Part II. Exploitable Gaps . . .
Abstracting the Accelerants of the New Economy
When traditional business and investment perspectives are confronted by accelerating change, they are slow to adapt to the new realities. For the individual investor, this creates many exploitable gaps in investors' expectations. In "New Rules for the New Economy," Kevin Kelly (1998) argued that the network economy has moved from change to flux, a space edging its way between order and chaos, a flux that is a creative force of both destruction and genesis. In flux, change is scaled upward: from changes in the game (new winners and losers-Barnes and Noble becomes a winner with its book superstores--a category killer, and Mom & Pop bookstores become losers.), to changes in the rules of the game (entirely new structures, like the Internet, emerge-Amazon.com is a new winner in multiple games, selling not only books but also music, videos, and who knows what else), to changes in how the rules of the game are changed (change morphs into creative destruction--eBooks emerge that may completely alter the structure of publishing, advertising, education, and the like). And, flux is not scary if you realize that for each niche, each job, or each company that is creatively destroyed, that the process of destruction, like Shiva's, becomes the creative genesis of two new niches, jobs, or companies that, in turn, spawn two more new niches. According to Kelly, in the industrial economy, success was self-limiting, following the law of diminishing returns; whereas, in the network economy, success is self-reinforcing, following the law of increasing returns.
The increasing expansion in the market capitalization of high technology stocks rests on the twin pillars of the microcosm and the telecosm. The laws of Moore and Metcalf enable the Internet, which is the fulcrum for these levers of exponential growth in a knowledge economy. In the new knowledge economy, you already, because you invest in high technology stocks, must recognize that these laws are the twin accelerants propelling increased productivity.
Moore's Law and Metcalfe's law encapsulate the exponential power of compounding. In 1965, Gordon Moore, an inventor of the integrated circuit and a founder of Intel, posited that the number of transistors on integrated circuits was increasing periodically every 18 (12-24) months. This doubling of capacity at lower cost was made possible by moving down the learning curves-ever smaller, faster, cheaper--of photolithography, decreasing line widths, and increasing wafer sizes. This rate of progress will continue for about 20 more years, until the insulators between lines are just a few atoms thick, before a discontinuous innovation is required. Ray Kurzweil (1999) noted that the exponential growth of computing did not start with Moore's law. He plotted forty-nine notable computing machines-five mechanical, three electro-mechanical, eleven vacuum tube, 12 discrete transistor, and eighteen integrated circuits--of the 20th Century on a graph to illustrate powers of ten Increases in Speed per One Thousand Dollars Cost. The exponential relationship held through out the Century, with a slight exponential increase of an exponential increase about the time that integrated circuits became available. The delay in noticing this effect is an example of the lily pond effect: any exponential effect is always virtually unnoticeable at first, becoming obvious only after many doublings.
Using this data as evidence for a more general thesis, Kurzweil (199, p. 30) posited, "The Law of Accelerating Returns: As order exponentially increases, time exponentially speeds up (that is, the time interval between salient events grows shorter as time passes)." Kurzweil sees technology as an evolutionary form of accelerating returns that, including Moore's law, is without limit. This is very good news for humankind and investors.
What's more, Metcalfe's Law has a higher rate of compounding. That is, the price-performance ratio is increasing more rapidly for fiber-doubling every 9 months, than for storage-doubling every 12 months, than for microprocessors-doubling every 18 months. Robert Metcalfe, inventor of Ethernet and founder of 3Com, is credited with formulating this eponymous law: the value of a network expands exponentially as its members (nodes, devices) increase arithmetically (V = N (N-1) = N-squared minus N). According to Metcalfe, "When you connect computers together, the cost of doing so is n, but the value is n-squared, because each of the machines that you hook up gets to talk to all of the other machines in the network." This means that the rise in network costs is linear, but at a key juncture-the critical mass crossover-value surpasses costs and continues to explode. A ten person network has a value of 90 ((10 X 10 = 100 - 10 = 90). When the network size doubles its value more than quadruples (380), when the base size increases from 10 to 100, the value expands a hundred-fold, from 100 to 1000, value explodes a million times.
Understanding Network Effects.
