副董事长的想法——过去与未来（芒格，50 周年特别信，2014）

Vice Chairman’s Thoughts – Past and Future

To the shareholders of Berkshire Hathaway Inc.:

I closely watched the 50-year history of Berkshire’s uncommon success under Warren Buffett. And it now seems appropriate that I independently supplement whatever celebratory comment comes from him. I will try to do five things. (1) Describe the management system and policies that caused a small and unfixably-doomed commodity textile business to morph into the mighty Berkshire that now exists, (2) Explain how the management system and policies came into being, (3) Explain, to some extent, why Berkshire did so well, (4) Predict whether abnormally good results would continue if Buffett were soon to depart, and (5) Consider whether Berkshire’s great results over the last 50 years have implications that may prove useful elsewhere. The management system and policies of Berkshire under Buffett (herein together called “the Berkshire system”)

were fixed early and are described below:

(1) Berkshire would be a diffuse conglomerate, averse only to activities about which it could not make useful predictions. (2) Its top company would do almost all business through separately incorporated subsidiaries whose CEOs would operate with very extreme autonomy. (3) There would be almost nothing at conglomerate headquarters except a tiny office suite containing a Chairman, a CFO, and a few assistants who mostly helped the CFO with auditing, internal control, etc. (4) Berkshire subsidiaries would always prominently include casualty insurers. Those insurers as a group would be expected to produce, in due course, dependable underwriting gains while also producing substantial “float” (from unpaid insurance liabilities) for investment. (5) There would be no significant system-wide personnel system, stock option system, other incentive system, retirement system, or the like, because the subsidiaries would have their own systems, often different. (6) Berkshire’s Chairman would reserve only a few activities for himself. (i)

He would manage almost all security investments, with these normally residing in Berkshire’s casualty insurers.

(ii)

He would choose all CEOs of important subsidiaries, and he would fix their compensation and obtain from each a private recommendation for a successor in case one was suddenly needed.

(iii)

He would deploy most cash not needed in subsidiaries after they had increased their competitive advantage, with the ideal deployment being the use of that cash to acquire new subsidiaries.

(iv)

He would make himself promptly available for almost any contact wanted by any subsidiary’s CEO, and he would require almost no additional contact.

(v)

He would write a long, logical, and useful letter for inclusion in his annual report, designed as he would wish it to be if he were only a passive shareholder, and he would be available for hours of answering questions at annual shareholders’ meetings.

(vi)

He would try to be an exemplar in a culture that would work well for customers, shareholders, and other incumbents for a long time, both before and after his departure.

(vii)

His first priority would be reservation of much time for quiet reading and thinking, particularly that which might advance his determined learning, no matter how old he became; and

(viii)

He would also spend much time in enthusiastically admiring what others were accomplishing.

(7) New subsidiaries would usually be bought with cash, not newly issued stock. (8) Berkshire would not pay dividends so long as more than one dollar of market value for shareholders was being created by each dollar of retained earnings. (9) In buying a new subsidiary, Berkshire would seek to pay a fair price for a good business that the Chairman could pretty well understand. Berkshire would also want a good CEO in place, one expected to remain for a long time and to manage well without need for help from headquarters. (10) In choosing CEOs of subsidiaries, Berkshire would try to secure trustworthiness, skill, energy, and love for the business and circumstances the CEO was in. (11) As an important matter of preferred conduct, Berkshire would almost never sell a subsidiary. (12) Berkshire would almost never transfer a subsidiary’s CEO to another unrelated subsidiary. (13) Berkshire would never force the CEO of a subsidiary to retire on account of mere age. (14) Berkshire would have little debt outstanding as it tried to maintain (i) virtually perfect creditworthiness under all conditions and (ii) easy availability of cash and credit for deployment in times presenting unusual opportunities. (15) Berkshire would always be user-friendly to a prospective seller of a large business. An offer of such a business would get prompt attention. No one but the Chairman and one or two others at Berkshire would ever know about the offer if it did not lead to a transaction. And they would never tell outsiders about it. Both the elements of the Berkshire system and their collected size are quite unusual. No other large corporation I know of has half of such elements in place. How did Berkshire happen to get a corporate personality so different from the norm? Well, Buffett, even when only 34 years old, controlled about 45% of Berkshire’s shares and was completely trusted by all the other big shareholders. He could install whatever system he wanted. And he did so, creating the Berkshire system. Almost every element was chosen because Buffett believed that, under him, it would help maximize Berkshire’s achievement. He was not trying to create a one-type-fits-all system for other corporations. Indeed, Berkshire’s subsidiaries were not required to use the Berkshire system in their own operations. And some flourished while using different systems. What was Buffett aiming at as he designed the Berkshire system?

