
1983 (Estimated) ........ 4.6 111.0
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 the reasons outlined in last year’s report, we expect
the poor industry experience of 1983 to be more or less typical
for a good many years to come. (As Yogi Berra put it: “It will be
deja vu all over again.”) That doesn’t mean we think the figures
won’t bounce around a bit; they are certain to. But we believe
it highly unlikely that the combined ratio during the balance of
the decade will average significantly below the 1981-1983 level.
Based on our expectations regarding inflation - and we are as
pessimistic as ever on that front - industry premium volume must
grow about 10% annually merely to stabilize loss ratios at
present levels.
Our own combined ratio in 1983 was 121. Since Mike Goldberg
recently took over most of the responsibility for the insurance
operation, it would be nice for me if our shortcomings could be
placed at his doorstep rather than mine. But unfortunately, as
we have often pointed out, the insurance business has a long
lead-time. Though business policies may be changed and personnel
improved, a significant period must pass before the effects are
seen. (This characteristic of the business enabled us to make a
great deal of money in GEICO; we could picture what was likely to
happen well before it actually occurred.) So the roots of the
1983 results are operating and personnel decisions made two or
more years back when I had direct managerial responsibility for
the insurance group.
Despite our poor results overall, several of our managers
did truly outstanding jobs. Roland Miller guided the auto and
general liability business of National Indemnity Company and
National Fire and Marine Insurance Company to improved results,
while those of competitors deteriorated. In addition, Tom Rowley
at Continental Divide Insurance - our fledgling Colorado
homestate company - seems certain to be a winner. Mike found him
a little over a year ago, and he was an important acquisition.
We have become active recently - and hope to become much
more active - in reinsurance transactions where the buyer’s
overriding concern should be the seller’s long-term
creditworthiness. In such transactions our premier financial
strength should make us the number one choice of both claimants
and insurers who must rely on the reinsurer’s promises for a
great many years to come.
A major source of such business is structured settlements -
a procedure for settling losses under which claimants receive
periodic payments (almost always monthly, for life) rather than a
single lump sum settlement. This form of settlement has
important tax advantages for the claimant and also prevents his
squandering a large lump-sum payment. Frequently, some inflation
protection is built into the settlement. Usually the claimant
has been seriously injured, and thus the periodic payments must
be unquestionably secure for decades to come. We believe we
offer unparalleled security. No other insurer we know of - even
those with much larger gross assets - has our financial strength.
We also think our financial strength should recommend us to
companies wishing to transfer loss reserves. In such
transactions, other insurance companies pay us lump sums to
assume all (or a specified portion of) future loss payments
applicable to large blocks of expired business. Here also, the
company transferring such claims needs to be certain of the
transferee’s financial strength for many years to come. Again,
most of our competitors soliciting such business appear to us to
have a financial condition that is materially inferior to ours.
Potentially, structured settlements and the assumption of
loss reserves could become very significant to us. Because of
their potential size and because these operations generate large
amounts of investment income compared to premium volume, we will
show underwriting results from those businesses on a separate
line in our insurance segment data. We also will exclude their
effect in reporting our combined ratio to you. We “front end” no
profit on structured settlement or loss reserve transactions, and
all attributable overhead is expensed currently. Both businesses
are run by Don Wurster at National Indemnity Company.
Insurance - GEICO
Geico’s performance during 1983 was as good as our own
insurance performance was poor. Compared to the industry’s
combined ratio of 111, GEICO wrote at 96 after a large voluntary
accrual for policyholder dividends. A few years ago I would not
have thought GEICO could so greatly outperform the industry. Its
superiority reflects the combination of a truly exceptional
business idea and an exceptional management.
Jack Byrne and Bill Snyder have maintained extraordinary
discipline in the underwriting area (including, crucially,
provision for full and proper loss reserves), and their efforts
are now being further rewarded by significant gains in new
business. Equally important, Lou Simpson is the class of the
field among insurance investment managers. The three of them are
some team.
We have approximately a one-third interest in GEICO. That
gives us a $270 million share in the company’s premium volume, an
amount some 80% larger than our own volume. Thus, the major
portion of our total insurance business comes from the best
insurance book in the country. This fact does not moderate by an
iota the need for us to improve our own operation.
Stock Splits and Stock Activity
We often are asked why Berkshire does not split its stock.
The assumption behind this question usually appears to be that a
split would be a pro-shareholder action. We disagree. Let me
tell you why.
One of our goals is to have Berkshire Hathaway stock sell at
a price rationally related to its intrinsic business value. (But
note “rationally related”, not “identical”: if well-regarded
companies are generally selling in the market at large discounts
from value, Berkshire might well be priced similarly.) The key to
a rational stock price is rational shareholders, both current and
prospective.
If the holders of a company’s stock and/or the prospective
buyers attracted to it are prone to make irrational or emotion-
based decisions, some pretty silly stock prices are going to
appear periodically. Manic-depressive personalities produce
manic-depressive valuations. Such aberrations may help us in
buying and selling the stocks of other companies. But we think
it is in both your interest and ours to minimize their occurrence
in the market for Berkshire.
To obtain only high quality shareholders is no cinch. Mrs.
Astor could select her 400, but anyone can buy any stock.
Entering members of a shareholder “club” cannot be screened for
intellectual capacity, emotional stability, moral sensitivity or
acceptable dress. Shareholder eugenics, therefore, might appear
to be a hopeless undertaking.
In large part, however, we feel that high quality ownership
can be attracted and maintained if we consistently communicate
our business and ownership philosophy - along with no other
conflicting messages - and then let self selection follow its
course. For example, self selection will draw a far different
crowd to a musical event advertised as an opera than one
advertised as a rock concert even though anyone can buy a ticket
to either.
Through our policies and communications - our
“advertisements” - we try to attract investors who will
understand our operations, attitudes and expectations. (And,
fully as important, we try to dissuade those who won’t.) We want
those who think of themselves as business owners and invest in
companies with the intention of staying a long time. And, we
want those who keep their eyes focused on business results, not
market prices.
Investors possessing those characteristics are in a small
minority, but we have an exceptional collection of them. I
believe well over 90% - probably over 95% - of our shares are
held by those who were shareholders of Berkshire or Blue Chip
five years ago. And I would guess that over 95% of our shares
are held by investors for whom the holding is at least double the
size of their next largest. Among companies with at least
several thousand public shareholders and more than $1 billion of
market value, we are almost certainly the leader in the degree to
which our shareholders think and act like owners. Upgrading a
shareholder group that possesses these characteristics is not
easy.
Were we to split the stock or take other actions focusing on
stock price rather than business value, we would attract an
entering class of buyers inferior to the exiting class of
sellers. At $1300, there are very few investors who can’t afford
a Berkshire share. Would a potential one-share purchaser be
better off if we split 100 for 1 so he could buy 100 shares?
Those who think so and who would buy the stock because of the
split or in anticipation of one would definitely downgrade the
quality of our present shareholder group. (Could we really
improve our shareholder group by trading some of our present
clear-thinking members for impressionable new ones who,
preferring paper to value, feel wealthier with nine $10 bills
than with one $100 bill?) People who buy for non-value reasons
are likely to sell for non-value reasons. Their presence in the
picture will accentuate erratic price swings unrelated to
underlying business developments.
We will try to avoid policies that attract buyers with a
short-term focus on our stock price and try to follow policies
that attract informed long-term investors focusing on business
values. just as you purchased your Berkshire shares in a market
populated by rational informed investors, you deserve a chance to
sell - should you ever want to - in the same kind of market. We
will work to keep it in existence.
One of the ironies of the stock market is the emphasis on
activity. Brokers, using terms such as “marketability” and
“liquidity”, sing the praises of companies with high share
turnover (those who cannot fill your pocket will confidently fill
your ear). But investors should understand that what is good for
the croupier is not good for the customer. A hyperactive stock
market is the pickpocket of enterprise.
For example, consider a typical company earning, say, 12% on
equity. Assume a very high turnover rate in its shares of 100%
per year. If a purchase and sale of the stock each extract
commissions of 1% (the rate may be much higher on low-priced
stocks) and if the stock trades at book value, the owners of our
hypothetical company will pay, in aggregate, 2% of the company’s
net worth annually for the privilege of transferring ownership.
This activity does nothing for the earnings of the business, and
means that 1/6 of them are lost to the owners through the
“frictional” cost of transfer. (And this calculation does not
count option trading, which would increase frictional costs still
further.)
All that makes for a rather expensive game of musical
chairs. Can you imagine the agonized cry that would arise if a
governmental unit were to impose a new 16 2/3% tax on earnings of
corporations or investors? By market activity, investors can
impose upon themselves the equivalent of such a tax.
Days when the market trades 100 million shares (and that
kind of volume, when over-the-counter trading is included, is
today abnormally low) are a curse for owners, not a blessing -
for they mean that owners are paying twice as much to change
chairs as they are on a 50-million-share day. If 100 million-
share days persist for a year and the average cost on each
purchase and sale is 15 cents a share, the chair-changing tax for
investors in aggregate would total about $7.5 billion - an amount
roughly equal to the combined 1982 profits of Exxon, General
Motors, Mobil and Texaco, the four largest companies in the
Fortune 500.
These companies had a combined net worth of $75 billion at
yearend 1982 and accounted for over 12% of both net worth and net
income of the entire Fortune 500 list. Under our assumption
investors, in aggregate, every year forfeit all earnings from
this staggering sum of capital merely to satisfy their penchant
for “financial flip-flopping”. In addition, investment
management fees of over $2 billion annually - sums paid for
chair-changing advice - require the forfeiture by investors of
all earnings of the five largest banking organizations (Citicorp,
Bank America, Chase Manhattan, Manufacturers Hanover and J. P.
Morgan). These expensive activities may decide who eats the pie,
but they don’t enlarge it.
(We are aware of the pie-expanding argument that says that
such activities improve the rationality of the capital allocation
process. We think that this argument is specious and that, on
balance, hyperactive equity markets subvert rational capital
allocation and act as pie shrinkers. Adam Smith felt that all
noncollusive acts in a free market were guided by an invisible
hand that led an economy to maximum progress; our view is that
casino-type markets and hair-trigger investment management act as
an invisible foot that trips up and slows down a forward-moving
economy.)
