How To Spot Increasing Returns Businesses

“My plan is to start from a completely blank state, taking nothing about technology for granted. I will build the argument piece by piece from three fundamental principles. The first will be the one I have been talking about: that technologies, all technologies, are combinations. This simply means that individual technologies are constructed or put together-combined-from components or assemblies or subsystems at hand. The second will be that each component of technology is itself in miniature a technology. This sounds odd and I will have to justify it, but for now think of it as meaning that because components carry out specific purposes just as overall technologies do, they too qualify as technologies. And the third fundamental principle will be that all technologies harness and exploit some effect or phenomenon, usually several.” – Brian Arthur

Growth is so hot right now. It sounds obvious, right? Of course, investors want growth. The dichotomy starts with how much to pay for that growth. Those on the shrewder side of Scrooge don’t want to pay anything for future growth (*takes a break from typing to raise hand*). “Growth” investors don’t mind paying up because addressable markets are huge, stocks only go up, and the Fed keeps printing money at 0% interest rates.

Yet this argument misses the point. Our goal as investors is to understand the mechanisms behind growth. To determine why it happens effortlessly in some companies and painstakingly in others.

In turn, we can buy businesses that look expensive but are cheap because growth is effortless. At the same time, it prevents us from buying painful growth enterprises.

Two economists can help us create this necessary framework: Paul Romer and Brian Arthur. The theorists’ seminal works were six years apart yet joined at the hip in our search for foundational growth principles.

And I know what you’re thinking … “But Brandon, macroeconomic theory is worthless to individual stocks!” And I agree! These two models apply at the micro-level as well as the macro level. It’s this dual-application that allows us to capture insights and develop a framework.

This framework cocktail needs a few ingredients:

  • Increasing returns to scale
  • Technology
  • Nonrival goods
  • Endogenous growth

The model allows us to understand the knowledge-economy growth framework.

Technological transformation enables companies to generate non-rival goods at scale spurring massive endogenous growth. This growth gives birth to increasing returns to scale as companies leverage their non-rival goods to create exponentially greater rival goods.

Three-Legged Growth Stools & Nonrival Goods

Growth theories are as old as Adam Smith’s Invisible Hand. Yet all these studies focused on two things: increasing specialization and labor. The problem with this logic is that those are static objects with finite resources. In other words, you can only have so much labor (equal to population growth). Returns suffer as soon as population growth slows and industry specialization permeates. Hence diminishing marginal returns.

“Following Smith, Marshall, and Young, most authors justified the existence of increasing returns based on increasing specialization and the division of labor. It is now clear that these changes in the organization of production cannot be rigorously treated as technological externalities. Formally, increased specialization opens new markets and introduces new goods. All producers in the industry may benefit from the introduction of these goods, but they are goods, not technological externalities.”

Romer’s seminal December 1989 paper describes the major difference in his model from conventional thought (emphasis mine):

“Growth in this model is driven by technological change that arises from intentional investment decisions made by profit-maximizing agents. The distinguishing feature of the technology as an input is that it is neither a conventional good nor a public good; it is a nonrival, partially excludable good. Because of the nonconvexity introduced by a nonrival good, price-taking competition cannot be supported, and instead, the equilibrium is one with monopolistic competition. The main conclusions are that the stock of human capital determines the rate of growth, that too little human capital is devoted to research in equilibrium, that integration into world markets will increase growth rates, and that having a large population is not sufficient to generate growth.

Romer argues that growth comes not from population size, but from the ability of that population to invest in research to create new technologies. These technologies become nonrival goods, which allows the company (or economy) to create a monopolistic environment.

There are three legs to Romel’s technological growth stool:

  1. Technological change (I.e., improvement in instructions for mixing raw materials) lies at the heart of economic growth
  2. Technological change arises in large part because of intentional actions taken by people who respond to market incentives (I.e., endogenous)
  3. Instructions for working with raw materials are inherently different from other economic goods (upfront fixed costs and zero marginal costs)

Technological Change Lies At The Heart of Economic Growth

Romer believes technological change spurs economic growth. But not any change. Romer emphasizes the accumulation of a partially excludable, nonrival input.