Loosely following Mauboussin in "Network to Net Worth," I will summarize his main ideas about network effects, adding a few amendments and additions. Network effects exist when the value of a good increases because the number of people using that good increases. This is possible because information and knowledge are non-rival goods that can be consumed by multiple agents. Moreover, multiple distributed agents simultaneously create value during a network's multiple interactions between these distributed agents. Networks permit many-to-one, one-to-one, one-to-many, and many-to-many potential interactions. Thus, the value of networks not only increases as the square of the two-way paths but also as the number of possible one-many and many-many paths increases. As a mesh, a "web" is more valuable than a "chain," just as using Internet protocol generates more value than a dedicated circuit. For example, a one-to-one connection between Paris and Berlin increases in value when the conversation becomes many-to-many by including all of the Ministers in the European Commonwealth in a conference call.
Direct network effects exist when the value of a good increases as more agents consume that good. Indirect network effects exist when the value of a good increases as the number or variety of complementary products (or complementor relationships) increases. Network effects vary in their level of intensity, from weak to strong. Weak network effects are reliant on the intrinsic usefulness of the product, service, or information to the customer, although the company may benefit additionally from a network effect, like the hub and spoke system of an airline that increases the numbers of customers on a flight. According to Mauboussin, "As investors, we are particularly interested in situations where network effects are both powerful and can be captured as shareholder value."
Apart from increases in size or reach, the driver of strong direct network effects is interactivity. Interactivity produces a strong network effect because increasing contact between its members allows the members to add value exponentially to the network. That is, as member contact increases arithmetically value expands exponentially, and, in addition, an emergent system-effect is created by the interactions of the distributed agents, which is also multiplicative, its emergent result becoming more than the sum of the parts.
A transaction network derives its value by creating a forum of exchange for its members. For example, in less than two years eBay grew from less than 1 million to 12.6 million users. In mid-1998, critical mass was achieved and growth exploded. EBay's market capitalization tracked the rate of its increase in membership. Users want to be where everyone else is. Buyers want sellers; sellers want buyers. Not only that, because eBay was first-to-scale in the auction market, neither free listings from Yahoo!, with its huge installed membership base, nor Amazon.com's entry, with its experience in satisfying e-Commerce users, could unseat eBay because of its first-mover dominance, its achievement of critical mass locked-in its increasing returns.
Once a community forms, value expands exponentially as the number of member interactions increase arithmetically. Not only do members add value to the community based upon the potentiality of interactions, but also the community becomes more valuable to the members as they interactively invest resources in it. As members increase the value of the network by contributing their individual resources of time, money, and affect, their commitment to the community deepens. An affect investment occurs when members invest their emotions in the community, those times when the community is found to be an exciting and enjoyable space. Mauboussin's transaction and community forms of "compatibility" are examples of strong, interactive, direct network effects.
Although all devices have an intrinsic function, a subclass of devices requires a complementary relationship with other devices, requiring the complementors to adapt to one another, to co-evolve. Devices, like a DVD player or an operating system, may require compatible ancillary devices, like DVD discs or software applications to increase value. The complementors exist in a value web ecology that is committed to providing a total value solution: everything must work for anything to work. Thus, Mauboussin's device "compatibility" can be re-categorized as a weak-to-moderately-strong, indirect network effect, where degree of strength is a function of the value added by complementors to a devices' intrinsic functions. The advantage of this revised distinction is that it increases parsimony as it sharpens the focus on interactivity.
Networks and Economic Value
For investors seeking excess returns, networks offer economic value by supplying multiple sources of revenues and by inducing nonlinear rates of growth that produce gaps in expectations that can be exploited. Brian Arthur's Law states, "Of networks there shall be few." Increasing returns for the network that is first to scale and reach critical mass creates a winner-take-most economic structure. Network building is a mad rush to gain members regardless of the initial costs of giveaways. This is so because if you are not the winner, becoming a near-monopoly, then you are doomed forever to an economic life as an also-ran. It is imperative to form the net-community, and then monetize it. Valuable services can leverage its value. At least six sources of economic value have been identified:
1. Commerce/transactions. Both direct sales to member or indirect sales by referrals to other sites for a fee are exploitable opportunities. The branded site brings acolytes under its wings or uses its prestige to refer to its satellites, saving the members of its community the cost of searching among a multiplicity of unfamiliar sites. Both the number of web pages and the number of unique visitors follow a power law, another exponential effect. The growth of membership is a leading indicator of future revenues. The few branded sites that hath many visitors receive many visits to their many pages. For example, in June 2000, Yahoo! reached 65% of the home and work audience in the U.S. for a combined time of 95 minutes per user, making it the stickiest of the web sites. Yahoo! reported 156 million unique users, 155 million registered users, and 680 million average daily page views, respectively up 95%, 138%, and 119% year over year; whereas, for the second quarter, revenues were 270 million, up 110%, net income was 74 million, up 174%, EPS, at .12, increased 140%, and free cash flow increased to 120 million, almost 45% of sales.
