Curated Insights 2018.09.14

Risk, uncertainty and ignorance in investing and business – Lessons from Richard Zeckhauser

People feel that 50% is magical and they don’t like to do things where they don’t have 50% odds. I know that is not a good idea, so I am willing to make some bets where you say it is 20% likely to work but you get a big pay-off if it works, and only has a small cost if it does not. I will take that gamble. Most successful investments in new companies are where the odds are against you but, if you succeed, you will succeed in a big way.” “David Ricardo made a fortune buying bonds from the British government four days in advance of the Battle of Waterloo. He was not a military analyst, and even if he were, he had no basis to compute the odds of Napoleon’s defeat or victory, or hard-to-identify ambiguous outcomes. Thus, he was investing in the unknown and the unknowable. Still, he knew that competition was thin, that the seller was eager, and that his windfall pounds should Napoleon lose would be worth much more than the pounds he’d lose should Napoleon win. Ricardo knew a good bet when he saw it.

…in any probabilistic exercise: the frequency of correctness does not matter; it is the magnitude of correctness that matters…. even though Ruth struck out a lot, he was one of baseball’s greatest hitters…. Internalizing this lesson, on the other hand, is difficult because it runs against human nature in a very fundamental way… The Babe Ruth effect is hard to internalize because people are generally predisposed to avoid losses. …What is interesting and perhaps surprising is that the great funds lose money more often than good funds do. The best VCs funds truly do exemplify the Babe Ruth effect: they swing hard, and either hit big or miss big. You can’t have grand slams without a lot of strikeouts.

Risk, which is a situation where probabilities are well defined, is much less important than uncertainty. Casinos, which rely on dice, cards and mechanical devices, and insurance companies, blessed with vast stockpiles of data, have good reason to think about risk. But most of us have to worry about risk only if we are foolish enough to dally at those casinos or to buy lottery cards….” “Uncertainty, not risk, is the difficulty regularly before us. That is, we can identify the states of the world, but not their probabilities.” “We should now understand that many phenomena that were often defined as involving risk – notably those in the financial sphere before 2008 – actually involve uncertainty.” “Ignorance arises in a situation where some potential states of the world cannot be identified. Ignorance is an important phenomenon, I would argue, ranking alongside uncertainty and above risk. Ignorance achieves its importance, not only by being widespread, but also by involving outcomes of great consequence.” “There is no way that one can sensibly assign probabilities to the unknown states of the world. Just as traditional finance theory hits the wall when it encounters uncertainty, modern decision theory hits the wall when addressing the world of ignorance.


Hank Paulson says the financial crisis could have been ‘much worse’

While Bear Stearns’ failure in normal markets would not hurt the U.S. economy, we believed that the system was too fragile and fear-driven to take a Bear Stearns bankruptcy. To those who argue that Bear Stearns created moral hazard and contributed to the Lehman failure, I believe just the opposite—that it allowed us to dodge a bullet and avoid a devastating chain reaction.

If Bear had failed, the hedge funds would have turned on Lehman with a vengeance. Lehman would have failed almost immediately and the result would have been much worse than Lehman’s September failure, which occurred after we had stabilized Fannie Mae and Freddie Mac and Bank of Americaacquired Merrill Lynch. I would hate to imagine what would have happened if this whole thing started before we’d stabilized Fannie and Freddie.

An interview with Tim Geithner on this topic was done recently at the Yale School of Management and he speaks much more authoritatively on the limits of the Fed powers than I, but here goes. While our responses may have looked inconsistent, Ben, Tim, and I were united in our commitment to prevent the failure of any systemically important financial institution. But we had a balkanized, outdated regulatory system without sufficient oversight or visibility into a large part of the modern financial system and without the necessary emergency powers to inject capital, guarantee liabilities, or wind down a non-banking institution. So we did whatever we could on a case-by-case basis.

For Lehman, we had no buyer and we needed one with the willingness and capacity to guarantee its liabilities. Without one, a permissible Fed loan would not have been sufficient or effective to stop a run. To do that, the Fed would have had to inject capital or guarantee liabilities and they had no power to do so. Now, here’s the point that I think a lot of people miss: In the midst of a panic, market participants make their own judgments and a Fed loan to meet a liquidity shortfall wouldn’t prevent a failure if they believed Lehman wasn’t viable or solvent. And no one believed they were.

