12. Heroes and Villains of the Globalization Era: Dot-Com Boom and Bust: When the Internet Became Big Business

My Name is Paul Allen: I Saw Possibilities Before They Had Names

I was born in Seattle, Washington, in 1953. I was the sort of kid who could disappear into science-fiction books and then spend hours wondering whether the machines in those stories might someday be possible. At Lakeside School, I met another student who shared my fascination with computers: Bill Gates. Computers were enormous, expensive machines in those days, and getting access to one felt almost like being allowed inside a laboratory of the future. Bill and I spent every hour we could programming. We were not simply interested in what computers could already do. We wanted to discover what they might do next.

The Computer on a Magazine Cover

I attended Washington State University but left before graduating because the computer revolution seemed to be moving faster than a classroom could contain it. Then, in late 1974, I saw something that changed everything: the Altair 8800, a small computer built around a microprocessor. I showed it to Bill. We immediately understood what it represented. Computers were becoming small enough and inexpensive enough that ordinary people might eventually own them. But machines needed software. We contacted the Altair's manufacturer and developed a version of BASIC that could run on it. That project helped launch Microsoft in 1975.

Microsoft Begins

Microsoft was not born as a giant corporation. It was a handful of people writing code, solving problems and trying to stay ahead of an industry that barely existed. Bill and I believed software would become extraordinarily important because every new generation of personal computers would need languages, operating systems and applications. We moved the company from Albuquerque back to the Seattle area and continued building software for the growing personal-computer market. I tended to look toward what might come next—the connections between computers, communications, information and entertainment—because computing never seemed to me like an isolated technology. It looked like the beginning of something much larger.

Leaving Microsoft, Not Technology

In 1982, I was diagnosed with Hodgkin lymphoma. That changes the way you think about time. I eventually left my day-to-day role at Microsoft in 1983, although I remained connected to the company as an owner and board member for years afterward. Leaving Microsoft did not mean I had stopped thinking about technology. Quite the opposite. It gave me the freedom to explore ideas that did not necessarily fit neatly inside one company. By the 1990s, Microsoft records listed me alongside ventures such as Asymetrix, Interval Research, Starwave and Vulcan Ventures. I wanted to investigate what would happen when computers, broadband communications, media and interactive entertainment began converging.

The Internet Was More Than Websites

During the 1990s, everyone began talking about the Internet, but I was interested in the larger system forming around it. The Internet was not simply a collection of webpages. It could become a distribution system for information, entertainment, communication and eventually almost every form of digital content. Through my investments and companies, I put money into cable systems, broadband, online media and technology businesses. Some ideas worked. Others arrived too early or never developed as expected. That is part of invention: you cannot explore the frontier while demanding that every expedition succeed.

After the Dot-Com Rush

When Internet companies began attracting enormous amounts of investment during the late 1990s, the excitement was understandable. Something genuinely revolutionary was happening. But revolutionary technology and successful businesses are not automatically the same thing. The dot-com crash eliminated many companies, yet it did not eliminate the Internet. Broadband expanded, digital media continued growing, and computing became increasingly connected. To me, that was the larger lesson: individual predictions can be wrong while the underlying direction remains right.

The Questions Became Bigger

Eventually my interests expanded beyond software and business. I funded scientific research, museums, conservation efforts and aerospace projects. In 2003, I helped establish the Allen Institute for Brain Science, beginning an effort to create large-scale, openly shared resources for understanding the brain. In 2011, I founded Stratolaunch because I remained fascinated by the possibility of changing how humanity reached space. The subjects had changed, but the instinct was the same: find a difficult problem, assemble talented people, give them tools and see whether something previously impossible could become possible.

Internet Changes from Research Network to Marketplace - Told by Paul Allen

By the middle of the 1990s, I had been watching computers evolve for more than twenty years. Bill Gates and I had started Microsoft because we believed computers would move from laboratories and corporations onto people's desks. Now I could see the next transition beginning. Those computers were connecting to one another, and the network connecting them was becoming something much larger than a tool for researchers. The Internet was beginning to look like a marketplace.