Economic goods follow a property matrix. Goods are either rival/non-rival and excludable/non-excludable. Let’s break this down.

Rival goods mean that if one person uses them, another person can’t. Perishable items are a good example of rival goods. If I take the last scoop of chocolate ice cream it means you can’t enjoy chocolate ice cream.

Nonrival goods are the opposite. Many people can use the good without diminishing the value to each individual. Think public transportation. I can hop on a public bus and receive the same utility regardless if the bus is full of people or empty.

Next is exclusivity. This is self-explanatory. A good is either exclusive or non-exclusive depending on how the owner lets a consumer use the good. Intellectual Property (IP) is an example of an exclusive good. Only those with the IP patent can use that technology.

Romer explains this concept through “design” (emphasis mine):

“The example of a nonrival input used in what follows is a design for a new good. The vast majority of designs result from the research and development activities of private, profit-maximizing firms. A design is, nonetheless, nonrival. Once the design is created, it can be used as often as desired, in as many productive activities as desired.”

This is the most important concept in Romer’s claims. Designs are ideas that transcend human capital. In turn, the cost structures inherent in ideas are different than that of human capital. Where human capital grows linearly, ideas (Romer’s “designs”) grow exponentially.

Why does this occur? Because human capital is a rival good and their ideas are nonrival goods. Romer explains (emphasis mine):

“In this sense, a design differs crucially from a piece of human capital like the ability to add. The design is nonrival but the ability to add is not. The difference arises because the ability to add is inherently tied to a physical object (a human body) whereas the design is not. The ability to add is rivalrous because the person who possesses this ability cannot be in more than one place at the same time; nor can this person solve many problems at once. By the argument noted above, rivalry implies that human capital is also excludable. Thus, human capital can be privately provided and traded in competitive markets. In contrast, the design is nonrival because it is independent of any physical object. It can be copied and used in as many different activities as desired.”

Knowing this we can create a hypothesis: If a company is profit-maximizing, they should use rival goods (human capital) to trade for nonrival goods (ideas, designs, etc.). This allows the profit-maximizing company to generate exponential returns from their base of rival goods.

Romer agrees with this hypothesis when he notes the expendability of rival goods like human capital in exchange for “eternal” nonrival ideas. This leads to two implications of nonrival goods on growth (from Romer’s piece):

  1. Nonrival goods can be accumulated without bound on a per capita basis, whereas a piece of human capital like the ability to add cannot. Each person has only a finite number of years that can be spent acquiring skills. When this person dies, the skills are lost, but any nonrival good that this person produces—a scientific law, a principle of mechanical, electrical and chemical engineering, a mathematical result, software, a patent, a mechanical drawing, or a blueprint—lives on after the person is gone.
  2. Treating knowledge as a nonrival good makes it possible to talk sensibly about knowledge spillovers, i.e. incomplete excludability.

Why Nonrivalry Matters: If nonrival inputs have productive value then the output cannot be a constant return to scale function of all its inputs taken together. This is because nonrival goods have zero incremental cost to replicate across a fixed rival input set.

Explaining Nonrival Exponential Growth

Say there’s Firm A that invests 10K hours of engineering time (rival good) to make a design (nonrival good) for a 20-megabyte hard disk drive (rival good).

In other words, 10K hours of rival goods translate into 20 megabytes of nonrival goods.

Now let’s assume Firm A spends $10M to build a factor and hire 100 workers to make these disks. At $10M Firm A generates 2 trillion megabytes of storage per year (100K units of the disk drive).

This is where we see the power of nonrival goods. Say we have a competing firm (Firm B). They spend 20K engineering hours to design a 30-megabyte hard disk drive. This hard disk drive uses the same labor and factory inputs (rival goods) as Firm A.

Here’s the difference. When Firm B scales its rival goods linearly, it’s nonrival goods scale exponentially. When Firm B doubles its output, it uses 20K hours, 2 factories, and 200 workers to make 6 trillion megabytes of storage. That’s 3x more than the original output of Firm A.

In other words, nonrival goods allow for increasing returns to scale.

