Berkshire 1979 Letter excerpts
1979 Operating Results
WARREN BUFFETT: We continue to feel that the ratio of operating earnings (before securities gains or losses) to shareholders’ equity with all securities valued at cost is the most appropriate way to measure any single year’s operating performance.
Measuring such results against shareholders’ equity with securities valued at market could significantly distort the operating performance percentage because of wide year-to-year market value changes in the net worth figure that serves as the denominator. For example, a large decline in securities values could result in a very low “market value” net worth that, in turn, could cause mediocre operating earnings to look unrealistically good. Alternatively, the more successful that equity investments have been, the larger the net worth base becomes and the poorer the operating performance figure appears. Therefore, we will continue to report operating performance measured against beginning net worth, with securities valued at cost.
On this basis, we had a reasonably good operating performance in 1979 - but not quite as good as that of 1978 - with operating earnings amounting to 18.6% of beginning net worth. Earnings per share, of course, increased somewhat (about 20%) but we regard this as an improper figure upon which to focus. We had substantially more capital to work with in 1979 than in 1978, and our performance in utilizing that capital fell short of the earlier year, even though per-share earnings rose. “Earnings per share” will rise constantly on a dormant savings account or on a U.S. Savings Bond bearing a fixed rate of return simply because “earnings” (the stated interest rate) are continuously plowed back and added to the capital base. Thus, even a “stopped clock” can look like a growth stock if the dividend payout ratio is low.
The primary test of managerial economic performance is the achievement of a high earnings rate on equity capital employed (without undue leverage, accounting gimmickry, etc.) and not the achievement of consistent gains in earnings per share. In our view, many businesses would be better understood by their shareholder owners, as well as the general public, if managements and financial analysts modified the primary emphasis they place upon earnings per share, and upon yearly changes in that figure.
Long Term Results
In measuring long term economic performance - in contrast to yearly performance - we believe it is appropriate to recognize fully any realized capital gains or losses as well as extraordinary items, and also to utilize financial statements presenting equity securities at market value. Such capital gains or losses, either realized or unrealized, are fully as important to shareholders over a period of years as earnings realized in a more routine manner through operations; it is just that their impact is often extremely capricious in the short run, a characteristic that makes them inappropriate as an indicator of single year managerial performance.
The book value per share of Berkshire Hathaway on September 30, 1964 (the fiscal year end prior to the time that your present management assumed responsibility) was $19.46 per share. At year end 1979, book value with equity holdings carried at market value was $335.85 per share. The gain in book value comes to 20.5% compounded annually. This figure, of course, is far higher than any average of our yearly operating earnings calculations, and reflects the importance of capital appreciation of insurance equity investments in determining the overall results for our shareholders. It probably also is fair to say that the quoted book value in 1964 somewhat overstated the intrinsic value of the enterprise, since the assets owned at that time on either a going concern basis or a liquidating value basis were not worth 100 cents on the dollar. (The liabilities were solid, however.)
We have achieved this result while utilizing a low amount of leverage (both financial leverage measured by debt to equity, and operating leverage measured by premium volume to capital funds of our insurance business), and also without significant issuance or repurchase of shares. Basically, we have worked with the capital with which we started. From our textile base we, or our Blue Chip and Wesco subsidiaries, have acquired total ownership of thirteen businesses through negotiated purchases from private owners for cash, and have started six others.
Just as the original 3% savings bond, a 5% passbook savings account or an 8% U.S. Treasury Note have, in turn, been transformed by inflation into financial instruments that chew up, rather than enhance, purchasing power over their investment lives, a business earning 20% on capital can produce a negative real return for its owners under inflationary conditions not much more severe than presently prevail.
If we should continue to achieve a 20% compounded gain - not an easy or certain result by any means - and this gain is translated into a corresponding increase in the market value of Berkshire Hathaway stock as it has been over the last fifteen years, your after-tax purchasing power gain is likely to be very close to zero at a 14% inflation rate. Most of the remaining six percentage points will go for income tax any time you wish to convert your twenty percentage points of nominal annual gain into cash.