The Internet Had Been Built for Something Else

For years, much of the American Internet's growth had been driven by government and university research. NSFNET, created by the National Science Foundation in 1986, connected researchers and supercomputer centers and eventually became a major backbone of the growing Internet. Commercial activity had originally been restricted on the federally funded backbone. Private networks gradually appeared, however, and policies changed as commercial use expanded. By the early 1990s, businesses were beginning to realize that a network designed to exchange research could also exchange something else: products, advertisements, entertainment, services, and money.

Then the Web Made the Internet Visible

The Internet existed long before most Americans understood what it was. The World Wide Web changed that. Mosaic, developed at the National Center for Supercomputing Applications, helped transform the experience by allowing ordinary users to navigate graphical webpages without needing to understand the complicated commands that earlier Internet users often encountered. Suddenly you could point, click, read text, and see images together on a screen. That seems ordinary today. In 1993 and 1994, it felt like opening a door. If people could easily move around this information space, companies could begin building things for them there.

Businesses Began Asking the Important Question

The question changed quickly from "What is the Internet?" to "How do we make money on it?" That was not as easy as it sounds. The Web had been built around open standards, not around cash registers. Companies had to determine how advertising might work, how customers could find businesses, and most importantly, how people could safely send payment information across a network. In 1994, the CommerceNet consortium was established to encourage Internet commerce, and secure credit-card transactions were demonstrated that same year. The pieces of an online economy were starting to appear.

Silicon Valley Notices the Web

Entrepreneurs saw the opportunity quickly. In 1994, Jim Clark and Marc Andreessen created the company that became Netscape Communications. Andreessen had helped develop Mosaic, and the new company's Netscape Navigator browser made accessing the Web easier for millions of people. This was an important change in attitude. The browser was no longer merely an interesting research tool. It was becoming the doorway through which customers could enter an entirely new commercial environment. Silicon Valley money began following that possibility.

I Wanted to Know What People Would Do Once They Arrived

My own interest was not simply in getting people connected. I wanted to know what they would do after they were connected. Through Starwave, we experimented with delivering information and entertainment over the Web. In 1994, Starwave introduced an experimental sports-information service, and in April 1995 we worked with ESPN to launch ESPNET SportsZone, the predecessor of ESPN.com. Imagine what that meant at the time. Instead of waiting for tomorrow's newspaper or the evening sports report, fans could begin looking for updated scores and information through their computers. The Internet was becoming a medium, not merely a network.

April 1995: A Quiet but Enormous Change

One of the most important events of this transition did not involve a flashy new website. On April 30, 1995, the National Science Foundation decommissioned the NSFNET backbone. By then, commercial Internet providers and privately operated networks were capable of carrying the traffic. The government-supported research backbone that had helped the Internet grow was no longer needed as the center of the system. By the end of 1994, NSFNET had been moving 17.8 trillion bytes of information per month; by 1995, roughly 100,000 public and private networks were operating around the country. The infrastructure of the Internet was moving decisively toward private commercial operation.

The Storefronts Were Appearing

At almost the same moment, entrepreneurs began building businesses that existed primarily on the Internet. Amazon began selling books online in 1995. eBay appeared that year as well. Thousands of entrepreneurs started imagining businesses without traditional storefronts, catalogs, or distribution systems. The Web offered something extraordinary: a small company with a computer server could potentially reach customers across the country—or eventually across the world. That possibility would soon attract billions of dollars from investors.

My Name is Don Valentine: I Bet on Markets Before They Became Obvious

I was born in New York City in 1932. I grew up in the Bronx and eventually studied chemistry at Fordham University, graduating in 1954. Chemistry taught me something useful: understand what something is made of before you decide what it can become. After college and military service, I found my way to California. The weather was certainly better, but that was not what kept me there. Electronics was changing, aerospace was booming, and California was becoming a place where technical people were building things that had never existed before. That interested me considerably more than following an established industry back East.