Contrast the hyperactive stock with Berkshire. The bid-and-
ask spread in our stock currently is about 30 points, or a little
over 2%. Depending on the size of the transaction, the
difference between proceeds received by the seller of Berkshire
and cost to the buyer may range downward from 4% (in trading
involving only a few shares) to perhaps 1 1/2% (in large trades
where negotiation can reduce both the market-maker’s spread and
the broker’s commission). Because most Berkshire shares are
traded in fairly large transactions, the spread on all trading
probably does not average more than 2%.
Meanwhile, true turnover in Berkshire stock (excluding
inter-dealer transactions, gifts and bequests) probably runs 3%
per year. Thus our owners, in aggregate, are paying perhaps
6/100 of 1% of Berkshire’s market value annually for transfer
privileges. By this very rough estimate, that’s $900,000 - not a
small cost, but far less than average. Splitting the stock would
increase that cost, downgrade the quality of our shareholder
population, and encourage a market price less consistently
related to intrinsic business value. We see no offsetting
advantages.
Miscellaneous
Last year in this section I ran a small ad to encourage
acquisition candidates. In our communications businesses we tell
our advertisers that repetition is a key to results (which it
is), so we will again repeat our acquisition criteria.
We prefer:
(1) large purchases (at least $5 million of after-tax
earnings),
(2) demonstrated consistent earning power (future
projections are of little interest to us, nor are
“turn-around” situations),
(3) businesses earning good returns on equity while
employing little or no debt,
(4) management in place (we can’t supply it),
(5) simple businesses (if there’s lots of technology, we
won’t understand it),
(6) an offering price (we don’t want to waste our time or
that of the seller by talking, even preliminarily,
about a transaction when price is unknown).
We will not engage in unfriendly takeovers. We can promise
complete confidentiality and a very fast answer - customarily
within five minutes - as to whether we’re interested. We prefer
to buy for cash, but will consider issuance of stock when we
receive as much in intrinsic business value as we give. We
invite potential sellers to check us out by contacting people
with whom we have done business in the past. For the right
business - and the right people - we can provide a good home.
* * * * *
About 96.4% of all eligible shares participated in our 1983
shareholder-designated contributions program. The total
contributions made pursuant to this program - disbursed in the
early days of 1984 but fully expensed in 1983 - were $3,066,501,
and 1353 charities were recipients. Although the response
measured by the percentage of shares participating was
extraordinarily good, the response measured by the percentage of
holders participating was not as good. The reason may well be
the large number of new shareholders acquired through the merger
and their lack of familiarity with the program. We urge new
shareholders to read the description of the program on pages 52-
53.
If you wish to participate in future programs, we strongly
urge that you immediately make sure that your shares are
registered in the actual owner’s name, not in “street” or nominee
name. Shares not so registered on September 28, 1984 will not be
eligible for any 1984 program.
* * * * *
The Blue Chip/Berkshire merger went off without a hitch.
Less than one-tenth of 1% of the shares of each company voted
against the merger, and no requests for appraisal were made. In
1983, we gained some tax efficiency from the merger and we expect
to gain more in the future.
One interesting sidelight to the merger: Berkshire now has
1,146,909 shares outstanding compared to 1,137,778 shares at the
beginning of fiscal 1965, the year present management assumed
responsibility. For every 1% of the company you owned at that
time, you now would own .99%. Thus, all of today’s assets - the
News, See’s, Nebraska Furniture Mart, the Insurance Group, $1.3
billion in marketable stocks, etc. - have been added to the
original textile assets with virtually no net dilution to the
original owners.
We are delighted to have the former Blue Chip shareholders
join us. To aid in your understanding of Berkshire Hathaway, we
will be glad to send you the Compendium of Letters from the
Annual Reports of 1977-1981, and/or the 1982 Annual report.
Direct your request to the Company at 1440 Kiewit Plaza, Omaha,
Nebraska 68131.
Warren E. Buffett
March 14, 1984 Chairman of the Board
Appendix
BERKSHIRE HATHAWAY INC.
Goodwill and its Amortization: The Rules and The Realities
This appendix deals only with economic and accounting Goodwill – not the goodwill of everyday usage. For example, a business may be well liked, even loved, by most of its customers but possess no economic goodwill. (AT&T, before the breakup, was generally well thought of, but possessed not a dime of economic Goodwill.) And, regrettably, a business may be disliked by its customers but possess substantial, and growing, economic Goodwill. So, just for the moment, forget emotions and focus only on economics and accounting.
When a business is purchased, accounting principles require that the purchase price first be assigned to the fair value of the identifiable assets that are acquired. Frequently the sum of the fair values put on the assets (after the deduction of liabilities) is less than the total purchase price of the business. In that case, the difference is assigned to an asset account entitled "excess of cost over equity in net assets acquired". To avoid constant repetition of this mouthful, we will substitute "Goodwill".
Accounting Goodwill arising from businesses purchased before November 1970 has a special standing. Except under rare circumstances, it can remain an asset on the balance sheet as long as the business bought is retained. That means no amortization charges to gradually extinguish that asset need be made against earnings.
The case is different, however, with purchases made from November 1970 on. When these create Goodwill, it must be amortized over not more than 40 years through charges – of equal amount in every year – to the earnings account. Since 40 years is the maximum period allowed, 40 years is what managements (including us) usually elect. This annual charge to earnings is not allowed as a tax deduction and, thus, has an effect on after-tax income that is roughly double that of most other expenses.
That’s how accounting Goodwill works. To see how it differs from economic reality, let’s look at an example close at hand. We’ll round some figures, and greatly oversimplify, to make the example easier to follow. We’ll also mention some implications for investors and managers.
Blue Chip Stamps bought See’s early in 1972 for \$25 million, at which time See’s had about \$8 million of net tangible assets. (Throughout this discussion, accounts receivable will be classified as tangible assets, a definition proper for business analysis.) This level of tangible assets was adequate to conduct the business without use of debt, except for short periods seasonally. See’s was earning about \$2 million after tax at the time, and such earnings seemed conservatively representative of future earning power in constant 1972 dollars.
Thus our first lesson: businesses logically are worth far more than net tangible assets when they can be expected to produce earnings on such assets considerably in excess of market rates of return. The capitalized value of this excess return is economic Goodwill.
In 1972 (and now) relatively few businesses could be expected to consistently earn the 25% after tax on net tangible assets that was earned by See’s – doing it, furthermore, with conservative accounting and no financial leverage. It was not the fair market value of the inventories, receivables or fixed assets that produced the premium rates of return. Rather it was a combination of intangible assets, particularly a pervasive favorable reputation with consumers based upon countless pleasant experiences they have had with both product and personnel.
Such a reputation creates a consumer franchise that allows the value of the product to the purchaser, rather than its production cost, to be the major determinant of selling price. Consumer franchises are a prime source of economic Goodwill. Other sources include governmental franchises not subject to profit regulation, such as television stations, and an enduring position as the low cost producer in an industry.
Let’s return to the accounting in the See’s example. Blue Chip’s purchase of See’s at \$17 million over net tangible assets required that a Goodwill account of this amount be established as an asset on Blue Chip’s books and that \$425,000 be charged to income annually for 40 years to amortize that asset. By 1983, after 11 years of such charges, the \$17 million had been reduced to about \$12.5 million. Berkshire, meanwhile, owned 60% of Blue Chip and, therefore, also 60% of See’s. This ownership meant that Berkshire’s balance sheet reflected 60% of See’s Goodwill, or about \$7.5 million.
In 1983 Berkshire acquired the rest of Blue Chip in a merger that required purchase accounting as contrasted to the "pooling" treatment allowed for some mergers. Under purchase accounting, the "fair value" of the shares we gave to (or "paid") Blue Chip holders had to be spread over the net assets acquired from Blue Chip. This "fair value" was measured, as it almost always is when public companies use their shares to make acquisitions, by the market value of the shares given up.
The assets "purchased" consisted of 40% of everything owned by Blue Chip (as noted, Berkshire already owned the other 60%). What Berkshire "paid" was more than the net identifiable assets we received by \$51.7 million, and was assigned to two pieces of Goodwill: \$28.4 million to See’s and \$23.3 million to Buffalo Evening News.
After the merger, therefore, Berkshire was left with a Goodwill asset for See’s that had two components: the \$7.5 million remaining from the 1971 purchase, and \$28.4 million newly created by the 40% "purchased" in 1983. Our amortization charge now will be about \$1.0 million for the next 28 years, and \$.7 million for the following 12 years, 2002 through 2013.
In other words, different purchase dates and prices have given us vastly different asset values and amortization charges for two pieces of the same asset. (We repeat our usual disclaimer: we have no better accounting system to suggest. The problems to be dealt with are mind boggling and require arbitrary rules.)
But what are the economic realities? One reality is that the amortization charges that have been deducted as costs in the earnings statement each year since acquisition of See’s were not true economic costs. We know that because See’s last year earned \$13 million after taxes on about \$20 million of net tangible assets – a performance indicating the existence of economic Goodwill far larger than the total original cost of our accounting Goodwill. In other words, while accounting Goodwill regularly decreased from the moment of purchase, economic Goodwill increased in irregular but very substantial fashion.
Another reality is that annual amortization charges in the future will not correspond to economic costs. It is possible, of course, that See’s economic Goodwill will disappear. But it won’t shrink in even decrements or anything remotely resembling them. What is more likely is that the Goodwill will increase– in current, if not in constant, dollars – because of inflation.
That probability exists because true economic Goodwill tends to rise in nominal value proportionally with inflation. To illustrate how this works, let’s contrast a See’s kind of business with a more mundane business. When we purchased See’s in 1972, it will be recalled, it was earning about \$2 million on \$8 million of net tangible assets. Let us assume that our hypothetical mundane business then had \$2 million of earnings also, but needed \$18 million in net tangible assets for normal operations. Earning only 11% on required tangible assets, that mundane business would possess little or no economic Goodwill.