AIG is a cautionary tale. We should not have let our financial regulatory system fail to keep up with modern financial markets. No single regulator had oversight visibility or adequate powers to deal with AIG. Its insurance companies were regulated at the state level, its holding company was like a giant hedge fund sitting on top of the insurance companies, and it was regulated by the ineffective Office of Thrift Supervision, which also regulated—get this—Countrywide, WaMu, IndyMac, GE Capital. They all selected their regulator. So you get the picture, it’s regulatory arbitrage.

And I’m concerned that some of the tools we effectively used to stave off disaster have now been eliminated by Congress. These include the ability of Treasury to use its exchange stabilization fund to guarantee the money market funds, the emergency lending authority the Fed used to avoid the failure of Bear and AIG, and the FDIC’s guarantee of bank liabilities on a systemwide basis, which was critical.

The global smartphone supply chain needs an upgrade

At the peak in October 2017, smartphone components accounted for over 33% of exports from Taiwan, 17% of those from Malaysia and 16% from Singapore. Smartphones comprise 6% of Chinese exports. Memory chips flow from South Korea and Vietnam; system chips from Malaysia, Taiwan and elsewhere; and displays from Japan and South Korea. Rich-world firms, such as Qualcomm, sell licences to use their intellectual property (IP). The parts are then assembled, mainly by armies of Chinese workers.

Apple and 13 of its chip suppliers earn over 90% of the total pool of profits from the Apple system. Meanwhile the tail of other firms doing more basic activities must pay for most workers, inventories and fixed assets (see chart). So they have in aggregate a weak return on equity, of 9%, and a net profit margin of just 2%. Their earnings have not risen for five years. They include assemblers such as Taiwan’s Hon Hai and niche component makers, some of which are visibly struggling. On August 22nd AAC Technologies, a specialist in making phones vibrate, said its second-quarter profits fell by 39% compared with the previous year.

Apple, Samsung and most semiconductor makers could ride out such tensions, with their high margins and cash-laden balance-sheets. But the long chain of other suppliers could not, given their razor-thin margins, big working-capital balances and fixed costs. Tariffs could push them into the red. Of the 132 firms, 52% would be loss-making if costs rose by just 5%. And a ZTE-style cessation of trade would be disastrous. If revenues dried up and the 132 firms continued to pay their own suppliers, short-term debts and wages, 28% of them would run out of cash within 100 days.

If you are running a big firm in the smartphone complex, you should be reimagining things in preparation for a less open world. In a decade, on its current trajectory, the industry will be smaller, with suppliers forced to consolidate and to automate production. It may also be organised in national silos, with production, IP, profits and jobs distributed more evenly around the world. Firms will need to adapt—or be swiped away.

The story of Box: A unicorn’s journey to public success

The early days of Box’s selling file sharing and collaboration have largely been replaced by big corporate wins. One measure of Box’s success is its penetration of the Fortune 500—from 52% in the second quarter of 2016 to 69% in the same quarter of fiscal 2019. About 58% of Box’s total revenue comes from enterprises of 2,000 employees or more.

In Box’s recently completed fiscal quarter, it closed 50 deals of more than $100,000, compared with 40 a year ago; 11 deals of more than $500,000, versus eight a year ago; and two deals of more than $1 million, compared with four a year ago. It expects a strong pipeline of seven-figure deals in the back half of this year.

But in encouraging its salespeople to pursue bigger deals, Box increasingly faces competition from deeper-pocketed competitors in a total addressable market pegged at $45 billion, based on market research by Gartner and IDC.

Soccer fans, your team is coming after you

At the time of its 2012 initial public offering, Man United counted 659 million fans worldwide. Analysts estimate the team’s revenue this year will be about 587 million pounds ($763 million) — just $1.16 per supporter. Twitter Inc. has just 338 million active monthly users, yet enjoys revenue of $2.4 billion and a market value of $27 billion.

