In 1983 Buffett created an owner’s manual for Berkshire shareholders. The current version is available at http://www.berkshirehathaway.com/ownman.pdf
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.
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.”)
Book value vs. intrinsic value
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:
- 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;
- 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.
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.
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 ⅙ 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 ⅔% 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 ½% (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.
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.
Regarding Goodwill and amortization (from the appendix of the 1983 letter)
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.
Leave a Reply