A business like that, therefore, might well have sold for the value of its net tangible assets, or for \$18 million. In contrast, we paid \$25 million for See’s, even though it had no more in earnings and less than half as much in "honest-to-God" assets. Could less really have been more, as our purchase price implied? The answer is "yes" – even if both businesses were expected to have flat unit volume– as long as you anticipated, as we did in 1972, a world of continuous inflation.
To understand why, imagine the effect that a doubling of the price level would subsequently have on the two businesses. Both would need to double their nominal earnings to \$4 million to keep themselves even with inflation. This would seem to be no great trick: just sell the same number of units at double earlier prices and, assuming profit margins remain unchanged, profits also must double.
But, crucially, to bring that about, both businesses probably would have to double their nominal investment in net tangible assets, since that is the kind of economic requirement that inflation usually imposes on businesses, both good and bad. A doubling of dollar sales means correspondingly more dollars must be employed immediately in receivables and inventories. Dollars employed in fixed assets will respond more slowly to inflation, but probably just as surely. And all of this inflation-required investment will produce no improvement in rate of return. The motivation for this investment is the survival of the business, not the prosperity of the owner.
Remember, however, that See’s had net tangible assets of only \$8 million. So it would only have had to commit an additional \$8 million to finance the capital needs imposed by inflation. The mundane business, meanwhile, had a burden over twice as large – a need for \$18 million of additional capital.
After the dust had settled, the mundane business, now earning \$4 million annually, might still be worth the value of its tangible assets, or \$36 million. That means its owners would have gained only a dollar of nominal value for every new dollar invested. (This is the same dollar-for-dollar result they would have achieved if they had added money to a savings account.)
See’s, however, also earning \$4 million, might be worth \$50 million if valued (as it logically would be) on the same basis as it was at the time of our purchase. So it would have gained \$25 million in nominal value while the owners were putting up only \$8 million in additional capital – over \$3 of nominal value gained for each \$1 invested.
Remember, even so, that the owners of the See’s kind of business were forced by inflation to ante up \$8 million in additional capital just to stay even in real profits. Any unleveraged business that requires some net tangible assets to operate (and almost all do) is hurt by inflation. Businesses needing little in the way of tangible assets simply are hurt the least.
And that fact, of course, has been hard for many people to grasp. For years the traditional wisdom – long on tradition, short on wisdom – held that inflation protection was best provided by businesses laden with natural resources, plants and machinery, or other tangible assets ("In Goods We Trust"). It doesn’t work that way. Asset-heavy businesses generally earn low rates of return – rates that often barely provide enough capital to fund the inflationary needs of the existing business, with nothing left over for real growth, for distribution to owners, or for acquisition of new businesses.
In contrast, a disproportionate number of the great business fortunes built up during the inflationary years arose from ownership of operations that combined intangibles of lasting value with relatively minor requirements for tangible assets. In such cases earnings have bounded upward in nominal dollars, and these dollars have been largely available for the acquisition of additional businesses. This phenomenon has been particularly evident in the communications business. That business has required little in the way of tangible investment – yet its franchises have endured. During inflation, Goodwill is the gift that keeps giving.
But that statement applies, naturally, only to true economic Goodwill. Spurious accounting Goodwill – and there is plenty of it around – is another matter. When an overexcited management purchases a business at a silly price, the same accounting niceties described earlier are observed. Because it can’t go anywhere else, the silliness ends up in the Goodwill account. Considering the lack of managerial discipline that created the account, under such circumstances it might better be labeled "No-Will". Whatever the term, the 40-year ritual typically is observed and the adrenalin so capitalized remains on the books as an "asset" just as if the acquisition had been a sensible one.
If you cling to any belief that accounting treatment of Goodwill is the best measure of economic reality, I suggest one final item to ponder.
Assume a company with \$20 per share of net worth, all tangible assets. Further assume the company has internally developed some magnificent consumer franchise, or that it was fortunate enough to obtain some important television stations by original FCC grant. Therefore, it earns a great deal on tangible assets, say \$5 per share, or 25%.
