What this covers

Patrick O'Shaughnessy talks with Carl Kawaja, a longtime Capital Group portfolio manager, starting with his two longest-held positions, TSMC and Vale. Over roughly 22.5 years they compounded at about 18.5 and 18 percent a year, against 7.5 to 8 percent for the All Country World Index. Kawaja explains each business's edge. Vale owns high-grade iron ore with high ferrous content, found in quantity only in Brazil and Western Australia, which China needs to blend with its weaker domestic ore. TSMC wins by serving many customers' specifications at once, like a restaurant with a huge menu, while Intel and Samsung optimized for the most popular variant.

From there the conversation moves to method. Kawaja describes the One Variable Test, the pull toward simplicity, and "echolocation" as a way to work through ambiguity. He cites Bill Miller on productive uncertainty and says he fears omission more than commission. Edward Tufte's one-graph reading of the Challenger O-ring data anchors an argument for visual emphasis in models. Shorter case studies include the MUFG chairman's blunt warning to Hank Paulson and AT&T's iPhone subsidy.

The contested material comes late. Kawaja's "Empire Strikes Back" theme holds that incumbents such as Target, GM, and Inditex can reinvent themselves, and he doubts the top five S&P 500 companies can keep growing from about 22 percent of index earnings toward 75 percent. He also compares autonomous vehicles to the early iPhone. The talk closes on Capital Group's river rafting analogy, Brunello Cucinelli's field of flowers, and his father's rule that there are no difficult decisions, only imperfect information.

Sharpest takeaway

Kawaja argues durable long-term investing comes from understanding a business at the unit-economics level, identifying the one variable that drives its outcome, and tolerating ambiguity and failure, while incumbents with entrenched infrastructure can reinvent themselves and be undervalued against the dominant growth stocks.

  • Most successful investments turn on one determinant variable, and the investor's job is to find it and have a differentiated view on it
  • Holding a position requires a core belief about what makes the business special, which should be revisited over time
  • Incumbent companies such as Target, GM, and Inditex can reinvent themselves, and the FAANG valuation gap may create opportunity

The claims · ranked67 claims · weighted by value

This asset isn't compiled yet

You're seeing its claims, ranked. Compile it to build the argument threads, weight them, and check each claim against your library — the full view.

0.86

Edward Tufte condensed the 13 pages of Morton Thiokol engineering data into one graph that plotted temperature against prior launches, revealing that cold-weather launches were risky, and Kawaja argues investors should similarly distill key variables into a clear visual.

causalhigh valueestablishednovelty 3/4durability 4/4· Carl

“Edward Tufte took those 13 pages of data and he put them in one graph”

0.80

Engineers at Morton Thiokol sent 13 pages of data arguing against the Challenger launch, but Tufte condensed the data into one graph with temperature that showed O-ring failure on cold days, which would have revealed the danger had it been visualized.

causalhigh valueestablishednovelty 2/4durability 4/4· Carl

“Edward Tufte took those 13 pages of data and he put them in one graph”

0.76

Investors retrospectively believe an outcome was obvious, which is a post-facto certainty that hides the real uncertainty they felt when they made the investment, so investors should remember how uncertain they were at the time.

causalhigh valueestablishednovelty 2/4durability 4/4· Carl

“we have this post-facto certainty that we say, 'Gosh, it was so clear to me.'”

0.73

Established incumbent companies can reinvent themselves to respond to disruptors, and the investment opportunity lies in the valuation gap between leaders and incumbents that are catching up on ecosystem, infrastructure, and talent advantages.

forecasthigh valuecontestednovelty 3/4durability 3/4· Carl

“what is the extent to which existing incumbent companies, the empire, can reinvent themselves to respond to the companies that are disrupting them”

0.69

Autonomous vehicles resemble the early iPhone in potential, and if autonomous navigation works for trucks and passenger cars, it could make driving a nostalgic activity over the next 50 years, much as horse ownership shrank from around 20 million horses in 1900 to fewer today.

forecasthigh valuecontestednovelty 3/4durability 2/4· Carl

“maybe roll this forward 50 years from now, driving the car yourself is going to be something that old guys do to indulge their nostalgia”

0.65

Colleague Mark Casey argued that digital platforms can keep growing because the global advertising market is far larger than big consumer-goods ad budgets, since local merchants and millions of new advertisers keep entering the ecosystem.