Exponential Growth More Prevalent … Thanks Tech!

Knowledge scales exponentially over time because each incremental worker has a larger stock of knowledge accumulated than the worker before her.

This is the same advantage experienced by machine learning/deep learning algorithms. The older the algorithm, the more data it has to learn from. Making it an even better algorithm with age than its next closest peer.

Increased knowledge produces better technology to make better goods. Production goods are rival. Yet the knowledge gained from the production of those goods is nonrival. Thus, Romer concludes all-new designs are nonrival, partially excludable inputs. Here’s his point (emphasis mine):

“If an inventor has a patented design on widgets, no one can make or sell widgets without signing a contract with the inventor. On the other hand, other inventors are free to spend time studying the patent application for the widget. In doing so, they may learn knowledge that helps in the design of widgets. The inventor of the widget cannot stop the inventors of widgets from learning from the design of a widget.”

The Formula For Maximum Profits

Romer didn’t know it at the time, but he explained why every investor loves the software (SaaS) business model. Through Romer’s growth model we see the power of software businesses at first principles.

Microsoft Office is a great example. The software was a partially excluded nonrival good. The software was free to use on any PC, but Microsoft alone housed the IP and rights to the software. In turn, the company was able to generate outsized profits thanks to the imperfect competition embedded in a nonrival + rival output model.

Charles I. Jones of Stanford University explains this concept (emphasis mine):

“A key to making these models applicable was the recognition that ideas, while nonrival, are not pure public goods. Recall that the definition of a pure public good is something that is both nonrival and nonexcludable. While nonrivalry is a property of the economic environment, excludability is a function of institutions and the decisions that societies make. Institutions such as the patent system – or even just trade secrets – can allow ideas to be partially excludable, at least for a certain period of time.”

Conceptualize nonrival ideas and perfect competition as a solar eclipse. There’s a finite point in which both align to create supernatural profits. The point where the moon engulfs the sun. When that happens, the company that moves first wins.

Jones continues his thought (emphasis mine):

“Successful innovators are awarded a patent that gives them the exclusive right to produce with their invention. This allows them to charge a mark-up over marginal cost, subject to imperfect competition, and to earn the profits that ultimately serve as the carrot that motivates the search for new ideas.”

The beauty of Romer’s model is we don’t have to limit our examples to patents. Nonrival ideas have varying flavors:

  • Network effects
  • Economies of Scale Shared
  • High Switching Costs
  • Lowest Cost Operator

We can think of Warren Buffett’s “moats” as Romer’s nonrival ideas that catalyze increasing returns to scale.

The Magic Formula (if there is one) is a dedicated investment in nonrival ideas (R&D and growth cap-ex) at the expense of rival good outputs (profits, operating margins, dividends). This idea is for long-term investors and entrepreneurs. It means losing profits today to grow your knowledge base (I.e., edge) for tomorrow.

Amazon runs this playbook better than almost any company. Their mission to offer the lowest prices while providing the best service is an example of trading rival goods for nonrival ideas (scale efficiencies shared).

Carvana (CVNA) does this too. Spending as much as possible to build the infrastructure and network  (nonrival ideas) to reap winner-take-most profits (rival outputs).

Reviewing Romer’s Model

Romer transformed the way we think about growth companies and what it means to invest in nonrival goods to create increasing returns to scale. We learned the three legs in the growth stool are:

  1. Technological change (I.e., improvement in instructions for mixing raw materials) lies at the heart of economic growth
  2. Technological change arises in large part because of intentional actions taken by people who respond to market incentives (I.e., endogenous)
  3. Instructions for working with raw materials are inherently different from other economic goods (upfront fixed costs and zero marginal costs)

Six years after Romer published his breakthrough growth theory, Brian Arthur penned his famous essay, Increasing Returns and the New World of Business In Harvard Business Review. Arthur leveraged Romer’s research and highlighted the importance of technology in the role of increasing returns to scale for nonrival idea generation.

Combined, the two men create one of the best mental models for understanding growth at the company level.