With such economics, it might sell for \$100 per share or more, and it might well also bring that price in a negotiated sale of the entire business.
Assume an investor buys the stock at \$100 per share, paying in effect \$80 per share for Goodwill (just as would a corporate purchaser buying the whole company). Should the investor impute a \$2 per share amortization charge annually (\$80 divided by 40 years) to calculate "true" earnings per share? And, if so, should the new "true" earnings of \$3 per share cause him to rethink his purchase price?
We believe managers and investors alike should view intangible assets from two perspectives:
-
In analysis of operating results – that is, in evaluating the underlying economics of a business unit – amortization charges should be ignored. What a business can be expected to earn on unleveraged net tangible assets, excluding any charges against earnings for amortization of Goodwill, is the best guide to the economic attractiveness of the operation. It is also the best guide to the current value of the operation’s economic Goodwill.
- In evaluating the wisdom of business acquisitions, amortization charges should be ignored also. They should be deducted neither from earnings nor from the cost of the business. This means forever viewing purchased Goodwill at its full cost, before any amortization. Furthermore, cost should be defined as including the full intrinsic business value – not just the recorded accounting value – of all consideration given, irrespective of market prices of the securities involved at the time of merger and irrespective of whether pooling treatment was allowed. For example, what we truly paid in the Blue Chip merger for 40% of the Goodwill of See’s and the News was considerably more than the \$51.7 million entered on our books. This disparity exists because the market value of the Berkshire shares given up in the merger was less than their intrinsic business value, which is the value that defines the true cost to us.
Operations that appear to be winners based upon perspective (1) may pale when viewed from perspective (2). A good business is not always a good purchase – although it’s a good place to look for one.
We will try to acquire businesses that have excellent operating economics measured by (1) and that provide reasonable returns measured by (2). Accounting consequences will be totally ignored.
At yearend 1983, net Goodwill on our accounting books totaled \$62 million, consisting of the \$79 million you see stated on the asset side of our balance sheet, and \$17 million of negative Goodwill that is offset against the carrying value of our interest in Mutual Savings and Loan.
We believe net economic Goodwill far exceeds the \$62 million accounting number.
伯克希尔·哈撒韦公司
致伯克希尔·哈撒韦公司股东：
过去一年，我们的登记股东人数从约1900人增加到约2900人。这一增长主要源于我们与蓝筹印花公司的合并，但“自然”增长的速度也在加快，让我们从几年前1000人的水平提升至此。
面对这么多新股东，有必要总结一下我们遵循的、与管理层-所有者关系相关的几条主要商业原则：
o 虽然我们是公司形式，但我们的态度是合伙制。查理·芒格和我把股东视为所有者合伙人，把我们自己视为管理合伙人。（由于我们持股规模较大，不管好坏，我们也是控股合伙人。）我们不把公司本身视为我们商业资产的最终所有者，而是把公司视为一个渠道，股东通过它拥有这些资产。
o 基于这种所有者导向，我们的董事会成员都是伯克希尔·哈撒韦的大股东。至少四位董事中，超过50%的家庭净资产以伯克希尔的持股形式存在。我们吃自己做的饭。
o 我们的长期经济目标（受到后面提到的一些条件的限制）是最大化每股内在商业价值的年增长率。我们不按规模来衡量伯克希尔的经济意义或业绩；我们按每股进展来衡量。我们确信未来每股进展的速度会下降——资本基数的扩大必然如此。但如果我们的增长率不能超过美国大型公司的平均水平，我们会感到失望。
o 我们更倾向于通过直接拥有一系列能够产生现金并持续获得高于平均水平的资本回报的企业来实现这一目标。其次是持有类似企业的部分股权，主要通过我们的保险子公司购买可流通普通股来实现。企业的价格和可得性，以及保险资本的需求，决定了任何一年的资本配置。
o 由于这种双重企业所有权方法，以及传统会计的局限性，合并报告收益可能无法真正揭示我们的经济表现。查理和我，作为所有者和经营者，几乎忽略这些合并数字。然而，我们也会向您报告我们所控制的每个主要企业的收益，这些数字我们认为非常重要。这些数据，连同我们将提供的关于各个企业的其他信息，通常应帮助您做出判断。
o 会计后果不影响我们的经营或资本配置决策。当收购成本相似时，我们更愿意购买2美元的标准会计原则下不可报告的收益，而不是1美元可报告的收益。这正是我们经常面临的选择，因为整个企业（其收益可以完全报告）的售价往往是其一小部分（其收益大部分不可报告）按比例价格的两倍。总的来说，随着时间的推移，我们预计这些未报告的收益将通过资本利得完全反映在我们的内在商业价值中。
o 我们很少使用大量债务，即使使用，我们也试图以长期固定利率为基础安排。我们会拒绝有趣的机会，也不愿过度杠杆化我们的资产负债表。这种保守主义损害了我们的业绩，但考虑到我们对保单持有人、存款人、贷款人以及那些将异常大比例的净资产托付给我们的众多股权人所承担的受托责任，这是唯一让我们感到安心的行为。
o 管理层的“愿望清单”不会以股东为代价来实现。我们不会以控制价格收购整个企业来实现多元化，而忽视对股东长期经济后果的影响。我们只用你的钱做我们用自己的钱会做的事，充分权衡你可以通过直接在股票市场购买来多元化自己投资组合所获得的价值。
o 我们认为，良好的意图应该定期对照结果进行检验。我们通过评估留存收益是否随着时间的推移为股东每保留1美元带来至少1美元的市场价值，来检验留存收益的智慧。迄今为止，这一检验已经通过。我们将继续以五年滚动方式应用它。随着我们净资产的增长，明智地使用留存收益变得更加困难。
o 我们只有在获得与付出相等的商业价值时才会发行普通股。这条规则适用于所有形式的发行——不仅是合并或公开募股，还包括债转股、股票期权和可转换证券。我们不会以与整个企业价值不一致的方式出售你公司的一小部分——而这正是发行股票的本质。
o 你应该充分意识到查理和我共有的一种态度会损害我们的财务表现：无论价格如何，我们完全没有兴趣出售伯克希尔拥有的任何优秀企业，并且非常不愿意出售次等企业，只要我们预期它们至少能产生一些现金，并且我们对其管理层和劳资关系感到满意。我们希望不再重复导致我们陷入这些次等企业的资本配置错误。对于通过重大资本支出可以使我们亏损企业恢复令人满意盈利能力的建议，我们持非常谨慎的态度。（预测会很诱人——倡导者会很真诚——但最终，在一个糟糕行业的重大额外投资，通常就像在流沙中挣扎一样有益。）然而，金拉米式管理风格（每轮丢弃你最无前途的业务）不是我们的风格。我们宁愿整体结果受点损失，也不愿这样做。
o 我们将在报告中坦诚相告，强调评估商业价值的重要利弊。我们的指导原则是：如果角色互换，我们会告诉你我们想知道的商业事实。我们对你至少应做到这一点。此外，作为一家拥有重要传媒业务的公司，如果我们在报道自己时采用的准确、平衡和深刻标准低于我们期望新闻人员报道他人时所应用的标准，那是不可原谅的。我们还相信坦诚对我们管理层也有好处：在公开场合误导他人的CEO最终可能在私下里误导自己。
o 尽管我们有坦诚的政策，但我们只会在法律要求的范围内讨论我们在可流通证券上的活动。好的投资想法是罕见的、有价值的，并且容易受到竞争性的侵占，就像好的产品或商业收购想法一样。因此，我们通常不会谈论我们的投资想法。这一禁令甚至延伸到我们已经卖出的证券（因为我们可能会再次买入）以及被错误传闻我们正在买入的股票。如果我们否认这些传闻，但在其他场合说“无可奉告”，那么“无可奉告”就成了确认。
教义问答到此结束，现在我们可以进入1983年的高潮——收购内布拉斯加家具城多数股权，以及与罗斯·布卢姆金及其家族的结缘。
内布拉斯加家具城
去年，在讨论那些头脑精明但肾上腺素旺盛的管理者如何争先恐后地进行愚蠢收购时，我引用了帕斯卡的话：“我注意到，人类的所有不幸都源于一个原因：他们不能安静地待在一个房间里。”
但即使是帕斯卡，也会为了布卢姆金夫人而离开房间。
大约67年前，23岁的布卢姆金夫人凭口才说服边境警卫，离开俄罗斯前往美国。她没有受过正规教育，甚至连小学都没上过，也不懂英语。在美国生活几年后，她的大女儿每天晚上教她白天在学校学到的单词，她由此学会了英语。