That combination - the inflation rate plus the percentage of capital that must be paid by the owner to transfer into his own pocket the annual earnings achieved by the business (i.e., ordinary income tax on dividends and capital gains tax on retained earnings) - can be thought of as an “investor’s misery index”. When this index exceeds the rate of return earned on equity by the business, the investor’s purchasing power (real capital) shrinks even though he consumes nothing at all. We have no corporate solution to this problem; high inflation rates will not help us earn higher rates of return on equity.
One friendly but sharp-eyed commentator on Berkshire has pointed out that our book value at the end of 1964 would have bought about one-half ounce of gold and, fifteen years later, after we have plowed back all earnings along with much blood,sweat and tears, the book value produced will buy about the same half ounce. A similar comparison could be drawn with Middle Eastern oil. The rub has been that government has been exceptionally able in printing money and creating promises, but is unable to print gold or create oil.
We intend to continue to do as well as we can in managing the internal affairs of the business. But you should understand that external conditions affecting the stability of currency may very well be the most important factor in determining whether there are any real rewards from your investment in Berkshire Hathaway.
Sources of Earnings
We again present a table showing the sources of Berkshire’s earnings. As explained last year, Berkshire owns about 60% of Blue Chip Stamps which, in turn, owns 80% of Wesco Financial Corporation. The table shows both aggregate earnings of the various business entities, as well as Berkshire’s share. All of the significant capital gains or losses attributable to any of the business entities are aggregated in the realized securities gain figure at the bottom of the table, and are not included in operating earnings.
Textile and Retailing
Our textile business also continues to produce some cash, but at a low rate compared to capital employed. This is not a reflection on the managers, but rather on the industry in which they operate. In some businesses - a network TV station, for example - it is virtually impossible to avoid earning extraordinary returns on tangible capital employed in the business. And assets in such businesses sell at equally extraordinary prices, one thousand cents or more on the dollar, a valuation reflecting the splendid, almost unavoidable, economic results obtainable. Despite a fancy price tag, the “easy” business may be the better route to go.
Both our operating and investment experience cause us to conclude that “turnarounds” seldom turn, and that the same energies and talent are much better employed in a good business purchased at a fair price than in a poor business purchased at a bargain price. Although a mistake, the Waumbec acquisition has not been a disaster. Certain portions of the operation are proving to be valuable additions to our decorator line (our strongest franchise) at New Bedford, and it’s possible that we may be able to run profitably on a considerably reduced scale at Manchester. However, our original rationale did not prove out.
Insurance Underwriting
We predicted last year that the combined underwriting ratio (see definition on page 36) for the insurance industry would “move up at least a few points, perhaps enough to throw the industry as a whole into an underwriting loss position”. That is just about the way it worked out. The industry underwriting ratio rose in 1979 over three points, from roughly 97.4% to 100.7%. We also said that we thought our underwriting performance relative to the industry would improve somewhat in 1979 and, again, things worked out as expected. Our own underwriting ratio actually decreased from 98.2% to 97.1%. Our forecast for 1980 is similar in one respect; again we feel that the industry’s performance will worsen by at least another few points. However, this year we have no reason to think that our performance relative to the industry will further improve. (Don’t worry - we won’t hold back to try to validate that forecast.)
Our reinsurance division, led by George Young, continues to give us reasonably satisfactory overall results after allowing for investment income, but underwriting performance remains unsatisfactory. We think the reinsurance business is a very tough business that is likely to get much tougher. In fact, the influx of capital into the business and the resulting softer price levels for continually increasing exposures may well produce disastrous results for many entrants (of which they may be blissfully unaware until they are in over their heads; much reinsurance business involves an exceptionally “long tail”, a characteristic that allows catastrophic current loss experience to fester undetected for many years). It will be hard for us to be a whole lot smarter than the crowd and thus our reinsurance activity may decline substantially during the projected prolonged period of extraordinary competition.