Digital marketing provides the opportunity for teams to put themselves in the middle of the sale of a service or product. It’s not simply about using a website or an app to sell fans more jerseys or baseball caps. It’s about turning the team into a platform, a way of connecting brands to customers, in the same way as Facebook Inc. and Alphabet Inc. already do.

Much in the way that price-comparison websites charge insurers or credit card companies for connecting them to customers, a sports team could, for example, offer its own exclusive video content with another provider’s mobile phone contract and take a cut of the proceeds. If that meant each fan were to spend just one more dollar a year with the club, it would provide a significant boost to sales.


Alibaba-backed apparel-sharing company YCloset brings sharing economy to a new level

Founded in December 2015, YCloset charges a monthly membership fee of 499 yuan and allows female users to rent unlimited clothes and accessories country-wide. Furthermore, users can choose to buy the apparel if they like to and prices fluctuate according to the rent count. Thus far, 75% of the income comes from membership fees and the remaining comes from sales of clothing. YCloset positions itself as a company that offers affordable luxury, professional and designer brand clothing. The company hopes to have the top famous brand to drive the long-tail brands.

In terms of business model, YCloset gradually shifted from one-time supplier purchase to brand partnerships with clothing companies. Brand partnerships allow revenue sharing between YCloset and their partners. To these clothing companies, YCloset gave them a new revenue, at the same time, they may get consumer insights from the data YCloset collects. In the future, YCloset will have joint marketing campaigns with the brands and assist in incubating new brands.

Autonomous delivery robots could lower the cost of last mile delivery by 20-fold

Last mile delivery – the delivery of goods from distribution hubs to the consumer – is the most expensive leg of logistics because it does not submit to economies of scale. The cost per last mile delivery today is $1.60 via human drivers but could drop precipitously to $0.06 as autonomous delivery robots proliferate.

Autonomous delivery robots are roughly seven times more efficient than electric vehicles on a mile per kilowatt basis. The major costs for autonomous delivery robots are hardware, electricity, and remote operators. Unlike in electric vehicles, the battery is not the largest cost component in slow moving robots. Air resistance is a function of velocity squared, suggesting that a robot traveling at four miles per hour loses much less energy than a car traveling at highway speeds to air resistance. As a result, rolling robots do not require large batteries, lowering both hardware and electricity costs relative to more traditional electric vehicles.

If rolling robots enable last mile delivery for $0.06 per mile, artificial intelligence could be advanced enough to improve their unit economics. A remote operator responsible for controlling robots in difficult or confusing situations probably will oversee roughly 100 robots, accounting for more than half of the cost per mile, as shown below. As autonomous capability improves, remote operators should be able to manage larger fleets of robots, bringing down the costs per robot.


Hospitals are fed up with drug companies, so they’re starting their own

A group of major American hospitals, battered by price spikes on old drugs and long-lasting shortages of critical medicines, has launched a mission-driven, not-for-profit generic drug company, Civica Rx, to take some control over the drug supply. Backed by seven large health systems and three philanthropic groups, the new venture will be led by an industry insider who refuses to draw a salary. The company will focus initially on establishing price transparency and stable supplies for 14 generic drugs used in hospitals, without pressure from shareholders to issue dividends or push a stock price higher.


Harvard Business School professor: Half of American colleges will be bankrupt in 10 to 15 years

There are over 4,000 colleges and universities in the United States, but Harvard Business School professor Clayton Christensen says that half are bound for bankruptcy in the next few decades. Christensen and co-author Henry Eyring analyze the future of traditional universities, and conclude that online education will become a more cost-effective way for students to receive an education, effectively undermining the business models of traditional institutions and running them out of business.

Christensen is not alone in thinking that online educational resources will cause traditional colleges and universities to close. The U.S. Department of Education and Moody’s Investors Service project that in the coming years, closure rates of small colleges and universities will triple, and mergers will double.

More than 90 per cent of Chinese teens access the internet through mobile phones, says report

The proportion of Chinese children under 10 years old who use the internet – which was only 56 per cent in 2010 – reached 68 per cent last year. More than 90 per cent of Chinese minors, those aged up to 18, can now access the internet through mobile phone and over 64 per cent of primary school kids have their own smartphones. Nearly 85 per cent of Chinese minors use WeChat, compared to only 48 per cent five years ago, but Chinese juveniles are still more fond of QQ, while Chinese adults prefer WeChat as a social app.