1937年，在卖旧衣服多年之后，布卢姆金夫人攒了500美元，用来实现她开家具店的梦想。看到芝加哥的美国家具城——当时全国家具批发活动的中心——她决定将自己的梦想命名为内布拉斯加家具城。
当一家只有500美元资金、没有地点或产品优势的企业与资金雄厚、根深蒂固的竞争对手抗衡时，她遇到了你能想到的所有障碍（以及一些你意想不到的）。在早期，当她的微薄资源耗尽时，“B夫人”（如今在奥马哈地区就像可口可乐或桑卡一样广为人知的个人商标）用一种商学院里学不到的方式应对：她干脆从家里卖掉家具和家电，以便按承诺偿还债权人。
奥马哈的零售商开始意识到B夫人能为顾客提供比他们以往更好的交易，于是向家具和地毯制造商施压，要求他们不要卖给她。但通过各种策略，她获得了商品并大幅降价。B夫人随后因违反公平贸易法被带上法庭。她不仅赢得了所有案件，还获得了无价的宣传。在一个案件结束时，她向法庭证明自己能够以远低于市场价的折扣盈利出售地毯，然后卖给了法官价值1400美元的地毯。
如今，内布拉斯加家具城在一个20万平方英尺的商店里，年销售额超过1亿美元。全国没有其他家具店接近这个销量。这家单店销售的家具、地毯和家电比奥马哈所有竞争对手的总和还要多。
在评估一项业务时，我经常问自己一个问题：假设我有充足的资本和熟练的人员，我是否愿意与它竞争。我宁愿与灰熊搏斗，也不愿与B夫人及其后代竞争。他们采购精明，运营费用率低得让竞争对手无法想象，然后将大部分节省让利于顾客。这是理想的企业——建立在为顾客提供卓越价值的基础上，而这又转化为其所有者的卓越经济效益。
B夫人既聪明又智慧，出于深远家族原因，她去年愿意出售企业。几十年来我一直钦佩这个家族和这家企业，交易很快达成。但现年90岁的B夫人不是那种回家冒险的人，用她的话说，“失去理智”。她仍是董事长，每周七天都在销售现场。地毯销售是她的专长。她个人销售的数量，足以成为其他地毯零售商的一个优秀部门总额。
我们购买了该业务90%的股份——留下10%给参与管理的家族成员——并将10%的期权授予一些关键的年轻家族管理人员。
他们真是出色的管理者。遗传学家应该为布卢姆金家族翻跟头。B夫人的儿子路易·布卢姆金多年来一直担任内布拉斯加家具城总裁，被广泛认为是全国最精明的家具和家电采购商。路易说他有一个最好的老师，B夫人说她有一个最好的学生。他们都说对了。路易和他的三个儿子都拥有布卢姆金的商业能力、职业道德，最重要的是，品格。最重要的是，他们是非常好的人。我们很高兴与他们成为合伙人。
公司业绩
1983年，我们的每股账面价值从737.43美元增加到975.83美元，增长了32%。我们从不把一年期的数字太当回事。毕竟，为什么行星绕太阳一周所需的时间要与商业行动产生回报所需的时间精确同步呢？相反，我们建议至少用五年作为衡量经济表现的大致标准。如果五年平均年增长率远低于同期美国工业整体的净资产收益率，红灯就应该开始闪烁。（如果出现这种情况，请注意我们的解释，正如歌德所说：“当思想失败时，词语就变得非常方便。”）
在现任管理层任职的19年间，账面价值从每股19.46美元增长到975.83美元，年复合增长率22.6%。考虑到我们目前的规模，不可能保持接近这一回报率的速度。那些不这样认为的人应该从事销售职业，但避免从事数学职业。
我们用账面价值来报告进展，是因为在我们的案例中（尽管绝非所有案例），它是内在商业价值增长的保守但相当充分的替代指标——这才是真正重要的衡量标准。账面价值作为记分卡的一个优点是容易计算，并且不涉及计算内在商业价值时所使用的主观（但重要）判断。然而，理解这两个术语——账面价值和内在商业价值——具有非常不同的含义是很重要的。
账面价值是一个会计概念，记录从实收资本和留存收益累积的财务输入。内在商业价值是一个经济概念，估计未来现金产出并折现到现值。账面价值告诉你投入了什么；内在商业价值估计可以取出什么。
一个类比可以说明这种区别。假设你花同样的钱送两个孩子上大学。每个孩子教育的账面价值（按财务投入衡量）是相同的。但未来回报的现值（内在商业价值）可能差异巨大——从零到教育成本的许多倍。同样，财务投入相同的企业最终的估值也可能天差地别。
在伯克希尔，当现任管理层在1965财政年度初接手时，每股19.46美元的账面价值大大高估了内在商业价值。所有这些账面价值都由纺织资产组成，这些资产平均而言无法赚取接近适当回报率的收益。用我们的类比来说，对纺织资产的投资就像对几乎浪费的教育投资。
然而现在，我们的内在商业价值大大超过了账面价值。有两个主要原因：
(1) 标准会计准则要求我们保险子公司持有的普通股按市值在账面上列示，但我们拥有的其他股票按总成本与市价孰低法列示。截至1983年底，后一组股票的市值超过账面价值7000万美元（税前），约5000万美元（税后）。这一超额部分属于我们的内在商业价值，但未计入账面价值计算；
(2) 更重要的是，我们拥有几家企业，它们拥有的经济商誉（应适当计入内在商业价值）远大于资产负债表上列示并反映在账面价值中的会计商誉。
商誉，无论是经济的还是会计的，都是一个深奥的话题，需要更多解释，这里不便展开。这封信后面的附录——“商誉及其摊销：规则与现实”——解释了为什么经济商誉和会计商誉可以而且通常差异巨大。
你可以一辈子都不考虑商誉及其摊销而过上充实而有意义的生活。但投资和管理的学生应该理解这个主题的细微差别。我自己的想法与35年前相比发生了巨大变化，当时我被教导偏好有形资产，回避价值主要依赖经济商誉的企业。这种偏见导致我犯了许多重要的商业错误——主要是错失良机，尽管相对较少是实际出手的错误。
凯恩斯指出了我的问题：“困难不在于新思想，而在于摆脱旧思想。”我摆脱旧思想被长期拖延，部分原因是我从同一位老师那里学到的大部分东西（一直）非常有价值。最终，直接和间接的商业经验使我产生了目前的强烈偏好：偏好拥有大量持久商誉且使用最少有形资产的企业。
我向那些熟悉会计术语并对理解商誉商业层面感兴趣的人推荐附录。无论你是否愿意阅读附录，你都应该知道，查理和我认为伯克希尔拥有远超账面价值所反映的重大经济商誉价值。
报告收益来源
下表显示了伯克希尔报告收益的来源。1982年，伯克希尔持有蓝筹印花公司约60%的股份，而1983年前六个月为60%，之后为100%。同样，伯克希尔在韦斯科的净权益在1982年和1983年前六个月为48%，1983年剩余时间为80%。由于这些所有权百分比的变化，表格的前两列最能反映潜在业务表现。
任何业务实体因资产异常出售而产生的所有重大收益和损失，均与证券交易汇总在表格底部附近的行中，不包括在营业利润内。（我们认为任何一年的已实现资本利得或损失数字毫无意义，但多年期间的已实现和未实现资本利得总额非常重要。）此外，商誉摊销不从特定业务中扣除，而是按照附录中概述的原因单独列示。
净利润
息前税前利润 税后
-------------------------------------- ------------------
总额 伯克希尔份额 伯克希尔份额
------------------ ------------------ ------------------
1983 1982 1983 1982 1983 1982
-------- -------- -------- -------- -------- --------
（千元省略）
营业利润：
保险集团：
承销.................... $(33,872) $(21,558) $(33,872) $(21,558) $(18,400) $(11,345)
净投资收益.............. 43,810 41,620 43,810 41,620 39,114 35,270
伯克希尔-万贝克纺织....... (100) (1,545) (100) (1,545) (63) (862)
关联零售店............... 697 914 697 914 355 446
内布拉斯加家具城(1)...... 3,812 -- 3,049 -- 1,521 --
喜诗糖果................. 27,411 23,884 24,526 14,235 12,212 6,914
布法罗晚报............... 19,352 (1,215) 16,547 (724) 8,832 (226)
蓝筹印花公司(2).......... (1,422) 4,182 (1,876) 2,492 (353) 2,472
韦斯科金融-母公司......... 7,493 6,156 4,844 2,937 3,448 2,210
互助储蓄与贷款........... (798) (6) (467) (2) 1,917 1,524
精密钢铁................. 3,241 1,035 2,102 493 1,136 265
利息债务................. (15,104) (14,996) (13,844) (12,977) (7,346) (6,951)
特殊GEICO分配............ 21,000 -- 21,000 -- 19,551 --
股东指定捐赠............. (3,066) (891) (3,066) (891) (1,656) (481)
商誉摊销................. (532) 151 (563) 90 (563) 90
其他..................... 10,121 3,371 9,623 2,658 8,490 2,171
-------- -------- -------- -------- -------- --------
营业利润合计............... 82,043 41,102 72,410 27,742 68,195 31,497
证券销售及资产异常销售..... 67,260 36,651 65,089 21,875 45,298 14,877
-------- -------- -------- -------- -------- --------
总利润................... $149,303 $ 77,753 $137,499 $ 49,617 $113,493 $ 46,374
======== ======== ======== ======== ======== ========
(1) 10月至12月
(2) 1982年和1983年不可比；合并中主要资产已转移。
关于韦斯科所拥有业务的讨论，请阅读第46-51页查理·芒格的报告。1983年底，查理接替路易·文森蒂担任韦斯科董事长，当时77岁的路易因健康原因退休。在某些情况下，“健康”是委婉说法，但在路易的案例中，除非健康原因，否则我们不会考虑让他退休。路易是个了不起的人，也是个出色的管理者。
表中报告的特殊GEICO分配源于该公司对其部分股票进行要约收购，从我们和其他股东手中买入。应GEICO的要求，我们投标了一定数量的股份，使交易后我们的持股比例与交易前相同。我们出售的比例性质使我们能够将收益视为股息。与个人不同，公司在股息收益中的净收入比资本利得要多得多，因为股息的有效联邦所得税率为6.9%，而资本利得为28%。
即使加上这一特殊项目，我们1983年从GEICO获得的总股息也远低于我们应占GEICO收益的份额。因此，从会计和经济角度来看，将赎回收益纳入报告收益是完全合适的。正因为这个项目金额大且不寻常，我们才提醒您注意。
显示收益来源的表格包括我们从所持有的非控股公司（其可流通权益证券）获得的股息。但该表格不包括这些公司保留的、属于我们所有权的收益。总的来说，随着时间的推移，我们预计这些未分配收益将通过市场价格反映出来，并一元对一元地增加我们的内在商业价值，就好像这些收益在我们的控制之下并作为我们利润的一部分报告一样。这并不意味着我们期望所有持股表现一致；有些会让我们失望，有些会带来惊喜。迄今为止，我们的经验比最初预期的要好。总体而言，我们每保留1美元的收益，就获得了远超过1美元的市场价值增长。
下表显示了1983年末我们在可流通股票中的净持仓。所有数字代表伯克希尔持仓的100%，以及韦斯科持仓的80%。归属于韦斯科少数股东的部分已被排除。
股数 成本 市值
------------- ---------- ----------