Well, over the years I diagnosed several important themes:

(1) He particularly wanted continuous maximization of the rationality, skills, and devotion of the most important people in the system, starting with himself. (2) He wanted win/win results everywhere--in gaining loyalty by giving it, for instance. (3) He wanted decisions that maximized long-term results, seeking these from decision makers who usually stayed long enough in place to bear the consequences of decisions. (4) He wanted to minimize the bad effects that would almost inevitably come from a large bureaucracy at headquarters. (5) He wanted to personally contribute, like Professor Ben Graham, to the spread of wisdom attained. When Buffett developed the Berkshire system, did he foresee all the benefits that followed? No. Buffett stumbled into some benefits through practice evolution. But, when he saw useful consequences, he strengthened their causes.

Why did Berkshire under Buffett do so well?

Only four large factors occur to me:

The constructive peculiarities of Buffett, The constructive peculiarities of the Berkshire system, Good luck, and The weirdly intense, contagious devotion of some shareholders and other admirers, including some in the press.

I believe all four factors were present and helpful. But the heavy freight was carried by the constructive peculiarities, the weird devotion, and their interactions. In particular, Buffett’s decision to limit his activities to a few kinds and to maximize his attention to them, and to keep doing so for 50 years, was a lollapalooza. Buffett succeeded for the same reason Roger Federer became good at tennis. Buffett was, in effect, using the winning method of the famous basketball coach, John Wooden, who won most regularly after he had learned to assign virtually all playing time to his seven best players. That way, opponents always faced his best players, instead of his second best. And, with the extra playing time, the best players improved more than was normal. And Buffett much out-Woodened Wooden, because in his case the exercise of skill was concentrated in one person, not seven, and his skill improved and improved as he got older and older during 50 years, instead of deteriorating like the skill of a basketball player does. Moreover, by concentrating so much power and authority in the often-long-serving CEOs of important subsidiaries, Buffett was also creating strong Wooden-type effects there. And such effects enhanced the skills of the CEOs and the achievements of the subsidiaries. Then, as the Berkshire system bestowed much-desired autonomy on many subsidiaries and their CEOs, and Berkshire became successful and well known, these outcomes attracted both more and better subsidiaries into Berkshire, and better CEOs as well. And the better subsidiaries and CEOs then required less attention from headquarters, creating what is often called a “virtuous circle.” How well did it work out for Berkshire to always include casualty insurers as important subsidiaries? Marvelously well. Berkshire’s ambitions were unreasonably extreme and, even so, it got what it wanted. Casualty insurers often invest in common stocks with a value amounting roughly to their shareholders’ equity, as did Berkshire’s insurance subsidiaries. And the S&P 500 Index produced about 10% per annum, pre-tax, during the last 50 years, creating a significant tailwind. And, in the early decades of the Buffett era, common stocks within Berkshire’s insurance subsidiaries greatly outperformed the index, exactly as Buffett expected. And, later, when both the large size of Berkshire’s stockholdings and income tax considerations caused the index-beating part of returns to fade to insignificance (perhaps not forever), other and better advantage came. Ajit Jain created out of nothing an immense reinsurance business that produced both a huge “float” and a large underwriting gain. And all of GEICO came into Berkshire, followed by a quadrupling of GEICO’s market share. And the rest of Berkshire’s insurance operations hugely improved, largely by dint of reputational advantage, underwriting discipline, finding and staying within good niches, and recruiting and holding outstanding people.