2. Advertising. Only those nodes attracting many members attract most advertising dollars. Not only that, networks can target advertising to users whose interests or buying patterns fit desirable profiles, measure results of clicking ads more accurately, and provide customer interaction sites. It's winner-take-all eyeballs, eyeballs, eyeballs at a branded location, location, location. For example, in the second quarter, Yahoo! had 3,675 advertisers, adding 300, including 2 more in the Fortune 50, or 29 out of 50.
3. Subscription. Very large or very specialized networks can charge a subscription fee. The value proposition offered is worth the fee; quality or convenience attracts consumers.
4. Data. Amassing either aggregated or personalized data is easier and cheaper for networks. They can use it themselves or sell the information to others for their use. When Amazon.com offers you other books that fit your interests, they achieve significant increases in sales-impulse buying for intellectuals.
5. Incubation. Because a network community is linked to a home site, it can be leveraged into new business opportunities, either directly by the company or by associates in their value web. These associates are charged fees, or a cut of their transactions fees is taken, for promoting them as satellites; opportunity to participate in the community can be leveraged by acquiring equity in the wannabe's before permitting access to the valuable community. In effect, the site is franchising new companies because of the power of its large net-community. For example, Yahoo! Shopping has 11,300 stores, has moved from fixed to variable fees, and added 800 new stores in the latest quarter. Additional sources of revenues are continually added, like Bluelight.com, a joint venture with K-Mart, B2B Marketplace, Yahoo! line of electronics, Merchant Auctions, Yahoo!Politics, Yahoo!Photos, Yahoo!Pepsi, wireless alliances with AT&T, Italia Mobile, and others, and CorporateYahoo!.
6. Globalization. Because the Internet is a global network, network brands are being extended internationally. And, scaled cheaply. Peter Lynch's old idea of buying a successful brick-and-mortar business that was replicating itself in multiple sites and cities is now up-dated as a global-click-strategy by AOL, Yahoo!, and e-Bay, because achieving critical mass profoundly reduces the costs of replicating new versions of existing services at an international scale. For example, with the recent launch of Yahoo!India, Yahoo! has 23 non-US global properties.
Supply-Side versus Demand-Size Economies of Scale. In the industrial economy, competition tended to produce large oligopolies; in the knowledge economy, competition produces natural monopolies. This difference is a function of distinct outcomes that are produced by different types of economics of scale.
On the one hand, in an industrial competition, positive feedback increased the economies of scale on the supply side as company's increased the size of their production runs, moving down the learning curve. However, as the economists Shapiro and Varian (1999, p. 179) pointed out, "Because traditional economies of scale based on manufacturing have generally been exhausted at scales well below market dominance . . . positive feedback based on supply-side economies of scale ran into natural limits, at which point negative feedback took over." The difficulties in managing enormous organizations put the brakes on positive feedback.
On the other hand, positive feedback on the demand-side is exponentially stronger. Natural monopolies form when knowledge companies have a sustainable competitive advantage that is generated either by proprietary knowledge or strong network effects. What is essential is achieving lock-in, by any means, including the initial contributions of luck, timing, or sponsorship, as well as any competitive advantages accruing from superior technology or first-mover effects.
In demand-side economies of scale, demand increases as the good becomes more popular, thereby generating increasing returns because the value of its products or services increases in parallel with it growing user base that continues to extend into new population pools. Moreover, according to Varian and Shapiro, growth on the demand-side also reduces the cost, making the cheaper product more attractive to new users-accelerating demand even more, creating a virtuous cycle of increasing returns. This difference between supply-side and demand-side economies of scale in the industrial and knowledge economies is enormous, increasing scale potential by orders of magnitude.
Mauboussin points out two immediate and critical implications of network effects as demand-side driven. First, once a network reaches critical mass, users want to join it at the expense of its competitors, creating nonlinear growth. Second, for networks based on intellectual property, there is no limit on market share, garnering market shares that are unthinkable in the industrial world. He concluded, "Natural monopolies are the progeny of strong network effects."