Curated Insights 2018.06.17

What helps or hurts investment returns? Here’s a ranking

An unexpected challenge in performing this exercise is a tendency for some elements to offset others. For example, changes in profits could be offset by widening or contracting price-earnings ratios; sentiment might offset valuation; returns tend to vary inversely with risk. Why does this matter? Because in the real world, one hand giveth while the other taketh away. This concept of cancellation matters a great deal to total portfolio returns.

The overall cost of a portfolio, compounded over 20 or 30 years, can add up to (or subtract) a substantial amount of the returns. One Vanguard Group study noted that a 110 basis-point expense ratio can cost as much as 25 percent of total returns after 30 years. That does not take into consideration other costs such as trading expenses, capital-gains taxes or account location (i.e., using qualified or tax-deferred accounts). The rise of indexing during the past decade is a tacit acknowledgment that on average, cost matters more than stock-picking prowess.

Those people born in 1948 not only managed to have their peak earning and investing years (35-65) coincide with multiple bull markets and interest rates dropping from more than 15 percent to less than 1 percent. They also lucked into a market that tripled in the decade before retirement.

Behavior and discipline > Humility and learning > Longevity and starting early > Valuation and year of birth > Asset allocation > Costs and expenses > Security selection


The forging of a skeptic

I think another thing people have gotten confused about is the sustainable competitive advantage and the moat. Durable competitive advantage and moats are not the same thing as brands. People sometimes use these terms interchangeably. I have also seen people ascribe competitive advantages to brands that don’t have them. For example, retailers — retailers have brands. We all know what Macy’s is, but retailing is fundamentally a bad business.

In essence, the merits of a brand are not the brand itself; they are the qualities of the product that create the consumer loyalty. What attracted him, ultimately, to Coca-Cola is that Coca-Cola’s formula make you more, not less, thirsty, and supposedly has been tested to prove that it doesn’t wear out the palate, no matter how much is consumed. This implies infinite sales potential. The cute commercials and cheery red logo create an association in people’s minds with those qualities. They aren’t what makes it Coca-Cola.

While there are moats that include brands, a brand is not a moat. The moat is whatever qualities are innate to the business that make it difficult to compete with

Worried about big tech? Chinese giants make America’s look tame

They have both funded ventures that offer online education, make electric cars and rent out bicycles. For the giants, such initiatives represent new opportunities for people to use their digital wallets — Ant Financial’s Alipay and Tencent’s WeChat Pay — and new ways to collect data on consumer behavior. Analysts at Sanford C. Bernstein counted 247 investment deals by Tencent in recent years and 156 by Alibaba, though given the pace of the companies’ deal-making, they said their database was “likely to be perennially incomplete.”

In a report this week, Morgan Stanley predicted that by 2027, the total market in China in which Alibaba could be making money will be worth $19 trillion — more than Amazon’s potential market worldwide.

‘As long as they’re unfriendly, it’s a sign they have confidence’

Keyence keeps up compound sales growth of 14 per cent a year (1986-2016) even with sales in the billions of dollars. It takes seemingly simple products such as barcode readers and sells them for five times the cost of manufacture.

Keyence’s first secret is its production outsourcing. It buys raw materials in bulk and sends them to component suppliers; it collects the components and sends them to assemblers and performs the final inspection of goods itself.

The second secret is what Keyence really sells: not a product, but a way to make a factory more efficient. Graeme McDonald, machinery analyst at Citigroup in Tokyo, says the group’s sales engineers “can often provide an idea of how to improve your manufacturing set-up literally on the site with an idea of the payback time and return on investment”. It offers quick victories — such as a sensor to replace manual inspection, for example — not risky projects. “The products they sell are not capital expenditure, they’re cost to the factory manager,” says Mr Noguchi. If the manager can save a $40,000 salary with a $20,000 gadget, they will sign off quickly, without worrying how much Keyence earns.

The products are high quality, if not necessarily unique. Keyence has a modest research budget and less than a tenth of the US patents held by rival automation companies such as Fanuc.