Then, later, as Berkshire’s nearly unique and quite dependable corporate personality and large size became well known, its insurance subsidiaries got and seized many attractive opportunities, not available to others, to buy privately issued securities. Most of these securities had fixed maturities and produced outstanding results. Berkshire’s marvelous outcome in insurance was not a natural result. Ordinarily, a casualty insurance business is a producer of mediocre results, even when very well managed. And such results are of little use. Berkshire’s better outcome was so astoundingly large that I believe that Buffett would now fail to recreate it if he returned to a small base while retaining his smarts and regaining his youth. Did Berkshire suffer from being a diffuse conglomerate? No, its opportunities were usefully enlarged by a widened area for operation. And bad effects, common elsewhere, were prevented by Buffett’s skills. Why did Berkshire prefer to buy companies with cash, instead of its own stock? Well, it was hard to get anything in exchange for Berkshire stock that was as valuable as what was given up. Why did Berkshire’s acquisition of companies outside the insurance business work out so well for Berkshire shareholders when the normal result in such acquisitions is bad for shareholders of the acquirer? Well, Berkshire, by design, had methodological advantages to supplement its better opportunities. It never had the equivalent of a “department of acquisitions” under pressure to buy. And it never relied on advice from “helpers” sure to be prejudiced in favor of transactions. And Buffett held self-delusion at bay as he underclaimed expertise while he knew better than most corporate executives what worked and what didn’t in business, aided by his long experience as a passive investor. And, finally, even when Berkshire was getting much better opportunities than most others, Buffett often displayed almost inhuman patience and seldom bought. For instance, during his first ten years in control of Berkshire, Buffett saw one business (textiles) move close to death and two new businesses come in, for a net gain of one. What were the big mistakes made by Berkshire under Buffett? Well, while mistakes of commission were common, almost all huge errors were in not making a purchase, including not purchasing Walmart stock when that was sure to work out enormously well. The errors of omission were of much importance. Berkshire’s net worth would now be at least $50 billion higher if it had seized several opportunities it was not quite smart enough to recognize as virtually sure things. The next to last task on my list was: Predict whether abnormally good results would continue at Berkshire if Buffett were soon to depart. The answer is yes. Berkshire has in place in its subsidiaries much business momentum grounded in much durable competitive advantage. Moreover, its railroad and utility subsidiaries now provide much desirable opportunity to invest large sums in new fixed assets. And many subsidiaries are now engaged in making wise “bolt-on” acquisitions. Provided that most of the Berkshire system remains in place, the combined momentum and opportunity now present is so great that Berkshire would almost surely remain a better-than-normal company for a very long time even if (1) Buffett left tomorrow, (2) his successors were persons of only moderate ability, and (3) Berkshire never again purchased a large business. But, under this Buffett-soon-leaves assumption, his successors would not be “of only moderate ability.” For instance, Ajit Jain and Greg Abel are proven performers who would probably be under-described as “world-class.” “World-leading” would be the description I would choose. In some important ways, each is a better business executive than Buffett. And I believe neither Jain nor Abel would (1) leave Berkshire, no matter what someone else offered or (2) desire much change in the Berkshire system.

Nor do I think that desirable purchases of new businesses would end with Buffett’s departure. With Berkshire now so large and the age of activism upon us, I think some desirable acquisition opportunities will come and that Berkshire’s $60 billion in cash will constructively decrease. My final task was to consider whether Berkshire’s great results over the last 50 years have implications that may prove useful elsewhere. The answer is plainly yes. In its early Buffett years, Berkshire had a big task ahead: turning a tiny stash into a large and useful company. And it solved that problem by avoiding bureaucracy and relying much on one thoughtful leader for a long, long time as he kept improving and brought in more people like himself. Compare this to a typical big-corporation system with much bureaucracy at headquarters and a long succession of CEOs who come in at about age 59, pause little thereafter for quiet thought, and are soon forced out by a fixed retirement age. I believe that versions of the Berkshire system should be tried more often elsewhere and that the worst attributes of bureaucracy should much more often be treated like the cancers they so much resemble. A good example of bureaucracy fixing was created by George Marshall when he helped win World War II by getting from Congress the right to ignore seniority in choosing generals. Sincerely, Charles T. Munger