Mauboussin continued this contrast between the two economies by describing two ways to scale a business, that is, to grow sales at a faster rate than "costs." The older tactic, which employed supply-side economies of scale, used physical capital more efficiently. The newer tactic expands the value of information goods, which characteristically have High Upfront Costs and Low Incremental Costs, through demand-side economies of scale. Recall how expensive it is to write a software program and how cheap it is to replicate a second copy or a million copies. When network effects are united with the cost characteristics of information goods, their union is explosive: a surge in demand-driven sales is now coupled with low incremental costs, a marriage that creates extensive shareholder value.
The distinction between network effects and high upfront/low incremental costs clarifies when you focus on the trigger for additional costs of reinvestment. In a physical network, constraints on capacity trigger reinvestment; whereas, in an information network, reinvestment is triggered by obsolescence because fresh information is highly valued. According to Mauboussin (2000, p. 12), "But the key [to explosive growth] is that while an information-based product or service is current, sales and costs become uncoupled."
By seeing through the stock market tumult and upheavals in business models, the investor can focus on the central question: Where are the loci of competitive advantage in a rapidly shifting economy? Mauboussin (p. 12) believes that "networks are one of the few sources of sustainable competitive advantage." Once dominant, networks are difficult to dislodge because of aggregated switching costs, the cost a user bears when switching from one system to another. These switching costs include the loss of interactions with others, learning a new system, and establishing an identity. Imagine that Silicon Investor banned the G&K thread; what would that cost us, individually and collectively? In a network, costs are aggregated across users. Even small switching costs among many users create the equivalent of gigantic switching cost for one user. Now, you might imagine that you could entice your fellows to join you on a new thread on a new site as an alternative to Silicon Investor, but the coordination hurdle would be immense. If you are the only who goes, no new value is added. Networks have Laws: Of networks, there shall be few; and that virtuous few, once chosen, shall demonstrate increasing returns while diminishing the returns of their rivals.
The Significance of Achieving Critical Mass. Epidemics form an S-curve when cumulative infections are plotted over time. Disease, like any innovation, starts slowly and increases at an accelerating rate if it takes hold. What matters most in whether an epidemic will form is the amount of interaction between agents and their relative susceptibility. This analogy holds for the diffusion of all innovations.
In his most recent edition of "Diffusion of Innovations," Everett Rogers (1998, p. 10 ff) defined diffusion "as the process by which an innovation is communicated through certain channels over time among the members of a social system." His definition included the four main elements of diffusion: (1) The Innovation--"an idea, practice, or object that is perceived as new by an individual or other unit of adoption; Innovations that are perceived by individuals as having greater relative advantage, compatibility, trialability, observability, and less complexity will be adopted more rapidly than other innovations. " (2) Communication Channels-where communication "is the process by which participants create and share information with one another to reach a mutual understanding," and channels are "the means by which messages get from one individual to another;" (3) Time-involved in three ways: (a) in the innovation-decision process by which an individual passes from first knowledge of an innovation through its adoption or rejection, (b) in the innovativeness [susceptibility] of an individual or other unit of adoption, and (c) in an innovation's rate of adoption in a system [the S-curve]; and (4) A Social System-"a set of interrelated units that are engaged in joint problem-solving to accomplish a common goal." According to Rogers (p. 259), "The S-shaped curve of diffusion 'takes off' once interpersonal networks become activated in spreading subjective evaluations of an innovation from peer to peer in a system. The part of the diffusion curve from about 10 percent adoption to 20 percent adoption is the heart of the diffusion process. After that point [critical mass], it is often impossible to stop the further diffusion of a new idea, even if one wished to do so." "Critical mass occurs at the point at which enough individuals have adopted an innovation so that the innovation's further rate of adoption becomes self-sustaining" (p. 313).
When there are non-interactive or indirect network effects based on device complementarities, the S-curve is flatter because the growing mass of users attracts additional users through their positive subjective evaluations of a device's value-added functionality. When there are interactive or direct network effects, the S-curve is steeper because the value of the net-community is enhanced beyond the intrinsic value of the innovative idea, practice or object: as interactions becomes generative-an emergent system effect--the value of continuing interactions expand explosively, creating surplus value for all adopters, early as well as late. Strong or weak network effects refer to the rate of adoption or the probability of reaching critical mass.