Fanuc in trouble? Talk to the (robot) hand

Fair enough, it’s a tough world for all iPhone dependents. Here’s a wrinkle in the bear-case thesis, though: Overseas shipments of robots and Robodrills from Yokohama, while down elsewhere, are up sharply to Asia. The volume of robots shipped by the port – mostly Fanuc’s – remains close to its highest in decades, at about 5,000 units in April. The company’s backlog of orders is near to its highest in more than two years, according to Bernstein analysts.

How e-commerce with drone delivery is taking flight in China

It is still waiting to earn back its investment in drone-delivery infrastructure, although it says that making a delivery by drone costs a fifth of the price than by man-and-van, once the driver’s labour is taken into account. Liu Qiangdong, JD’s chief executive, says drone delivery will cut costs by 70% once it is scaled up across the country. Villagers tend to buy washing powder, accessories for their phones, maternity goods and fresh food. The firm has made 20,000 such deliveries to date.

JD may have added drones to daily Chinese village life, but whether they will make financial sense for the company over time remains to be seen. Current models of drone are pricey, although JD says the cost will gradually come down as it scales up the network and builds more drones (it plans to sell those it makes to other firms, as well as use them for its operations). The government approves of its operations in rural areas, and is planning to build a new train station in Suqian next to JD’s drone base. If JD can use drone delivery to cut its costs and attract rural shoppers, that will help the firm compete with its arch-rival in e-commerce, Alibaba, which has not, as yet, seen the value of drone delivery. JD hopes that will prove to be a mistake.


Internet lending is booming in China

The balance of online consumer loans in China has grown about fivefold between 2015 and 2017, reaching 350 billion yuan ($54.6 billion), according to Chinese research company Analysys. According to a survey conducted by research specialist Analysys in December 2017, people between the ages of 24 and 35 accounted for more than 70% of consumer borrowers in China.

Chinese consumers, especially people born in 1980 and later, are less squeamish than their older peers about buying on credit. But the total balance of consumer loans in China is still about 60% lower than that in the U.S. and is expected to continue growing. Analysys estimates that the balance of internet loans in China will more than double to 720 billion yuan in 2019, compared with 350 billion yuan in 2017. That flow of credit will likely give a lift to the Chinese consumer market.

The scooter economy

The mistake in Kalanick’s thinking is two-fold: First, up-and-until the point that self-driving cars are widely available — that is, not simply invented, but built-and-deployed at scale — Uber’s drivers are its biggest competitive advantage. Kalanick’s public statements on the matter hardly evinced understanding on this point. Second, bringing self-driving cars to market would entail huge amounts of capital investment. For one, this means it would be unlikely that Google, a company that rushes to reassure investors when it loses tens of basis points in margin, would do so by itself, and for another, whatever companies did make such an investment would be highly incentivized to maximize utilization of said investment as soon as possible. That means plugging into the dominant transportation-as-a-service network, which means partnering with Uber.

My contention is that Uber would have been best-served concentrating all of its resources on its driver-centric model, even as it built relationships with everyone in the self-driving space, positioning itself to be the best route to customers for whoever wins the self-driving technology battle.

Why you should read those boring 10-K filings

The vast majority of the text changes are concentrated in the Management Discussion and Analysis (MD&A) of the 10-K. These disclosures also tend to be more negative than positive, perhaps because the reports are typically drafted by lawyers who tilt toward disclosing negative trends more than positive ones. When the authors applied natural language text processing to evaluate the changes, they found that 86 percent reflected negative sentiment shifts and only 14 percent positive shifts. Furthermore, the text differences contain useful information for predicting future earnings: Changes in the 10-K written text today predict earnings surprises in the future.

Given this negative bias to the textual changes and their ability to predict future earnings, the study shows that companies with 10-K text modifications experience noticeably lower future stock returns than other firms. For example, the authors construct a portfolio that goes long on companies with no material textual changes and shorts firms that contain such changes. That portfolio earns an abnormal positive return of up to 7 percent per year above the market.

Curated Insights 2017.07.16

Tom Russo on shareholder value creation and global brands

Competition is both global and local. The fight is typically fair, but not always. In India, for instance, Nestlé suffered a below-the-belt punch with its famous Maggi instant soup business. It suffered a massive goodwill assault by local forces that alleged the product would kill young children because of ingredients that were toxic. There is absolutely no hard science to support the claim. Yet, it ran across all 250 newspapers in India, and Nestlé’s brand preference score for the Maggi product dropped from 98% to 8%.