Selling the Future at Fairchild

I worked as a sales engineer and eventually joined Fairchild Semiconductor around 1960. That was where I received my real education in technology. Silicon semiconductors were beginning to transform electronics, but inventing a product was only half the job. Someone had to determine who needed it, how large the market could become, and how to sell it. I spent years building Fairchild's sales operation and watching engineers create increasingly powerful components. I learned to look beyond the technology itself. A clever invention serving a tiny market might remain a clever invention. A useful product entering an enormous market could create an industry.

National Semiconductor and Learning to Build Companies

In 1967, I became a founding vice president of sales and marketing at National Semiconductor. By then I understood that technology companies did not succeed simply because their engineers were intelligent. They needed customers, distribution, discipline, pricing, management, and a reason for people to buy what they were producing. I began personally investing in young technology companies while I was still working in the semiconductor business. Entrepreneurs came looking for advice and money. Eventually I realized there was a business hiding inside those conversations.

Building Sequoia

In 1972, I started the venture operation that became Sequoia Capital. The name mattered. A sequoia grows enormously large and can survive for generations. That was the kind of company I wanted to finance. I was not interested in giving money to every bright engineer with an interesting idea. I wanted companies addressing markets large enough to matter. Our first venture fund totaled about $3 million, and among our early investments were Atari and Apple. Later came businesses such as Oracle, Electronic Arts, LSI Logic, and many others. Venture capital was still a small business then. There was no established Silicon Valley formula. We were developing the formula while we were using it.

Products, Markets, and People

People sometimes assume venture capitalists invest primarily in charismatic founders. That was never my starting point. I wanted to know the market. How large could it become? Who cared about the product? What problem was being solved? Management could sometimes be strengthened; a nonexistent market was considerably harder to repair. I could be demanding because the market was demanding. Entrepreneurs were asking investors to risk money on businesses that frequently had little more than technology, ambition, and a prediction about the future. My job was not to applaud the presentation. My job was to determine whether the prediction made sense.

Cisco and the Road to the Internet

Cisco became one of the investments that best demonstrated that philosophy. Sequoia financed Cisco in the 1980s when computer networking was still poorly understood outside technical circles. But the problem was becoming obvious: computers were multiplying, networks were multiplying, and those networks needed ways to communicate with one another. Cisco was building equipment that addressed a problem capable of becoming enormous. I served as chairman as the company grew, and by the 1990s networking had become part of the infrastructure underneath the Internet revolution. That is the kind of opportunity investors spend their careers trying to recognize before everyone else recognizes it.

Watching Silicon Valley Become Silicon Valley

By the 1990s, the small technology community I had entered decades earlier had become an economic phenomenon. Money poured into technology. Companies went public quickly. The Internet created another generation of entrepreneurs convinced they could reinvent entire industries. Some were right. Others merely had expensive ideas financed during an unusually enthusiastic period. Sequoia continued investing, but the fundamental question never changed: Is this a company capable of becoming important, or merely a company participating in something fashionable?

The Netscape IPO Ignites the Dot-Com Gold Rush - Told by Don Valentine

I had been investing in technology companies long enough to know that Silicon Valley occasionally experiences a moment when everybody suddenly sees the same opportunity. It happened with semiconductors. It happened with personal computers. On August 9, 1995, another one of those moments arrived. A 15-month-old company called Netscape Communications went public, and before that trading day was finished, Wall Street had demonstrated just how much money investors were willing to place behind the Internet.

A Browser Becomes a Business

Netscape had been founded by Jim Clark and Marc Andreessen in 1994. Andreessen had helped develop Mosaic, one of the graphical browsers that made navigating the World Wide Web considerably easier. Netscape's Navigator browser quickly became dominant, with contemporary estimates putting its share of the browser market around 70 percent and its users around seven million. Much of Navigator had been distributed free because Netscape hoped the browser would establish its position while the company earned money from commercial software and Internet servers. To an investor, the important fact was not simply that people liked the product. Netscape was positioning itself at the entrance to a potentially enormous new market.