The potential magnitude of an indirect network effect is a function of value-added by a device and how many nonadopters are untapped but in the social cluster of adopters. The potential magnitude of an interactive direct network effect is a function of value-generated by the net-community and how many nonadopters remain untapped in the social cluster of adopters. The power of network effects resides in the cascade of adoption through social clusters that are interconnected to other social clusters. The small-world effect and the Internet insure the rapid proliferation of adopted innovations and the dominance of locked-in networks.
Mauboussin (p. 16) summarized three key points: (a) "Interactive innovations, including networks, grow at a faster rate than noninteractive ones." (b) "An individual's adoption threshold is defined by how many other people in his or her social cluster engage in an activity before they join in as well." (c) "The small-world effect shows us how communications technology makes us more connected to one another than ever before."
Mauboussin noted that critical mass was completely mapped by three alternative descriptions: (1) The inflection point-or-elbow in an S-curve; (2) The transition from the early adopters to the early majority-or Moore's "crossing the chasm;" and (3) The tipping point, where incremental share comes at incrementally lower costs.
Because the concept of network effects" is nested in Moore's fundamental and integrated models, at home in the Valued-Based Economic Model and the model of the Technology Adoption Life Cycle, some gorilla gamers may want to consider the argument that network effects produce a sustainable competitive advantage. Any network effect that achieves critical mass becomes locked-in while it simultaneously creates high switching costs. Achieving a critical mass is described alternatively as reaching the tipping point, or the inflection point in the elbow of the S-curve, or entering hypergrowth after crossing the chasm. The assessment of the relative investment risk of lock-in that is achieved by proprietary open architecture or by network effects that have achieved critical mass is complex. Caution urges the risk-averse to side with the gorilla; whereas, the temptation of increased excess returns urges the risk-embracing to side with the light business models of large, expanding networks. Each investor must make his/her own choices.
Fifteen Takeaways: Exploitable Gaps in Expectations.
Lets review the exploitable gaps in expectations between traditional investors and knowledgeable investors that are generated by the VBEM and the sustainable competitive advantages of proprietary open architectures with high switching costs or locked-in network effects with high switching costs. The exploitable gaps in expectations include:
1. Inflection points create exponential growth, whereas most investors assume only linear growth.
2. Inflection points can be anticipated when diffusion reaches a10 to 20% level of technology adoption.
3. Excess returns (ROIC - WACC) create value, not accounting earning growth.
4. Excess returns drive stock prices, making those prices appear high to traditional investors.
5. Cash flow exceeds earnings growth for knowledge-based companies, and visa versa for industrial companies, which creates gaps in their perceived value.
6. Natural monopolies, which insure high GAPS (spreads) or excess returns by their pricing power, are common outcomes when knowledge companies achieve critical mass, igniting a pattern of increasing returns, which lengthens value-duration.
7. A network's mix of business assets affects how it scales: knowledge companies with high upfront and low incremental costs, in contrast to physical networks with moderate upfront and moderate incremental costs, scale more easily, creating a gap in expectations as they increase their operating margins rapidly and sharply upward.
8. Because demand-side scaling increases market share beyond levels that can be attained through traditional supply-side economies of scale, a gap in expectations about the level of attainable market share is often present.
9. Time, as a competitive advantage period, is a neglected value driver, causing traditional investors to underestimate how long a dominant company's value-duration will last, leading them to hold lightly rather than tightly when any slippage in price occurs.
10. Light business models generate cash for new investments, which can scale-up and scope-up far more rapidly and extensively than those with traditional expectations believe to be possible.
11. Interactivity drives direct network effects to levels of value that are not anticipated by traditional investors.
12. All knowledge companies, and particularly network companies, have the potential to create real options that are not being taken into account by traditional investors.
13. Management of the cash conversion cycle reduces the cost of capital, which drives value in ways that are often not anticipated by investors with an obsessive focus on accounting earnings.
14. Global scaling will increase the value of branded Internet service providers far beyond traditional expectations.
15. The exponential power of Gorillas or dominant networks will drive market capitalization of companies with a lock-in to previously unattained levels, creating a gap because traditional investor's continually expect a correction.
I hope this helps. I know writing it helped me.
Don |