（千元省略）
690,975 Affiliated Publications, Inc. .... $ 3,516 $ 26,603
4,451,544 General Foods Corporation(a) ..... 163,786 228,698
6,850,000 GEICO Corporation ................ 47,138 398,156
2,379,200 Handy & Harman ................... 27,318 42,231
636,310 Interpublic Group of Companies, Inc. 4,056 33,088
197,200 Media General .................... 3,191 11,191
250,400 Ogilvy & Mather International .... 2,580 12,833
5,618,661 R. J. Reynolds Industries, Inc.(a) 268,918 341,334
901,788 Time, Inc. ....................... 27,732 56,860
1,868,600 The Washington Post Company ...... 10,628 136,875
---------- ----------
$558,863 $1,287,869
所有其他普通股持仓................. 7,485 18,044
---------- ----------
普通股合计........................ $566,348 $1,305,913
========== ==========
(a) 韦斯科持有这些公司的股份。
基于现有持股和当前股息率——排除任何特殊项目如去年GEICO的比例赎回——我们预计1984年该组合的报告股息约为3900万美元。我们还可以非常粗略地估计该组合保留的、归属于我们所有权的收益：今年可能总计约6500万美元。这些留存收益可能对证券的市场价格没有即时影响。然而，随着时间的推移，我们觉得它们将有真正意义。
除了已经提供的数据，关于我们控制的业务的信息出现在第40-44页的管理层讨论中。其中最重要的是布法罗晚报、喜诗糖果和保险集团，我们将在此特别关注。
布法罗晚报
首先，澄清一下：我们的公司名称是布法罗晚报公司，但自一年多前我们开始发行晨报以来，报纸名称是布法罗新闻。
1983年，布法罗新闻略微超过了其10%的税后目标利润率。有两个原因：(1)州所得税成本低于正常水平，因为有大额亏损结转，现已完全使用；(2)新闻纸每吨成本大幅下降（一个意想不到的偶然事件，1984年将逆转）。
尽管我们1983年的利润率与布法罗新闻这样的报纸平均水平大致相当，但考虑到布法罗的经济和零售环境，该报的业绩仍然是一项重大成就。
布法罗集中了大量重工业，这一经济领域在最近的衰退中受到的打击尤其严重，并且复苏滞后。随着布法罗消费者遭受损失，该报的零售客户也受到冲击。过去几年它们的数量减少，许多幸存者削减了广告行数。
在这种环境下，布法罗新闻有一个非凡的优势：公众的接受度，这由报纸的“渗透率”——社区内每天购买报纸的家庭百分比——来衡量。我们的比率非常出色：在截至1983年9月30日的六个月中，布法罗新闻在美国100家最大报纸中周一至周五渗透率排名第一（排名基于审计发行量局编制的“城市区域”数据）。
在解读排名时，重要的是要注意到许多大城市有两家报纸，在这种情况下，任何一家的渗透率必然低于像布法罗这样只有一家报纸的情况。然而，100家最大报纸的名单包括许多只有一个城市的报纸。在这些报纸中，布法罗新闻在全国排名第一，远远领先于许多全国最知名的日报。
在相同大报纸的周日版中，布法罗新闻的渗透率排名第三——领先许多知名报纸10到20个百分点。布法罗并非一直如此。下面我们显示布法罗1977年之前几年的周日发行量，与当前时期相比。在早期阶段，周日版是《信使快报》（当时布法罗新闻没有出版周日版）。现在，当然是布法罗新闻。
平均周日发行量
--------------------------
年份 发行量
---- -----------
1970 314,000
1971 306,000
1972 302,000
1973 290,000
1974 278,000
1975 269,000
1976 270,000
1984（当前） 376,000
我们认为报纸的渗透率是衡量其特许经营权实力的最佳指标。在地理区域中对主要本地零售商具有异常高渗透率、在其他地方发行量相对较少的报纸，对这些零售商来说是异常高效的购买选择。低渗透率报纸向广告商传达的信息远没有那么有说服力。
我们认为，三个因素在很大程度上解释了布法罗新闻在社区的非凡接受度。其中，第2点和第3点也可能解释了与《信使快报》作为唯一周日版时相比，布法罗新闻周日版受欢迎的原因：
(1) 第一点与布法罗新闻的优点无关。布法罗的迁出和迁入都相对较低。稳定的人口对社区活动更感兴趣、更投入，而流动人口则不然——因此对本地日报的内容更感兴趣。增加城市人口的进出流动，渗透率就会下降。
(2) 布法罗新闻以编辑质量和诚信著称，这得益于我们长期主编、传奇人物阿尔弗雷德·基尔希霍费尔的锤炼，并由默里·莱特保持和发扬。这一声誉对我们成功建立周日版、对抗根深蒂固的竞争对手至关重要。而没有周日版，布法罗新闻长期无法生存。
(3) 布法罗新闻名副其实——它提供非常丰富的新闻。1983年，我们的“新闻版面”（编辑内容——非广告）占报纸内容的50%（不包括预先印刷的插页）。在主导市场且规模相当或更大的报纸中，我们只知道一家的新闻版面比例超过布法罗新闻。没有全面的数据，但抽样显示平均百分比在30多的水平。换句话说，每页内容，我们的组合给读者提供比典型报纸多25%以上的新闻。这种新闻丰富的组合是有意为之。一些出版商为了提高利润率，在过去十年削减了新闻版面。我们保持了，并将继续这样做。如果撰写和编辑得当，充分的新闻供给使我们的报纸对读者更有价值，并有助于我们非凡的渗透率。
尽管布法罗新闻的特许经营权实力强劲，但ROP广告行数（印刷在报纸页面内的广告，与预印插页相对）的增长将非常困难。1983年，我们的预印插页大幅增长：行数从930万增加到1640万，收入从360万美元增加到810万美元。这些增长与全国趋势一致，但在我们的情况下，由于《信使快报》停刊后我们获得的业务而扩大。
总的来说，从ROP转向预印插页对我们有负面的经济影响。预印插页的盈利能力较低，而且业务更容易受到替代递送方式的竞争。此外，ROP行数的减少意味着新闻的绝对空间减少（因为新闻版面比例保持不变），从而降低了报纸对读者的实用性。
斯坦·利普西在年中接替亨利·厄本担任布法罗新闻的出版商。在引入周日版后那段诉讼和亏损的黑暗日子里，亨利从未退缩——当时引入周日版的明智性受到许多业内人士的质疑，包括我们内部的一些人。亨利受到布法罗商界的钦佩，受到所有为他工作的人的钦佩，也受到查理和我的钦佩。斯坦与亨利共事多年，自1969年以来一直为伯克希尔·哈撒韦工作。他亲身参与了报纸业务的各个具体方面，从编辑到发行。我们不可能做得更好了。
喜诗糖果店
喜诗糖果的财务业绩依然出色。该业务拥有宝贵而稳固的消费者特许经营权，以及同样宝贵和稳固的管理者。
近年来，喜诗面临着两个重要问题，至少其中一个正接近解决。那个问题涉及成本，不包括原材料成本。近年来我们在原材料成本上有所受益，当然我们的竞争对手也是如此。总有一天我们会遇到相反方向的令人不快的意外。实际上，原材料成本基本上超出我们的控制，因为我们理所当然地会购买我们能买到的最好的配料，无论其价格水平如何变化。我们认为产品质量是神圣的。
但其他类型的成本更可控，而正是在这个领域我们遇到了问题。按每磅计算，我们的成本（不包括原材料成本）在过去几年中以显著高于整体价格水平的增长速度增长。扭转这一趋势对我们的竞争地位和盈利潜力至关重要。
最近几个月，我们对成本实施了更好的控制，我们确信1984年这些成本的增长率将低于通胀率。这种信心源于我们与查克·哈金斯管理才能长期合作的经验。我们从接手那天起就让查克负责，他的记录简直非凡，如下表所示：
52-53周财年 营业 利润 糖果磅数 年末开店数
约至12月31日 销售收入 税后利润 糖果销量 年末开店数
------------------- ------------ ----------- ---------- -----------
1983（53周）...... $133,531,000 $13,699,000 24,651,000 207
1982 .............. 123,662,000 11,875,000 24,216,000 202
1981 .............. 112,578,000 10,779,000 24,052,000 199
1980 .............. 97,715,000 7,547,000 24,065,000 191
1979 .............. 87,314,000 6,330,000 23,985,000 188
1978 .............. 73,653,000 6,178,000 22,407,000 182
1977 .............. 62,886,000 6,154,000 20,921,000 179
1976（53周）...... 56,333,000 5,569,000 20,553,000 173
1975 .............. 50,492,000 5,132,000 19,134,000 172
1974 .............. 41,248,000 3,021,000 17,883,000 170
1973 .............. 35,050,000 1,940,000 17,813,000 169
1972 .............. 31,337,000 2,083,000 16,954,000 167
正如表格所示，我们面临的另一个问题是最近无法实现糖果销量的有意义增长。整个行业也有同样的问题。但多年来我们在这一方面表现优于行业，而现在没有。
过去四年中，我们零售店的磅数销量几乎每年都没什么变化，尽管店铺数量每年都略有增加（分销费用也是如此）。当然，美元销售额增加了，因为我们大幅提高了价格。但我们认为零售趋势最重要的衡量标准是每家店铺售出的单位数量，而不是美元销售额。在同店基础上（仅计算全年都开业的店铺），所有数字调整到52周年份，1983年磅数下降了0.8%。这一小幅下降是自1979年以来我们最好的同店表现；自那以来的累计下降约为8%。大宗订单量约占我们总量的25%，在1970年代经历了非常大的磅数增长后，近年来趋于平稳。
我们不确定这种销量持平——无论是在零售店领域还是大宗订单领域——在多大程度上是由于我们的定价政策，在多大程度上是由于行业销量停滞、经济衰退以及我们已在主要销售区域拥有的异常市场份额。我们1984年的涨价幅度比过去几年温和得多，我们希望明年能向您报告更好的销量数据。但我们没有预测的依据。
尽管存在销量问题，喜诗的优点很多且重要。在我们的主要销售区域——西部地区，我们糖果的受欢迎程度远超任何竞争对手。事实上，我们相信大多数巧克力爱好者更喜欢它，即使价格是其他糖果的两三倍。（糖果和股票一样，价格和价值可能不同；价格是你付出的，价值是你得到的。）我们店铺的客户服务质量——由我们在全国范围内自营，而非特许经营——与产品本身一样好。愉快、乐于助人的员工与盒子上的标志一样成为喜诗的商标。在需要雇佣大约2000名季节性工人的业务中，这可不是小成就。我们不知道有哪个同等规模的组织能提供比查克·哈金斯及其同事更优质的客户服务。
由于我们在1984年涨价幅度很小，我们预计喜诗今年的利润与1983年大致持平。
保险——受控业务
我们既经营保险公司，也在一家我们不经营的保险业务——GEICO——中拥有重大经济利益。所有结果可以简单总结：总的来说，我们经营的、其承保结果反映了几年前由我负责决策的那些公司，结果绝对糟糕。幸运的是，我不影响其政策的GEICO简直是光芒四射。你从这个总结中得出的推论是正确的。我在几年前犯了一些严重错误，现在报应来了。
该行业经历了很长时间以来最糟糕的承保年份，如下表所示：