causalhigh valuecontestednovelty 3/4durability 2/4· Carl

“The advertising market's a lot bigger than Proctor & Gamble and Unilever”

0.65

The top five S&P 500 companies, Apple, Microsoft, Amazon, Facebook, and Google, account for about 22 percent of S&P 500 earnings, and if they grow around 15 to 20 percent a year while the index compounds far slower, they would reach about 75 percent, which Kawaja considers unlikely.

forecasthigh valuecontestednovelty 3/4durability 2/4· Carl

“if these companies do, not only are they 22% of S&P 500 earnings now, but they're going to be 75% of S&P 500 earnings. And that seems unlikely to me.”

0.63

The seaborne-traded iron ore market is effectively controlled by three significant producers in Australia and one major producer in Brazil, which gives the group pricing power in the global market.

factualhigh valueestablishednovelty 2/4durability 2/4· Carl

“there are three significant producers in Australia, there's one significant producer in Brazil and they basically control the global seaborne traded iron ore market”

0.58

In successful investments usually only one variable ends up mattering, so the investor should identify that one thing, and should invest only if that variable is likely to be determinant and the investor has superior, differentiated information about it.

normativehigh valuespeaker onlynovelty 3/4durability 4/4· Carl

“often find in successful investments really just one thing mattered”

0.58

Kawaja worries more about errors of omission, the opportunities missed through insufficient work or dismissal, than about errors of commission, because the investment business rewards swinging for big wins and missed opportunities are the costlier mistakes.

normativehigh valuespeaker onlynovelty 3/4durability 4/4· Carl

“that is the errors of omission that bothered me the most”

0.58

When the chairman of MUFJ warned Hank Paulson that the firm might fail during a capital-raising meeting, Kawaja took that candor, combined with the stock already being down, as a reason he wanted to buy Japanese banks after due diligence, because the extreme pessimism signaled that the situation was not as hopeless as it sounded.

causalhigh valuespeaker onlynovelty 4/4durability 2/4· Carl

“before you make this investment with us, I think I have to warn you that we may fail to succeed as a firm”

0.58

Over roughly 22.5 years, TSMC compounded at about 18.5 percent a year and Vale compounded at about 18 percent a year in dollars, versus about 7.5 to 8 percent for the All Country World Index.

factualhigh valueestablishednovelty 2/4durability 1/4· Carl

“Taiwan semiconductor has compounded at about 18 and a half percent a year”

0.57

Vale's competitive advantage comes from high-grade, high-ferrous-content iron ore that exists in significant quantities only in Brazil's Amazon region and Western Australia, and that is needed to blend with China's lower-quality domestic ore, making Vale's moat more durable than major technology platforms.

causalhigh valuespeaker onlynovelty 3/4durability 3/4· Carl

“iron ore in the quantities and of the grade at which they find it really primarily only exists significantly in two places in the world”

0.57

AT&T's subsidy of iPhones was rational because the phones attracted data-hungry customers whose net present value exceeded the subsidy cost, and the market underestimated this as a phone-subsidy story when it was really a data-revenue story.

causalhigh valuespeaker onlynovelty 3/4durability 3/4· Carl

“The net present value of acquiring those customers is worth much more than the subsidy of getting them.”

0.57

Bill Miller said that feeling terrible about a portfolio, with some holdings that had run up and others that were cheap but deteriorating, was a sign of productive uncertainty, and that he did well when he felt this miserable state of uncertainty.

normativehigh valuespeaker onlynovelty 3/4durability 3/4· Carl

“That miserable state of uncertainty is kind of a good place to be.”

