The Open Curriculum

The Market · Level 2

Poor Charlie’s Almanack

Charlie Munger (ed. Peter D. Kaufman) · 2005

Thesis

Good judgment in business and life comes less from brilliance than from carrying the few big ideas of every basic discipline, checking decisions against the predictable ways human psychology misfires, and acting heavily on the rare occasions when the odds are clearly favorable.

The main ideas

The latticework of mental models

Munger argues that worldly wisdom comes from knowing the big ideas of the basic disciplines, including mathematics, accounting, engineering, physics, biology, psychology and microeconomics, and using them together. Isolated facts do not help; they have to hang on a latticework of models. In his 1994 talk at USC Business School he says that "eighty or ninety important models will carry about 90 percent of the freight in making you a worldly-wise person," and that a mere handful of those carry very heavy freight.

The mechanism is the one in the proverb he cites: to a man with a hammer, every problem looks pretty much like a nail. A specialist who knows one discipline sees every problem through it and misses the causes that belong to other fields. Munger treats this as the main defect of professional training. Because most important problems cross disciplines, the thinker with more models makes fewer forced errors.

His prescription for the models themselves is to learn the elementary version of each thoroughly, including compound interest, basic probability, and the reliable principles of each field, and to use them routinely, not only when a problem looks like it belongs to that field.

Implication: when analyzing a business, a policy or a person, run the problem through the main models of several disciplines before trusting the answer your own specialty produces.

Interests first, then psychology

Munger analyzes a decision in two passes. First, the rational one: what are the factors that really govern the interests involved, rationally considered. Second, the psychological one: what subconscious influences, which are usually useful but often malfunction, are pushing the people involved toward misjudgment, including himself.

His catalog for the second pass is "The Psychology of Human Misjudgment," first given as a talk in 1995 and revised for the book, which lists 25 standard causes of misjudgment. They include reward and punishment superresponse, liking and disliking, doubt avoidance, inconsistency avoidance, envy, reciprocation, social proof, contrast-misreaction and deprival superreaction. Munger's point is not that any one tendency is surprising but that each is predictable, and that a checklist catches what intuition misses.

The 25th is the lollapalooza tendency: extreme consequences come from several tendencies acting together in favor of the same outcome. Munger holds that the most extreme outcomes, in good businesses and in disasters, almost always involve such a combination, so the analyst should look for where several forces align in the same direction.

Implication: when an outcome is far larger than any single cause would explain, a stock mania, a cult, a runaway product, look for several psychological tendencies stacking. When you want to produce a big result yourself, arrange for several to work together.

Incentives

Munger places reward and punishment superresponse first in his list and says that although he has been in the top 5% of his age cohort in understanding the power of incentives, he has underestimated it all his life. People respond to what they are paid for, not to what they are asked for, and they rationalize the result.

His cases: Federal Express could not get its night shift to move packages through the hub quickly enough while it paid workers by the hour. Management then paid them by the shift and let them go home when the work was done, and the problem disappeared. At Xerox, Joe Wilson found that a new, better copier was selling worse than an older, inferior one. The commission arrangement gave salesmen a large incentive to sell the old machine.

The mechanism is double. Incentives change effort directly, and they also bend perception: people come to believe that what pays them is right. Munger extends this to professional advice, where the adviser's compensation shapes the advice sincerely, not just cynically.

Implication: before diagnosing a performance problem as a people problem, look at what the people are paid for. When taking advice, identify how the adviser gets paid and discount accordingly.

Inversion

Many problems are best solved backward. Munger borrows the mathematician Carl Jacobi's instruction to invert, always invert, and the story of a rustic who wished he knew where he was going to die, so that he would never go there. Asking what would guarantee failure is often easier and more complete than asking what would produce success.

His 1986 commencement speech at the Harvard School in Los Angeles applies it to life. He quotes Johnny Carson's prescriptions for guaranteed misery: ingest chemicals to alter mood or perception, envy, and resent. He adds four of his own: be unreliable; learn only from your own personal experience, minimizing what you learn vicariously from the good and bad experience of others; go down and stay down when you suffer your first, second or third severe reverse; and ignore the story of the rustic, meaning never invert.

Inversion works because failure modes are fewer and more concrete than paths to success, and avoiding a known cause of ruin needs no brilliance. Munger's broader claim is that he and Buffett gained more from consistently avoiding stupidity than from seeking brilliance.

Implication: before planning how to make a project, investment or relationship succeed, list what would reliably make it fail, and check the plan against that list first.

The business is the return

Over long periods, a stock's return converges on the return on capital of the underlying business. In the USC talk Munger puts numbers on it: if a business earns 6% on capital over 40 years and you hold it for those 40 years, you will not make much more than 6% even if you bought it at a huge discount. If a business earns 18% on capital over 20 or 30 years, even an expensive-looking purchase price produces an excellent result.

This is why he steered Buffett away from buying poor businesses because they were cheap. The formative case in the book is See's Candies, bought in 1972 for $25 million, which showed them that a business with a strong brand and pricing power was worth paying up for.

The same logic governs who keeps the gains from improvement. In a commodity business, gains from better productivity go to customers, not owners. When Berkshire owned a textile mill, people brought Buffett a new loom that did twice as much work. His answer, as Munger tells it, was that he hoped it did not work, because if it did he would have to close the mill: every competitor would buy the same loom and the savings would pass to the buyers of cloth.

Implication: value a business by what it can earn on reinvested capital over decades, and ask of every productivity gain whether the owner or the customer will keep it.

Few big bets

Munger argues that an investor needs only a few great decisions, made when the odds are clearly favorable, and should otherwise do nothing. He repeats Buffett's image of a punch card with only 20 slots, one for each investment you may make in a lifetime; with so few punches, you would think hard about each and put real money behind it. Of Berkshire, he says the top ten insights account for most of its accumulated billions.

The mechanism combines the latticework and the misjudgment catalog. Most of the time the investor has no edge. Opportunities where the analysis is clear and the price is wrong are rare, and diversifying across many mediocre bets dilutes the few good ones. The discipline is patience; the model he borrows is the racetrack's pari-mutuel system, in which the odds reflect the crowd's betting and the winner is the bettor who waits for a mispriced horse.

Implication: keep a very high bar for new commitments, and when one clears it, size it to matter. Treat activity as a cost, not a sign of diligence.

Solving backward from the target

In "Practical Thought About Practical Thought?" (1996), Munger demonstrates his method on a problem: it is 1884, a man named Glotz offers $2 million for half the equity of a new non-alcoholic beverage business, and the business must be worth $2 trillion by 2034. Munger lists five helpful notions: decide the no-brainer questions first, use numerical fluency, invert, use the elementary academic ideas of several disciplines together, and expect big results only from a combination of factors.

He then works backward from the target. By 2034 there will be about 8 billion people. If the new beverage and its imitators flavor 25% of the water humans drink, and the company takes half of that market, it sells 2.92 trillion eight-ounce servings a year. At a net of four cents a serving, it earns $117 billion, which supports the $2 trillion valuation. The rest of the talk shows how conditioning, social proof and a protected trademark could produce that result, which is what Coca-Cola did.

Implication: for any ambitious goal, fix the end number first, work out the arithmetic it requires, and only then ask which mechanisms could deliver it.

Limits

Related