The century of biology

When engineering, biology, and computer science come together it elevates bio to a “read/write” paradigm. That is, you don’t just read the code of biology but you can also write, or design, with it.

…when you looked at a cell type that was normal, and you looked at its related diseased counterpart, you could see changes in the differential set of genes that were being expressed in one cell type vs. another (the cell program). Sometimes there’s not a mutated gene that’s causing the disease — it’s not just a bug or “bad gene” in the code (genomics) — it’s the cell running the wrong program, which is where epigenomics comes in again. The gene doesn’t just turn on and off like a light switch to cause disease, it can be more like a dimmer, going higher or lower (too much or too little of a gene). What causes a healthy cell to shift in to a diseased state, whether it’s cells multiplying uncontrollably in cancer or dormant immune cells suddenly getting activated in autoimmune disease? The ability to understand how the genome is being deployed or regulated within a cell — is it the wrong dosage or level for a particular set of genes? — becomes a very interesting new avenue to help us get to the right drug, for the right patient, at the right dose, and at the right time.

This leads us to the second derivatives of read/write, which is to enable insight/design. We’re not just generating data on the read side, we’re also fundamentally understanding it deeply in ways that we couldn’t before, especially with multiple data streams and machine learning to help make sense of it all. And on the write side, we’re not just editing, we’re also designing with biology. In the future, biology can become its own creative medium of sorts.

…our ability to read/write biology will disrupt a wide range of industries. In addition to its obvious impact across health, we’re increasingly using biology for manufacturing. Eventually it’s going to impact areas people don’t typically think about as “biological” — like textiles, architecture, and many more areas — in ways we can’t even conceive of yet.


Scientists can use CRISPR to store images and movies in bacteria

CRISPR was invented billions of years ago, as a way for bacteria to defend themselves against viruses. The bacteria grab the DNA of invading viruses, incorporating it into their own genomes. That viral DNA always gets inserted in the same place, and new sequences get added after old ones, as if the bacteria were stacking books on a shelf. They use these archives to guide an enzyme called Cas9, which cuts and disables any viral DNA that matches the stored sequences.

So, first and foremost, CRISPR is a kind of genetic memory—a system for storing information. And that information doesn’t have to be the DNA of viruses. Scientists can now encode any digital file in the form of DNA, by converting the ones and zeroes of binary code into As, Cs, Gs, and Ts of the double helix.

Recording information in DNA isn’t new: That’s effectively what living things have been doing since the dawn of life itself. More recently, scientists have realized that DNA makes the perfect storage medium. It takes up so little space that you could fit all the world’s data in the back of a truck. It’s durable, provided it’s kept cold, dry, and dark. And it is immune to obsolescence: DVDs and Blu-Rays will eventually go the way of cassettes and laser discs, but humans will always have the desire and means to read DNA.​


In China, shoppers buy bad loans online with their groceries

For 4.15 million yuan ($610,000), customers on the site can bid for the debt of a steelmaker from Zhejiang. The company has failed to pay back a 9.95 million-yuan loan, including interest, so a distressed asset manager is auctioning it off to the highest online bidder.

“Financial technology and e-commerce in China has reached a high level of sophistication. Online platforms are leveling the playing field in the distressed debt market as it means everybody gets access to the same information.”

“Conducting NPL auctions online has increasingly become a trend. More investors are using Taobao as a platform because of the simplicity, transparency and confidentiality of the bidders’ identity.”

But bad-loan investing isn’t like trading equities or even ordinary debt, which raises questions over the opening up of the market to rank-and-file investors.

The most important moat

“A Bezos Moat is premised on the idea that the customer is willingly and is frequently entering into a commercial transaction with the company because the customer is deriving more value from the transaction than he or she is paying for.

“A Buffett Moat attempts to identify companies that will be the only one (or one of a few) available in a commercial landscape, so that the customer is, in effect, forced to transact with these companies (i.e. only bridge, only newspaper, only soft drink option).”