保费增长率（%） 含保单持有人分红的综合比率
------------- --------------------
1972 .................... 10.2 96.2
1973 .................... 8.0 99.2
1974 .................... 6.2 105.4
1975 .................... 11.0 107.9
1976 .................... 21.9 102.4
1977 .................... 19.8 97.2
1978 .................... 12.8 97.5
1979 .................... 10.3 100.6
1980 .................... 6.0 103.1
1981 .................... 3.9 106.0
1982（修订）.......... 4.4 109.7
1983（估计）.......... 4.6 111.0
来源：Best's Aggregates and Averages。
Best的数据反映了几乎整个行业的经验，包括股份公司、相互公司和互惠公司。综合比率表示总保险成本（已发生损失加费用）与保费收入的比较；比率低于100表示承保利润，高于100表示亏损。
由于去年报告中概述的原因，我们预计1983年糟糕的行业经验在未来许多年或多或少都是典型。（正如约吉·贝拉所说：“这将是又一次似曾相识。”）这并不意味着我们认为数字不会有所波动；它们肯定会。但我们认为该十年剩余时间内综合比率平均显著低于1981-1983年水平的可能性非常小。基于我们对通胀的预期——我们在这方面仍然像以往一样悲观——行业保费收入必须每年增长约10%才能勉强稳定在当前的损失率水平。
我们1983年自己的综合比率为121。由于迈克·戈德堡最近接手了保险业务的大部分责任，如果我们的缺点能归到他而不是我的名下，那对我来说会很好。但不幸的是，正如我们经常指出的，保险业务有很长的前置时间。尽管业务政策可能改变，人员可能改善，但必须经过相当长的时间才能看到效果。（这一业务特征使我们能够在GEICO赚很多钱；我们可以在实际发生之前就描绘出可能发生的事情。）因此，1983年结果的根源在于两年或更早之前的经营和人事决策，那时我直接负责保险集团的管理责任。
尽管我们整体结果糟糕，但我们的几位经理确实做得非常出色。罗兰·米勒引导了国家赔偿公司和国家火灾与海洋保险公司的汽车和一般责任险业务取得了改善的结果，而竞争对手的结果却恶化了。此外，大陆分水岭保险公司的汤姆·罗利——我们羽翼未丰的科罗拉多本州公司——似乎肯定会成为赢家。迈克在一年多前找到了他，他是一次重要收购。
我们最近变得活跃起来——并希望变得更加活跃——在再保险交易中，买方最关心的应该是卖方的长期信誉。在这样的交易中，我们首要的财务实力应该使我们成为索赔人和保险公司的首选，他们必须在未来许多年里依赖再保险公司的承诺。
这类业务的一个主要来源是结构性结算——一种解决损失的程序，索赔人获得定期付款（几乎总是每月，终身），而不是一次性总付结算。这种结算形式对索赔人有重要的税收优势，也防止他们挥霍大笔一次性付款。通常，结算中包含一些通胀保护。通常索赔人受了重伤，因此定期付款必须在未来几十年内绝对安全。我们相信我们提供了无与伦比的安全性。我们不知道有其他保险公司——即使那些总资产大得多的——有我们的财务实力。
我们还认为，我们的财务实力应该让我们成为希望转移损失准备金的公司的推荐选择。在这样的交易中，其他保险公司支付一次性款项给我们，让我们承担适用于大量已过期业务的所有（或特定部分）未来损失支付。在这里，转移此类索赔的公司也需要确保受让方在未来许多年的财务实力。同样，大多数征求此类业务的竞争对手在我们看来财务状况远不如我们。
潜在来看，结构性结算和损失准备金承担可能对我们非常重要。由于它们的潜在规模，以及这些业务相比保费收入会产生大量投资收益，我们将把这些业务的承保结果在我们的保险分部数据中单独列示。我们也会在向您报告综合比率时排除它们的影响。我们不对结构性结算或损失准备金交易预先确认任何利润，所有相关间接费用当期列支。这两项业务都由国家赔偿公司的唐·沃斯特负责。
保险——GEICO
GEICO在1983年的表现与我们的保险表现一样糟糕。相比行业111的综合比率，GEICO在自愿为保单持有人分红大额计提后，综合比率为96。几年前我不会想到GEICO能如此大幅超越行业。它的优势反映了真正卓越的商业理念和卓越的管理相结合。
杰克·伯恩和比尔·斯奈德在承保领域保持了非凡的纪律（包括关键的全额和适当的损失准备金计提），他们的努力现在正通过新业务的大幅增长得到进一步回报。同样重要的是，卢·辛普森在保险投资经理中是顶尖的。他们三个是了不起的团队。
我们在GEICO拥有约三分之一的权益。这使我们在该公司的保费收入中占2.7亿美元，比我们自己的保费收入高出约80%。因此，我们总保险业务的主要部分来自全国最好的保险业务。这一事实并没有丝毫减轻我们改善自身业务的需求。
股票拆分与股票活动
我们经常被问到为什么伯克希尔不拆分股票。这个问题背后的假设通常是拆分是对股东有利的行为。我们不同意。让我告诉你为什么。
我们的目标之一是让伯克希尔·哈撒韦的股票价格与其内在商业价值理性相关。（但请注意“理性相关”，而不是“相同”：如果备受推崇的公司一般在市场上以大幅低于价值的价格出售，伯克希尔也可能被类似定价。）理性股价的关键在于理性的股东，无论是现有股东还是潜在股东。
如果一家公司的股东和/或吸引来的潜在买家倾向于做出非理性或基于情绪的决定，那么一些相当愚蠢的股价就会周期性出现。躁郁症性格产生躁郁症估值。这种异常现象可能有助于我们买卖其他公司的股票。但我们认为，在伯克希尔的市场中，最大限度地减少这种情况的发生符合你和我们的共同利益。
获得高质量股东并不容易。阿斯特夫人可以选择她的400人，但任何人都可以买任何股票。股东“俱乐部”的进入成员不能对其智力能力、情绪稳定性、道德敏感性或可接受的着装进行筛选。因此，股东优生学似乎是一项无望的事业。
然而，在很大程度上，我们觉得如果我们始终如一地传达我们的业务和所有权理念——并且不传递相互矛盾的信息——然后让自我选择顺其自然，就可以吸引和维持高质量的所有权。例如，自我选择会把完全不同的人群吸引到一场被宣传为歌剧的音乐活动，而不是摇滚音乐会，尽管任何人都可以买任何一种门票。
通过我们的政策和沟通——我们的“广告”——我们试图吸引能够理解我们的运营、态度和期望的投资者。（同样重要的是，我们试图劝阻那些不会理解的人。）我们希望那些把自己视为企业所有者、并以长期持有为目的投资公司的人。而且，我们希望那些把目光集中在业务成果而不是市场价格上的人。
拥有这些特征的投资者是少数，但我们拥有非凡的一群。我相信超过90%——可能超过95%——的股份由五年前就是伯克希尔或蓝筹印花股东的人持有。我猜想超过95%的股份由那些持股至少是其第二大持股两倍的投资者持有。在拥有至少几千名公众股东且市值超过10亿美元的公司中，我们在股东像所有者一样思考和行动的程度方面几乎肯定是领先的。提升已经具备这些特征的股东群体并不容易。
如果我们拆分股票或采取其他关注股价而非商业价值的行动，我们将吸引一批比退出卖家低劣的进入买家。在1300美元的价格上，很少有投资者买不起一股伯克希尔。如果我们将1股拆成100股，让潜在的单股买家能够购买100股，他们会更好吗？那些这样认为并因此或因预期拆分而购买股票的人，一定会降低我们现有股东群体的质量。（我们真的能通过用一些现有头脑清晰的股东换来那些喜欢纸片胜过价值、觉得拥有九张10美元钞票比一张100美元钞票更富有的易受影响的新股东来改善股东群体吗？）为了非价值原因购买的人很可能为了非价值原因出售。他们在画面中的存在会加剧与基础业务发展无关的波动性价格波动。
我们将努力避免吸引短期关注我们股价的买家的政策，并努力遵循吸引关注商业价值的知情长期投资者的政策。就像你在一个由理性知情投资者组成的市场购买你的伯克希尔股票一样，你应该有机会——如果你愿意的话——在同样的市场出售。我们将努力保持它的存在。
股票市场的讽刺之一是强调活动。经纪人使用“可销售性”和“流动性”等术语，赞扬高换手率的公司（那些无法填满你口袋的人会自信地填满你的耳朵）。但投资者应该明白，对庄家有利的，对顾客不一定有利。过度活跃的股票市场是企业的小偷。
例如，考虑一家典型公司，假设备净资产收益率为12%。假设其股票年换手率极高，达到100%。如果每次买卖各抽取1%的佣金（低价股可能更高），并且股票按账面价值交易，那么我们假设公司的所有者每年将支付公司净资产的2%作为转移所有权的代价。这种活动对企业的收益毫无帮助，意味着其中1/6通过“摩擦性”转移成本损失给了所有者。（这个计算没有考虑期权交易，这会使摩擦成本进一步增加。）
所有这些都使得游戏音乐椅相当昂贵。你能想象如果政府对公司或投资者的收益征收16又2/3%的新税，会引起多么痛苦的呼喊吗？通过市场活动，投资者可以对自己施加等同于这种税收的税。
市场交易日成交量1亿股的日子（包括场外交易，这种成交量今天异常低）对所有者来说是诅咒而非祝福——因为它们意味着所有者换椅子要支付两倍于5000万股日的代价。如果1亿股日持续一年，每次买卖平均成本为每股15美分，那么投资者的总换椅税将达到约75亿美元——大致相当于埃克森、通用汽车、美孚和德士古这四家《财富》500强最大公司1982年利润的总和。
这些公司在1982年底的净资产总额为750亿美元，占整个《财富》500强净资产和净利润的12%以上。根据我们的假设，投资者每年为了满足他们对“金融翻转”的嗜好而放弃所有这些巨额资本的收益。此外，每年超过20亿美元的投资管理费——为换椅建议支付的费用——需要投资者放弃五大银行组织（花旗集团、美洲银行、大通曼哈顿、汉华实业和摩根大通）的所有收益。这些昂贵的活动可能决定谁吃馅饼，但它们不会扩大馅饼。
（我们知道一种扩大馅饼的论点，说这种活动改善了资本配置过程的合理性。我们认为这种论点似是而非，总的来说，过度活跃的股票市场会破坏理性的资本配置，并成为馅饼缩小器。亚当·斯密认为自由市场中所有非共谋行为都由一只看不见的手引导，使经济实现最大进步；我们的观点是，赌场式市场和触发式投资管理就像一只看不见的脚，绊倒并拖慢前进的经济。）
将过度活跃的股票与伯克希尔对比。我们股票的买卖价差目前约为30点，略高于2%。根据交易量大小，伯克希尔卖方的收入与买方的成本之差可能从4%（仅涉及少量股票的交易）到大约1.5%（在大额交易中，协商可以减少做市商价差和经纪人佣金）不等。由于大多数伯克希尔股票在相当大额交易中交易，所有交易的平均价差可能不超过2%。
与此同时，伯克希尔股票的真实换手率（排除交易商间交易、赠与和遗赠）可能每年3%。因此，我们的所有者总共每年支付伯克希尔市值的大约6/100的1%作为转移特权费用。根据这个非常粗略的估计，那是90万美元——不是小数目，但远低于平均水平。拆分股票将增加这一成本，降低我们股东群体的质量，并鼓励市场价格与内在商业价值的相关性降低。我们看不到任何抵消优势。
杂项
去年在这一部分，我放了一个小广告来鼓励收购候选者。在我们的通信业务中，我们告诉广告客户，重复是产生效果的关键（确实如此），所以我们将再次重复我们的收购标准。
我们偏好：
(1) 大额收购（至少500万美元税后利润），
(2) 经过验证的持续盈利能力（未来预测对我们几乎没有兴趣，转型情况也是如此），
(3) 净资产收益率良好且很少或没有使用债务的企业，
(4) 现有管理层（我们不能提供），
(5) 简单的业务（如果技术含量很高，我们就无法理解），
(6) 报价（我们不想浪费自己或卖家的时间，在价格未知的情况下即使初步讨论交易也不行）。
我们不会进行恶意收购。我们可以承诺完全保密和非常快速的答复——通常在五分钟内——关于我们是否感兴趣。我们更倾向于现金收购，但当我们获得与付出相等的内在商业价值时，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往来的人来核实我们。对于合适的业务——和合适的人——我们可以提供一个好的归宿。
* * * * *