0.57

Kawaja suspects Vale's competitive advantage in iron ore may be more durable than the advantages of popular internet platforms like Google, Amazon, and Facebook, because steel has been essential since the Iron Age and no substitute has displaced it.

forecasthigh valuespeaker onlynovelty 3/4durability 3/4· Carl Kawaja

“I actually suspect that Vale's competitive advantage is more durable than any of those companies”

0.57

TSMC's durable advantage is not only doing extraordinarily hard semiconductor manufacturing but also satisfying many different customers' specifications at once, like a restaurant with a huge menu, while Intel and Samsung optimized for only the most popular variant and lost share when customers needed alternatives.

causalhigh valuespeaker onlynovelty 3/4durability 3/4· Carl

“TSMC is actually able to do something so difficult as that, but also at the same time, to satisfy the needs of multiple customers”

0.57

The key market theme is whether existing incumbent companies, the empire, can reinvent themselves to respond to disruptors, and some established retailers and automakers are reaching tipping points where store networks, fulfillment, and electrification allow them to compete.

normativehigh valuespeaker onlynovelty 3/4durability 3/4· Carl

“what is the extent to which existing incumbent companies, the empire, can reinvent themselves to respond to the companies that are disrupting them”

0.57

Capital Group resembles a river rafting trip: new hires drop in at random points, may hit rapids, and rely on experienced colleagues to pull them back into the boat, and they will eventually reach clear water and make their reputation if the firm has a long time horizon.

causalhigh valuespeaker onlynovelty 3/4durability 3/4· Carl

“sooner or later you will hit clear water. Like you will hit a long smooth stretch of clear water and you will make your bones.”

0.57

Brunello Cucinelli bought the neighboring logistics facility next to his Solomeo factory for about $60 million and demolished it to plant a field of grass and flowers, reasoning that people who make beautiful clothes need to see something beautiful.

factualhigh valuespeaker onlynovelty 3/4durability 3/4· Carl

“how can I ask people to make beautiful clothes when they don't look at something beautiful that inspires them?”

0.56

Lessons from non-investment reading such as Walt Whitman and Rilke have more staying power than investment formulas, because in uncertain markets, when a holding falls, it is a fundamental wisdom rather than a DCF model that keeps an investor committed.

normativehigh valuespeaker onlynovelty 2/4durability 4/4· Carl

“it's not your DCF model that's going to keep you in it”

0.56

Great leaders teach people to want certain things, such as being ethical, hardworking, and creative, and then let them figure out the path, similar to Saint-Exupery's idea of teaching people to long for the endless sea rather than assigning them tasks.

normativehigh valuespeaker onlynovelty 2/4durability 4/4· Carl

“teach them to long for the endless immensity of the sea”

0.56

Edward Tufte's The Visual Display of Quantitative Information is the most practically useful book for an investing career because it teaches critical thinking about variables and how to present ideas, and analysts should use visual emphasis in models so that key numbers such as gross margin stand out.

normativehigh valuespeaker onlynovelty 2/4durability 4/4· Carl

“The Visual Display of Quantitative Information, has a dry title, is the single probably most practically useful book you can read to distinguish yourself in this job”

0.55

Kawaja's father held that there are no difficult decisions, only decisions with imperfect information, so whenever an investor cannot decide, the investor should force themselves to gather more information.

normativehigh valuespeaker onlynovelty 3/4durability 4/4· Carl

“there are no decisions that are difficult to make”

0.54

When a CEO the investor liked leaves for another job or retires, the investor should generally sell, because the executive has effectively written a letter saying to sell and may know something the investor does not.

normativehigh valuespeaker onlynovelty 3/4durability 3/4· Carl

“when a CEO has a good job in his day, a good job at a firm that he likes and he leaves, he has written you a letter saying to sell the stock”

0.54

Capital Group was founded in the 1930s by John Lovelace, who moved money out of the market in 1929 after sensing froth, relocated to Los Angeles, and ran the firm for 26 years without making money, a long-horizon tradition that still shapes its management.

factualestablishednovelty 2/4durability 4/4· Carl

“he ran the firm for 26 years without making money”

0.53

A pediatric neurologist, Dr. Bill Stennis, showed belief in a young Kawaja who was facing lifelong seizure medication by delivering a gift and a reassurance that he would have a normal life, and Kawaja concludes that the kindest thing someone can do is believe in you.

normativehigh valuespeaker onlynovelty 2/4durability 4/4· Carl

“Sometimes the kindest thing someone can do for you is to believe in you.”