The conventional wisdom is that insurance underwriting overall will be poor in 1980, but that rates will start to firm in a year or so, leading to a turn in the cycle some time in 1981. We disagree with this view. Present interest rates encourage the obtaining of business at underwriting loss levels formerly regarded as totally unacceptable. Managers decry the folly of underwriting at a loss to obtain investment income, but we believe that many will. Thus we expect that competition will create a new threshold of tolerance for underwriting losses, and that combined ratios will average higher in the future than in the past.
To some extent, the day of reckoning has been postponed because of marked reduction in the frequency of auto accidents - probably brought on in major part by changes in driving habits induced by higher gas prices. In our opinion, if the habits hadn’t changed, auto insurance rates would have been very little higher and underwriting results would have been much worse. This dosage of serendipity won’t last indefinitely.
Our forecast is for an average combined ratio for the industry in the 105 area over the next five years. While we have a high degree of confidence that certain of our operations will do considerably better than average, it will be a challenge to us to operate below the industry figure. You can get a lot of surprises in insurance. Nevertheless, we believe that insurance can be a very good business. It tends to magnify, to an unusual degree, human managerial talent - or the lack of it. We have a number of managers whose talent is both proven and growing. (And, in addition, we have a very large indirect interest in two truly outstanding management groups through our investments in SAFECO and GEICO.) Thus we expect to do well in insurance over a period of years. However, the business has the potential for really terrible results in a single specific year. If accident frequency should turn around quickly in the auto field, we, along with others, are likely to experience such a year.
Insurance Investments
In recent years we have written at length in this section about our insurance equity investments. In 1979 they continued to perform well, largely because the underlying companies in which we have invested, in practically all cases, turned in outstanding performances. Retained earnings applicable to our insurance equity investments, not reported in our financial statements, continue to mount annually and, in aggregate, now come to a very substantial number. We have faith that the managements of these companies will utilize those retained earnings effectively and will translate a dollar retained by them into a dollar or more of subsequent market value for us. In part, our unrealized gains reflect this process.
Below we show the equity investments which had a year end market value of over $5 million:
We currently believe that equity markets in 1980 are likely to evolve in a manner that will result in an underperformance by our portfolio for the first time in recent years. We very much like the companies in which we have major investments, and plan no changes to try to attune ourselves to the markets of a specific year.
Since we have covered our philosophy regarding equities extensively in recent annual reports, a more extended discussion of bond investments may be appropriate for this one, particularly in light of what has happened since year end. An extraordinary amount of money has been lost by the insurance industry in the bond area - notwithstanding the accounting convention that allows insurance companies to carry their bond investments at amortized cost, regardless of impaired market value. Actually, that very accounting convention may have contributed in a major way to the losses; had management been forced to recognize market values,its attention might have been focused much earlier on the dangers of a very long-term bond contract.
Ironically, many insurance companies have decided that a one-year auto policy is inappropriate during a time of inflation, and six-month policies have been brought in as replacements. “How,” say many of the insurance managers, “can we be expected to look forward twelve months and estimate such imponderables as hospital costs, auto parts prices, etc.?” But, having decided that one year is too long a period for which to set a fixed price for insurance in an inflationary world, they then have turned around, taken the proceeds from the sale of that six-month policy, and sold the money at a fixed price for thirty or forty years.
For the last few years our insurance companies have not been a net purchaser of any straight long-term bonds (those without conversion rights or other attributes offering profit possibilities). There have been some purchases in the straight bond area, of course, but they have been offset by sales or maturities. Even prior to this period, we never would buy thirty or forty-year bonds; instead we tried to concentrate in the straight bond area on shorter issues with sinking funds and on issues that seemed relatively undervalued because of bond market inefficiencies.
However, the mild degree of caution that we exercised was an improper response to the world unfolding about us. You do not adequately protect yourself by being half awake while others are sleeping. It was a mistake to buy fifteen-year bonds, and yet we did; we made an even more serious mistake in not selling them (at losses, if necessary) when our present views began to crystallize. (Naturally, those views are much clearer and definite in retrospect; it would be fair for you to ask why we weren’t writing about this subject last year.)