约96.4%的符合条件的股份参与了我们的1983年股东指定捐款计划。根据该计划进行的捐款总额——于1984年初发放，但全部在1983年列支——为3,066,501美元，共有1353家慈善机构受益。尽管按参与股份百分比衡量的回应异常好，但按参与持有人百分比衡量的回应却不那么好。原因可能是通过合并获得的大量新股东对该计划不熟悉。我们敦促新股东阅读第52-53页对该计划的描述。
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名义注册，而不是以“街道”或代名人名义。未在1984年9月28日这样注册的股票将没有资格参与任何1984年计划。
* * * * *
蓝筹印花/伯克希尔合并顺利完成。每家公司不到0.1%的股份投票反对合并，没有要求评估的请求。1983年，我们从合并中获得了一些税收效率，并预计未来会获得更多。
合并中一个有趣的侧面：伯克希尔现在有1,146,909股流通股，而现任管理层接管业务的1965财政年度初为1,137,778股。如果你当时拥有公司1%的股份，现在你拥有0.99%。因此，今天的所有资产——布法罗新闻、喜诗、内布拉斯加家具城、保险集团、13亿美元的可流通股票等等——都是在几乎没有对原始所有者造成净稀释的情况下添加到原始纺织资产上的。
我们很高兴前蓝筹印花股东加入我们。为了帮助您了解伯克希尔·哈撒韦，我们很乐意向您发送1977-1981年年报的信件汇编和/或1982年年报。请将请求发送至公司地址：1440 Kiewit Plaza, Omaha, Nebraska 68131。
沃伦·E·巴菲特
1984年3月14日 董事长
附录
伯克希尔·哈撒韦公司
商誉及其摊销：规则与现实
本附录仅涉及经济和会计商誉——而非日常用语中的商誉。例如，一家企业可能深受大多数顾客喜爱甚至爱戴，但可能没有经济商誉。（AT&T在分拆前普遍受到好评，但没有一毛钱的经济商誉。）而且，遗憾的是，一家企业可能不受顾客欢迎，但拥有可观且不断增长的经济商誉。所以，暂时忘记情感，只关注经济学和会计。
当一家企业被收购时，会计准则要求将收购价格首先分配到所收购的可辨认资产的公允价值。通常，资产公允价值之和（扣除负债后）低于总收购价格。在这种情况下，差额被分配给一个名为“成本超过所收购净资产权益的金额”的资产账户。为了避免不断重复这一拗口的说法，我们将称为“商誉”。
1970年11月之前购买的商业产生的会计商誉具有特殊地位。除非在罕见情况下，只要所购买的业务被保留，它就可以作为资产负债表上的资产。这意味着不需要从收益中扣除摊销费用来逐渐消除该资产。
但对于1970年11月之后的购买，情况不同。当这些购买产生商誉时，必须在不超过40年内通过每年等额的收益扣除来摊销。由于40年是允许的最长期限，40年是管理层（包括我们）通常选择的期限。这种每年对收益的扣除不允许作为税收减免，因此对税后收入的影响大约是大多数其他费用的两倍。
这就是会计商誉的运作方式。为了看清它如何与经济现实不同，让我们看一个身边的例子。我们将取整一些数字，并大大简化，以使例子更容易理解。我们还将提到一些对投资者和管理者的启示。
蓝筹印花公司在1972年初以2500万美元收购了喜诗，当时喜诗约有800万美元的有形净资产。（在整个讨论中，应收账款将被归类为有形资产，这对商业分析是一个合适的定义。）这一水平的有形资产足以在不需要债务的情况下经营业务，除了季节性短期。当时喜诗税后利润约为200万美元，这些利润似乎保守地代表了以1972年不变美元计算的未来盈利能力。
因此，我们的第一个教训：当企业可以预期在这些资产上获得远高于市场回报率的收益时，其价值在逻辑上远高于有形净资产。这种超额收益的资本化价值就是经济商誉。
在1972年（现在也是），很少有企业能够预期在有形净资产上持续获得25%的税后收益，就像喜诗那样——而且是通过保守的会计和没有财务杠杆做到的。产生这种溢价回报率的并不是存货、应收款或固定资产的公允市场价值。而是无形资产的组合，特别是基于消费者对产品和人员无数次愉快经历而形成的广泛的有利声誉。
这种声誉创造了一个消费者特许经营权，使得产品对购买者的价值，而不是其生产成本，成为销售价格的主要决定因素。消费者特许经营权是经济商誉的主要来源。其他来源包括不受利润监管的政府特许经营权，如电视台，以及在一个行业中作为低成本生产者的持久地位。
让我们回到喜诗的例子中的会计问题。蓝筹印花以超过有形净资产1700万美元的价格收购喜诗，这要求在蓝筹印花的账簿上建立一个1700万美元的商誉资产账户，并且每年从收益中扣除42.5万美元，持续40年摊销该资产。到1983年，经过11年这样的扣除，1700万美元已减少到约1250万美元。与此同时，伯克希尔持有蓝筹印花60%的股份，因此也持有喜诗60%的股份。这一所有权意味着伯克希尔的资产负债表反映了喜诗商誉的60%，即约750万美元。
1983年，伯克希尔通过合并收购了蓝筹印花其余股份，这需要购买会计处理，而不是某些合并允许的“权益结合法”处理。在购买会计下，我们给予（或“支付给”）蓝筹印花持有人的股份的“公允价值”必须分摊到从蓝筹印花收购的净资产上。这种“公允价值”通常是按公开公司使用其股份进行收购时，以所放弃股份的市场价值衡量的。
“购买”的资产包括蓝筹印花拥有的40%的一切（如前所述，伯克希尔已拥有另外60%）。伯克希尔“支付”的比我们收到的可辨认净资产多5170万美元，被分配到两块商誉：2840万美元给喜诗，2330万美元给布法罗晚报。
因此，合并后，伯克希尔留下了一个喜诗的商誉资产，包含两个组成部分：1971年购买剩余的750万美元，以及1983年“购买”40%新产生的2840万美元。我们未来的摊销费用将在未来28年每年约100万美元，以及在接下来的12年（2002年至2013年）每年约70万美元。
换句话说，不同的购买日期和价格给了我们同一资产的两部分截然不同的资产价值和摊销费用。（我们重复我们通常的免责声明：我们提不出更好的会计系统。要处理的问题令人头疼，需要武断的规则。）
但经济现实是什么？一个现实是，自收购喜诗以来每年在收益表中作为成本扣除的摊销费用并非真正的经济成本。我们知道这一点，因为去年喜诗在约2000万美元的有形净资产上赚取了1300万美元的税后利润——这一业绩表明存在远大于我们会计商誉原始总成本的经济商誉。换句话说，虽然会计商誉从购买那一刻起定期减少，但经济商誉以不规则但非常实质性的方式增加。
另一个现实是，未来的年度摊销费用将与经济成本不符。当然，喜诗的经济商誉有可能消失。但它不会以均匀的递减或任何类似的方式减少。更可能的是，商誉会增加——以当前美元计，如果不是不变美元的话——因为通胀。
这种可能性之所以存在，是因为真正的经济商誉往往名义价值随通胀比例上升。为了说明这一点，让我们将喜诗这类业务与更普通的业务进行对比。回想一下，我们在1972年购买喜诗时，它在800万美元的有形净资产上赚了约200万美元。假设我们假设的普通业务当时也有200万美元的收益，但需要1800万美元的有形净资产进行正常运营。在所需有形资产上仅赚取11%，这种普通业务几乎没有经济商誉。
因此，这样的业务很可能会按其有形净资产的价值出售，即1800万美元。相比之下，我们为喜诗支付了2500万美元，尽管它的收益并不更多，而“真实”资产只有一半。正如我们的购买价格所暗示的那样，少真的可以是多吗？答案是“是”——即使两个业务都预计单位销量持平——只要你像我们在1972年那样预期一个持续通胀的世界。
为了理解原因，想象一下随后价格水平翻倍对两个业务的影响。两者都需要将名义收益翻倍到400万美元才能保持与通胀同步。这似乎不是什么大问题：以双倍于之前的价格出售相同数量的单位，假设利润率不变，利润也必然翻倍。
但关键的是，要实现这一点，两个业务都可能需要将名义有形资产投资翻倍，因为通胀通常会对好企业和坏企业施加这种经济要求。销售额翻倍意味着必须立即在应收款和存货中投入相应更多的资金。固定资产中的资金将对通胀反应较慢，但几乎同样确定。所有这些通胀所需的投资都不会提高回报率。这种投资的动机是企业的生存，而不是所有者的繁荣。
然而，记住喜诗只有800万美元的有形净资产。因此它只需要额外投入800万美元来满足通胀带来的资本需求。与此同时，普通业务负担是两倍多——需要1800万美元的额外资本。
尘埃落定之后，现在每年赚400万美元的普通业务可能仍然值其有形资产的价值，即3600万美元。这意味着其所有者每投资1美元新资本，只获得了1美元的名义价值。（如果他们把钱存入储蓄账户，也会得到同样的美元对美元结果。）
而喜诗，同样赚400万美元，如果按我们购买时的相同基础估值（逻辑上会如此），可能值5000万美元。因此，当所有者仅投入800万美元额外资本时，它获得了2500万美元的名义价值——每投入1美元，获得超过3美元的名义价值。
然而，即使这样，喜诗这类业务的所有者也因通胀被迫额外投入800万美元资本，仅仅是为了保持实际利润不变。任何需要一些有形净资产才能运营且没有杠杆的业务（几乎所有业务都如此）都会受到通胀的伤害。只需要很少有形资产的业务只是受伤最轻。
当然，这一事实对许多人来说很难理解。多年来，传统智慧——传统悠久，智慧不足——认为通胀保护最好由拥有自然资源、厂房和机器或其他有形资产的企业提供（“我们信任商品”）。实际上并非如此。资产密集型业务通常回报率低——这些回报率往往勉强提供足够资本来满足现有业务的通胀需求，没有剩余用于实际增长、分配给所有者或收购新业务。
相比之下，在通胀年代建立起来的重大商业财富中，不成比例的部分来自那些拥有持久价值无形资产、同时对有形资产要求较少的业务。在这种情况下，收益以名义美元大幅上升，而这些美元大部分可用于收购额外业务。这一现象在通信业务中尤其明显。该业务对有形投资要求很少——但其特许经营权却能持久。在通胀期间，商誉是一份不断给予的礼物。
但这句话自然只适用于真正的经济商誉。虚假的会计商誉——周围有很多——则是另一回事。当过度兴奋的管理层以愚蠢的价格收购一项业务时，前面描述的相同的会计细节会被遵守。由于它无处可去，愚蠢行为最终落在商誉账户中。考虑到创造该账户的管理缺乏纪律，在这种情况下，它可能更适合被标记为“无商誉”。无论用什么术语，通常40年的仪式会被遵守，这样资本化的肾上腺素作为“资产”留在账面上，好像收购是明智的。
如果你仍然相信商誉的会计处理是衡量经济现实的最佳标准，我建议你考虑最后一个项目。
假设一家公司每股净资产20美元，全部为有形资产。进一步假设该公司内部开发了某种宏伟的消费者特许经营权，或者幸运地通过原始FCC授权获得了一些重要的电视台。因此，它在有形资产上赚了很多，比如每股5美元，即25%。
基于这样的经济状况，它可能卖到每股100美元或更多，并且很可能在整体业务协商出售中也能卖到这个价格。
假设一位投资者以每股100美元买入股票，实际上为商誉支付了每股80美元（就像收购整个公司的企业买家一样）。投资者是否应该每年计提2美元的摊销费用（80美元除以40年）来计算“真实”每股收益？如果是这样，新的每股3美元的“真实”收益是否应该让他重新考虑购买价格？
我们认为管理者和投资者都应从两个角度来看待无形资产：
-
在分析经营业绩时——即在评估业务单元的潜在经济状况时——应忽略摊销费用。一家企业预期能在无杠杆有形净资产上赚取的收益（排除商誉摊销对收益的任何扣除）是衡量该业务经济吸引力的最佳指南。它也是衡量该业务经济商誉当前价值的最佳指南。
-
在评估业务收购的明智性时，也应忽略摊销费用。它们既不应从收益中扣除，也不应从业务成本中扣除。这意味着永远以全额成本（在任何摊销之前）来查看购买的商誉。此外，成本应定义为所给予的所有对价的全额内在商业价值——而不是记录的会计价值——无论合并时所涉证券的市场价格如何，也无论是否允许权益结合法处理。例如，在蓝筹印花合并中，我们为喜诗和布法罗新闻40%的商誉实际支付的价格远多于账面上记录的5170万美元。这种差异之所以存在，是因为合并中放弃的伯克希尔股票的市场价值低于其内在商业价值，而内在商业价值才定义了我们的真实成本。
基于视角（1）看起来是赢家的业务，在视角（2）下可能黯然失色。好业务不一定总是好购买——尽管它是寻找好购买的好地方。
我们将努力收购那些从（1）衡量具有优秀经营经济性，且从（2）衡量能提供合理回报的业务。会计后果将被完全忽略。
截至1983年底，我们会计账面上的商誉净额为6200万美元，包括你在资产负债表资产方看到的7900万美元，以及抵消我们持有的互助储蓄与贷款权益账面价值的1700万美元负商誉。
我们相信净经济商誉远超过6200万美元的会计数字。