0.53

To hold a stock for a long period one needs a core belief that what the business does is special, reminded periodically, and if that changes and the business goes away or the investor was wrong, the investor should sell, rather than selling for timing reasons.

normativehigh valuespeaker onlynovelty 2/4durability 4/4· Carl

“it's kind of tough to hang on to something unless you have some core belief that what they do is special”

0.53

Kawaja's errors of omission usually came from not doing enough work and not being open-minded, as with his dismissal of Tesla in 2000 when students' convincing short report and their desire to own the product signaled something powerful that he failed to investigate.

causalhigh valuespeaker onlynovelty 2/4durability 4/4· Carl

“I was too lazy and too closed minded to do more work myself”

0.52

Kawaja uses echolocation by voicing investment ideas inside the firm and listening for the feedback, which lets him refine ideas; strong pushback has saved him from mistakes, while ideas that get no response may be underexplored.

normativehigh valuespeaker onlynovelty 2/4durability 3/4· Carl

“I'm like finding my way through the cave”

0.52

Many candidates from leading business schools have not made mistakes or experienced much failure, and those who cannot process failure should choose a different job, because investing requires tolerance for being wrong frequently.

normativehigh valuespeaker onlynovelty 2/4durability 3/4· Carl

“These candidates, they haven't made a mistake, their whole lives and not much has gone wrong.”

0.52

Analyst earnings models usually render every number in the same font, so a model with about 520 numbers hides which figures matter; varying font size or weight, such as making a gross margin shift from 42% to 67% bigger, would direct attention.

normativehigh valuespeaker onlynovelty 2/4durability 3/4· Carl

“every number is the same font. If the gross margin's going to go from 42% to 67%, make the 67 bigger than the 42”

0.52

At Google's 2004 IPO, Mark Casey's model assumed operating margin compression, but Kawaja argued Google would need to hire every computer PhD graduating from U.S. universities over five years to achieve the forecast expense growth, so he built a model twice as bullish.

causalhigh valuespeaker onlynovelty 3/4durability 3/4· Carl

“They literally need to hire every computer PhD who's graduated in the United States for the last five years”

0.51

Google and Facebook were about 10 percent of world ad spend, which was around $2.5 trillion, and the ad market was far larger than the budgets of consumer-goods giants such as Procter & Gamble, Unilever, Coke, and Nike, because millions of small merchants had not yet advertised online.

factualhigh valuespeaker onlynovelty 3/4durability 2/4· Carl

“The advertising market's a lot bigger than Proctor & Gamble and Unilever”

0.50

Around 2007 and 2008, an analyst argued that iPhone buyers would keep their phones and fill them with photos, which contradicted the then-common view that phones were replaced frequently, and this unit-level insight supported Kawaja's investment in Apple.

causalhigh valuespeaker onlynovelty 2/4durability 3/4· Carl

“no one was going to give up their iPhones”

0.50

Capital Group tolerates analysts whose early results were poor, sometimes waiting five to seven years before judging them, and a loss early in a career can be a useful mind-focuser, because a long horizon gives people time to figure out the business.

normativehigh valuespeaker onlynovelty 2/4durability 3/4· Carl

“we've waited five years, six years, seven years to sort of see whether they were, in fact, bad”

0.50

Investors who have profited from a stock begin to love it and convince themselves it has a huge moat and an excellent CEO, which blinds them, so maintaining some uncertainty is necessary even for investments that have worked.

causalhigh valuespeaker onlynovelty 2/4durability 3/4· Carl

“you start really loving the things that have done well and convincing yourself that they have a huge mote”

0.49

General Motors under CEO Mary Barra is reinventing itself by investing heavily in electric vehicles, including electric trucks and an electric Hummer, and by pursuing autonomous driving through Cruise, which Kawaja views as an Empire Strikes Back example.

forecasthigh valuespeaker onlynovelty 2/4durability 2/4· Carl

“Invest incredibly heavily push all the poker chips on the table behind being successful on electric vehicles”

0.48

Buffett's horse analogy shows that technological change can shrink a once-large category, from roughly 20 million horses to about 6 million, so not all incumbents will survive disruption.

factualestablishednovelty 1/4durability 4/4· Carl Kawaja

“there were 20 million horses in America in 1900”

0.47

Bunge, a soybean crusher, can be understood by comparing the cost of a 60-pound bushel of soybeans (around $12 to $13) with revenue from roughly 44 pounds of meal at about 23 cents a pound and 11 pounds of oil at 60 to 70 cents a pound, and when oil fell from 70 to 15 cents, the impact on profits became immediately computable.

factualhigh valuespeaker onlynovelty 2/4durability 3/4· Carl

“a bushel of soybeans weigh 60 pounds, and when you crush it, it produces 44 pounds of soybean meal and 11 pounds of soybean oil”

0.47

Kawaja's father taught him that if a restaurant is struggling, its prices are too low, because higher prices with fewer customers generate more profit, which Kawaja uses as a basic lens on pricing power.

normativehigh valuespeaker onlynovelty 2/4durability 3/4· Carl

“If restaurants are at fault it's because your prices are too low.”