The Knowledge Economy Spurs Increasing Returns

“But steadily and continuously in this century, Western economies have undergone a transformation from bulk-material manufacturing to design and use of technology — from the processing of resources to processing of information, from application of raw energy to application of ideas. As this shift has occurred, the underlying mechanisms that determine economic behavior have shifted from ones of diminishing to ones of increasing returns.” – Brian Arthur

Like Romer, Arthur understood the power of shifting from capital to knowledge-based economies. Before we dive in, this isn’t to say that investors can’t make money buying “capital” first businesses. There’s plenty of cash lying on the ground for those willing to buy great manufacturing businesses at less than replacement cost.

Arthur understood this dichotomy in these two economic frameworks (emphasis mine):

“The two worlds have different economics. They differ in behavior, style, and culture. They call for different management techniques, strategies, and codes of government regulation. They call for different understandings.”

Let’s try to understand these differences.

The Old Economy: Standard Pricing & Fierce Competition

Alfred Marshall coined the term “diminishing returns” in the late 1880s and 1890s. During that time, business meant raw materials. Metal, dyes, coal, lumber, chemicals, commodities. These raw materials were then manufactured and sold at prices determined by the market, not the company. These businesses were perfectly competitive and thus easy to model.

As Arthur notes, Marshall’s early model still lives over 100 years later. He notes some nuances like brand names and product differentiation. But the premise remains: profits and margins are small for the perfectly competitive firm (emphasis mine):

“But typically, if these companies try to expand, they run into some limitation: in numbers of consumers who prefer their brand, in regional demand, in access to raw materials. So no company can corner the market. And because such products are normally substitutable for one another, something like a standard price emerges. Margins are a thing and nobody makes a killing.”

That doesn’t sound like a great business now does it. Fortunately, there’s another option. The Knowledge Economy.

The Knowledge Economy: Returns Beget Returns

Like Romer’s three-legged growth stool, Arthur developed the six-legged insect of increasing returns (and yes, I picked insect because they only have six legs):

  • Market instability (market tilts to favor a product that gets ahead)
  • Many potential outcomes (luck, skill, or chance)
  • Unpredictability
  • Ability to lock in a market
  • The possible predominance of an inferior product
  • Fat profits for the winner

If a company has all these attributes, you’ve got yourself a big, fat increasing returns insect.

There’s a few examples that come to mind: Facebook, Microsoft DOS, Bloomberg Terminal, Oracle Databases, Stock Market Exchanges

We can tie Romer’s theory to Arthur’s above observations: Truly great companies use nonrival ideas to generate competitively advantaged rival outputs in markets predisposed to increasing return abilities.

Arthur identified three attributes (or nonrival ideas) winner-take-most businesses have:

1. Up-front costs

High-tech products cost a lot to develop. Arthur uses the Windows hard drive disk as an example. The first Windows disk cost $50M to produce. The next disk? A mere $3. That’s increasing returns to scale thanks to massive upfront R&D investment.

Dennis Hong talks about this in his interview with Value DACH. Hong (one of the sharpest investors in the game,) loves high capital intensity businesses for this exact reason. R&D investments are a competitive advantage if they translate to increased sales/margins.

2. Network Effects

The value of the product or service increases with each incremental user on the platform or product. Think social media or product marketplaces like Etsy, Redbubble, eBay, etc. Arthur used Java as an example, “Java has competitors. But the more it gains prevalence, the more likely it will emerge as a standard.”

3. Customer Groove-in

Once you train a customer how to use a product with a steep learning curve, each incremental update requires less training and a stickier user base. This means products that gain market share in the initial adoption phase have a greater tendency to take more share as the market expands. Thus locking in customers.

Bloomberg Terminals are a great example of this attribute. It’s cumbersome to use with an archaic UI/UX. Yet thousands of investors use the software every day.

We know what it takes to create increasing returns businesses (our insect model). And we know what those companies look like (high upfront costs, network effects, customer groove-in). How can we see if a company is executing on this increasing returns strategy? Arthur offers two clues.