Of course, we must hold significant amounts of bonds or other fixed dollar obligations in conjunction with our insurance operations. In the last several years our net fixed dollar commitments have been limited to the purchase of convertible bonds. We believe that the conversion options obtained, in effect, give that portion of the bond portfolio a far shorter average life than implied by the maturity terms of the issues (i.e., at an appropriate time of our choosing, we can terminate the bond contract by conversion into stock).
This bond policy has given us significantly lower unrealized losses than those experienced by the great majority of property and casualty insurance companies. We also have been helped by our strong preference for equities in recent years that has kept our overall bond segment relatively low. Nevertheless, we are taking our lumps in bonds and feel that, in a sense, our mistakes should be viewed less charitably than the mistakes of those who went about their business unmindful of the developing problems.
Harking back to our textile experience, we should have realized the futility of trying to be very clever (via sinking funds and other special type issues) in an area where the tide was running heavily against us.
We have severe doubts as to whether a very long-term fixed-interest bond, denominated in dollars, remains an appropriate business contract in a world where the value of dollars seems almost certain to shrink by the day. Those dollars, as well as paper creations of other governments, simply may have too many structural weaknesses to appropriately serve as a unit of long term commercial reference. If so, really long bonds may turn out to be obsolete instruments and insurers who have bought those maturities of 2010 or 2020 could have major and continuing problems on their hands. We, likewise, will be unhappy with our fifteen-year bonds and will annually pay a price in terms of earning power that reflects that mistake.
Some of our convertible bonds appear exceptionally attractive to us, and have the same sort of earnings retention factor (applicable to the stock into which they may be converted) that prevails in our conventional equity portfolio. We expect to make money in these bonds (we already have, in a few cases) and have hopes that our profits in this area may offset losses in straight bonds.
And, of course, there is the possibility that our present analysis is much too negative. The chances for very low rates of inflation are not nil. Inflation is man-made; perhaps it can be man-mastered. The threat which alarms us may also alarm legislators and other powerful groups, prompting some appropriate response.
Furthermore, present interest rates incorporate much higher inflation projections than those of a year or two ago. Such rates may prove adequate or more than adequate to protect bond buyers. We even may miss large profits from a major rebound in bond prices. However, our unwillingness to fix a price now for a pound of See’s candy or a yard of Berkshire cloth to be delivered in 2010 or 2020 makes us equally unwilling to buy bonds which set a price on money now for use in those years. Overall, we opt for Polonius (slightly restated): “Neither a short-term borrower nor a long-term lender be.
Prospects
Last year we said that we expected operating earnings in dollars to improve but return on equity to decrease. This turned out to be correct. Our forecast for 1980 is the same. If we are wrong, it will be on the downside. In other words, we are virtually certain that our operating earnings expressed as a percentage of the new equity base of approximately $236 million, valuing securities at cost, will decline from the 18.6% attained in 1979. There is also a fair chance that operating earnings in aggregate dollars will fall short of 1979; the outcome depends partly upon the date of disposition of the bank, partly upon the degree of slippage in insurance underwriting profitability, and partly upon the severity of earnings problems in the savings and loan industry.
We continue to feel very good about our insurance equity investments. Over a period of years, we expect to develop very large and growing amounts of underlying earning power attributable to our fractional ownership of these companies. In most cases they are splendid businesses, splendidly managed, purchased at highly attractive prices.
Your company is run on the principle of centralization of financial decisions at the top (the very top, it might be added), and rather extreme delegation of operating authority to a number of key managers at the individual company or business unit level. We could just field a basketball team with our corporate headquarters group (which utilizes only about 1500 square feet of space).
This approach produces an occasional major mistake that might have been eliminated or minimized through closer operating controls. But it also eliminates large layers of costs and dramatically speeds decision-making. Because everyone has a great deal to do, a very great deal gets done. Most important of all, it enables us to attract and retain some extraordinarily talented individuals - people who simply can’t be hired in the normal course of events - who find working for Berkshire to be almost identical to running their own show.
We have placed much trust in them - and their achievements have far exceeded that trust.