0.44

A single insight about a product or the math can reveal that the total investable market is roughly ten times larger than others assume, which is the 'golden needle' that justifies the research effort.

causalhigh valuespeaker onlynovelty 2/4durability 2/4· Carl

“you realize that the total investible market is really 10 times bigger than you thought”

0.44

Apparel customers like to try on and return products, so Inditex's physical store presence and its fitting experience give it an advantage in the shift to online retail.

causalhigh valuespeaker onlynovelty 2/4durability 2/4· Carl

“a lot of people like to go into the store and they like to try on the product”

0.36

Amazon, which looks like a virtual e-commerce company, was buying large amounts of steel for its headquarters construction, showing that even new-economy firms depend on physical materials.

factualestablishednovelty 2/4durability 2/4· Carl

“guess what they're buying a ton of?”

0.36

The Chinese government has spent tens of billions of dollars on semiconductor manufacturing without much success, showing that manufacturing capability at TSMC's level is hard to replicate with capital alone.

factualestablishednovelty 2/4durability 2/4· Carl

“The Chinese government has thrown tens of billions into semiconductor manufacturing without much success.”

0.35

Kawaja performs better as an investor in businesses he understands, and he tries to explain to himself a few key metrics that drive each company, because sometimes he discovers after the fact that he did not understand a business.

normativeestablishednovelty 1/4durability 3/4· Carl

“I find I do better as an investor in businesses that I understand.”

0.35

Smartphones became primarily data devices as voice became a commodity, with users streaming sports and video, which made data-heavy usage the economic core of the phone business.

factualestablishednovelty 1/4durability 3/4· Carl

“Phones became all about data and voice became a commodity”

0.35

The earlier generation of antiseizure medication Kawaja took as a child dulled reaction times, had a sedating effect, and made him feel ostracized, whereas modern seizure treatment has improved greatly over the last 50 years.

factualestablishednovelty 1/4durability 3/4· Carl

“I was on an earlier generation of the medicine that kind of dulled your reaction times and had a very sedating effect”

0.34

Annual reports written by CEOs in their own words are important because employees read them and decide who they want to be, so Kawaja values CEO letters that shape organizational identity.

normativespeaker onlynovelty 2/4durability 3/4· Carl

“they read it and they're like, 'Oh wait, that's who I want to be.'”

0.33

George Saunders's reflection suggests that the regrets that matter most are the times one had the chance to be kind and did not, which Kawaja takes as a reason to practice kindness more.

normativespeaker onlynovelty 1/4durability 4/4· Carl

“What he regrets are opportunities in his life that he had to be kind when he didn't exercise it.”

0.32

Luxury brands such as Chanel, Louis Vuitton, Prada, and Hermes derive value from heritage, and Brunello Cucinelli created a luxury brand from scratch through vision, which Kawaja considers remarkable.

causalspeaker onlynovelty 2/4durability 3/4· Carl

“no one had Brunello Cucinelli bags 50 years ago, but he has, out of whole cloth, created this incredible luxury brand”

0.32

Luxury purchases are driven partly by heritage narratives, such as Katharine Hepburn carrying a Louis Vuitton bag, so newer brands must build cachet through vision rather than history.

causalspeaker onlynovelty 2/4durability 3/4· Carl Kawaja

“one of the reasons someone spends $2,500 on a Louis Vuitton bag is that Katherine Hepburn had one”

0.31

Target partnered with Amazon to sell online, and Kawaja argues Target should have built its own online capabilities earlier, but it is now adjusting by using stores for fulfillment and online sales.

causalspeaker onlynovelty 2/4durability 2/4· Carl Kawaja

“I think Target should have been making those investments themselves early on”

0.31

Kawaja's colleague Galen Hoskin battled aggressive prostate cancer and died from it, and Kawaja considers the time spent helping him one of the great gifts of his life.