Active Strategies for Increasing Return Market Domination

There are two things companies can do if they want to capture the profits of increasing returns business:

  • Hit the market first
  • Have superb technology

Yet as Arthur explains, having these two attributes isn’t enough to secure market leadership (emphasis mine):

“These maxims are true but do not guarantee success. Prodigy was first into the on-line services market but was passive in building its subscriber base to take advantage of increasing returns. As a result, it has fallen from its leading position and currently lags the other services. As for technology, Steve Jobs’s NeXT workstation was superb. But it was launched into a market already dominated by Sun Microsystems and Hewlett-Packard. It failed.”

To dominate a market, a company has to move fast. They have to command broad market penetration and user adoption. Arthur’s strategy for doing so sounds like a SaaS prophecy.

Strategy 1: Discount & Conquer

His first strategy is “to discount heavily initially to build up an installed base.” How many software companies use this strategy? Dropbox and Box.com come to mind. Slack also did this by offering a free version to its users with limited capacity and functions.

Here’s the biggest problem: if you don’t turn those free users into paying customers all discounts are worthless.

Strategy 2: Create The Most Fun Playground

Another strategy Arthur mentions is creating the greatest incentives for others to use your platform. Think of two/three sided marketplaces or software developing platforms. Since most technology runs on ancillary pieces of software, those that create the easiest integration system win.

This is what Salesforce does with its acquisition strategy. Come to Salesforce and easily integrate all other technology apps in one place.

NetWare did the same thing in the late 80s. Here’s Arthur’s take (emphasis mine):

“NetWare’s success depended on attracting software applications to run on NetWare — which was a part of the ecology outside the company’s control. So it set up incentives for software developers to write for NetWare rather than for its rivals. The software writers did just that. And by building NetWare’s success, they ensured their own.”

Again we see similarities with Carvana’s business model. Offer customers the highest price for their used car to build inventory which they can then use to create a more attractive used-car online marketplace. Carvana is (in effect) paying an up-front lock-in tax for the potential to capture the entire market.

Strategy 3: Linking & Leveraging

Linking and Leveraging builds on the ecological model in Strategy 2. In short, a company takes a customer from one “node” in their business to the next “node”. These “nodes” occur during software/product updates. Microsoft shifted its customers from DOS to Windows, then to Windows 95.

Apple does this to an extent, too. Moving customers from the iPhone 5 through 12 Pro.

The key to this web-based strategy is to let other companies (dependents) lock-in their products on top of the dominant product. Arthur explains (emphasis mine):

“The key in web building is the active management of the cross-company mutual feedback … It also means that, rather than attempting to take over all products in the ecology, dominant players in a web should allow dependent players to lick in their dependent products by piggybacking on the web’s success.”

Visa (V) and Mastercard (MA) do this better than almost any company. By providing the payment processing web to the world, they allowed dependent companies (e-commerce, POS systems, digital payments) to create an even stronger rails system.

Bringing Everything Together

Though six years apart, Romer and Arthur’s work creates a beautiful mosaic for analyzing and understanding how growth works on the company level.

But it’s more than that. Their research gave us the tools to study business models at the frontier of innovation and technological breakthroughs. They help us wrangle the idea of businesses foregoing profits for the sake of long-term competitive advantages.

Romer and Arthur assert that increasing returns businesses play a longer game than their opponents. They trade rival goods for nonrival ideas. Profits for moats.

Here’s how you can install these ideas in your investment process:

  • Focus on top-line and gross profit growth, not net income
  • Study businesses that have the three legs of growth and six legs of increasing returns to scale (look for chairs and insects!)
  • Find businesses trading rival goods for non-rival ideas
  • Invest in industries/markets predisposed to increasing returns to scale

That’s the best part of this business. Go learn at the edge of technology. Make bets on companies trading short-term gains for long-term rewards. Invest in leaders who understand the power of increasing returns to scale through nonrival ideas.

Further reading:

http://tuvalu.santafe.edu/~wbarthur/Papers/HBR.pdf
http://www.economia.ufpr.br/Eventos/Downloads/Minicurso2b.pdf
http://www2.csudh.edu/gsterling/complex.pdf
Paul Romer: Ideas, Nonrivalry and Endogenous Growth
http://www.dklevine.com/archive/refs42232.pdf