factualspeaker onlynovelty 1/4durability 4/4· Carl

“a dear Capital colleague, a friend named Galen Hoskin who battled prostate cancer”

0.31

Rilke's line that perhaps all dragons are princesses waiting to see us act with beauty and courage encourages viewing problems as opportunities, although some problems are simply problems.

normativespeaker onlynovelty 1/4durability 4/4· Carl

“perhaps all the dragons in our lives are princesses who are only waiting to see us just once, act with beauty and courage”

0.30

Carl joined Capital Group in 1991 despite a competing offer that would have paid more than twice as much, and he stayed for 30 years because the firm's culture fit him.

factualspeaker onlynovelty 1/4durability 3/4· Carl

“the other place that I could've gone to was going to pay me more than twice as much as I made for going to Capital”

0.28

Optimizing each decision by searching for the best deal, like constantly re-shopping rentals, produces short-term gratification that resembles social media and can substitute for the more rewarding long-term choice of buying a house and enjoying it over years.

normativespeaker onlynovelty 2/4durability 2/4· Carl

“it's like short-term gratification, it's like social media or something”

0.27

Investors should bring the personal interests and traits that make them distinctive into their investment process, since this individuality creates an edge.

normativespeaker onlynovelty 1/4durability 3/4· Carl

“you need to bring the things that make Patrick O'Shaughnessy, Patrick O'Shaughnessy, to your investing process because that makes you distinctive”

0.27

Steel has been used since the iron age and makes tall buildings, cars, planes, and cities possible, and Kawaja believes nothing is likely to displace it, so the odds are very good that steel demand endures.

forecastspeaker onlynovelty 1/4durability 3/4· Carl

“It's been around a long time and nothing's come along to displace it so I think the odds are very good that nothing will.”

0.26

Whether buying well or holding well is harder depends on personality, and Kawaja finds holding harder because his bargain-hunting temperament inherited from his father pulls him toward constant switching, while his mother's stubbornness helps him hold.

normativespeaker onlynovelty 1/4durability 2/4· Carl

“probably holding onto things is harder because my nature is a bit reflective of my dad”

0.26

Kawaja's cancer diagnosis, which he viewed as a dragon, produced research relationships, a hospital board seat, and oncology investments that became valuable, so adversity can open unexpected opportunities.

factualspeaker onlynovelty 1/4durability 2/4· Carl

“it's been a princess for me or a prince depending on your gender”

0.26

A former Capital colleague, arguably the best investor Kawaja saw there, joked he had more one-way bungee jumps than anyone, signaling that top investors must tolerate doing terribly for periods.

factualspeaker onlynovelty 2/4durability 2/4· Carl

“he'd had more one-way bungee jumps than anyone else at Capital”

0.26

Kawaja frames the question of whether to bet on the top five dominant stocks, which he calls the Lakers, or the rest of the index, which he calls the field, and notes that stars can be injured while the field often wins.

normativespeaker onlynovelty 2/4durability 2/4· Carl Kawaja

“would you take the other 495? Would you take the field or would you take the Lakers?”

0.25

Of roughly 2,000 working hours per year, Kawaja estimates only two or three are genuinely productive insight moments, and the remaining time is spent searching for those moments through meetings that often leave him with no conviction.

factualspeaker onlynovelty 1/4durability 3/4· Carl

“only like two or three hours were the good ones, the other 1,997 hours were looking around for the good hours”

0.19

At an iron ore price of around $220 per ton, shipping ore from Brazil to China is still affordable relative to the value of the ore, which had been around $60 per ton not long ago.

factualspeaker onlynovelty 2/4durability 0/4· Carl

“Currently, the iron ore price is around $220 a ton. Not that long ago, it was $60 a ton”

0.18

Kawaja liked merchant power generators such as Calpine for a time, when the thesis was right, and then the stock went terribly wrong, illustrating that some investment ideas are totally wrong.

factualspeaker onlynovelty 1/4durability 1/4· Carl

“For a while, I really like merchant power generators and Calpine”

0.17

Carl switched to an electric car partly because the time spent at gas stations felt wasteful, even though he optimizes small time savings, and he is not going back to combustion engines.

factualspeaker onlynovelty 0/4durability 2/4· Carl

“I'm never going back to combustion engine car again”