Wall Street Could Barely Price It

The original IPO prospectus contemplated shares at roughly $12 to $14. Demand became so intense that the price was raised repeatedly, eventually reaching $28, while the offering itself increased from 3.5 million shares to five million. By opening day, investors reportedly had placed orders for approximately 100 million shares. Think about that ratio: five million shares available and demand for roughly twenty times that amount. You did not need an elaborate financial model to understand what was happening. Investors were no longer merely buying Netscape. They were trying to buy a position in the Internet itself.

Then the Bell Rang

When Netscape finally began trading on August 9, the stock did not open at $28. It opened at $71. During the first hour it reached about $75 before falling back, and it eventually closed at $58.25. More than 13.8 million shares changed hands during the day. At the closing price, Netscape's total market value was estimated at roughly $1.9 billion. This was a company that had never reported a profit and had generated only $16.6 million in revenue during the first six months of 1995. Those numbers should make any investor stop and think.

What Investors Were Really Buying

Traditional investors liked profits, assets, established customers, and predictable cash flow. Netscape forced Wall Street to consider something different: future market position. Investors were betting that the Internet would become enormous and that companies establishing themselves early might someday control extremely valuable pieces of it. That reasoning was not foolish. The Internet really was beginning a historic expansion. But there is an enormous difference between correctly predicting the future of an industry and correctly valuing every company participating in it. That distinction would become increasingly important.

Silicon Valley Heard the Starting Gun

Entrepreneurs noticed what happened immediately. So did venture capitalists. Netscape demonstrated that a young Internet company could move from startup to public corporation with astonishing speed. Jim Clark himself later described the IPO partly as a marketing event, while venture capitalist John Doerr had recognized early that graphical Internet navigation represented a potentially enormous opportunity. After the IPO, founders across Silicon Valley could point toward Netscape and make a very attractive argument: fund us now, help us grow rapidly, take us public, and perhaps Wall Street will value our company the same way.

The Rules Begin to Change

This was where things became interesting—and dangerous. If investors rewarded growth faster than profitability, entrepreneurs naturally concentrated on growth. If market share produced enormous valuations, companies would spend aggressively to acquire market share. Venture capital became fuel. The public markets could provide even more. A startup no longer necessarily had to spend many years becoming profitable before Wall Street would listen. It could instead tell investors a compelling story about how large its future market might become. Netscape had shown that the market was willing to hear that story.

Gold Had Been Discovered

I would not say Netscape single-handedly created the dot-com boom. Technology stocks were already rising, Internet usage was already expanding, and venture investors were already financing Internet businesses. But August 9, 1995, became an unmistakable signal. The Internet was no longer merely a technological revolution. It had become an investment phenomenon. Money would now race toward browsers, search engines, online stores, advertising companies, Internet service providers, networking businesses, and ideas that had barely progressed beyond a business plan. Some would build extraordinary companies. Others would burn through extraordinary amounts of money. That is what happens in a gold rush: discovering that gold exists is only the beginning. The next question is how many people will convince themselves that every patch of dirt contains it.

Venture Capital Floods into Internet Startups - Told by Don Valentine

I had been investing in technology companies for decades, and by 1995 something was changing in Silicon Valley. Entrepreneurs were no longer walking into venture-capital offices primarily with ideas for semiconductors, computers, or traditional software. Now they were carrying plans for Internet directories, online stores, communications networks, advertising businesses, and services that had no obvious equivalent in the physical world. After Netscape's spectacular 1995 public offering, investors understood that Wall Street was willing to put extraordinary values on young Internet companies. The race to find the next one had begun.

The Money Starts Moving

Venture capital was already growing rapidly. American venture investors put roughly $7.4 billion into companies during 1995, nearly half of it in information technology. In 1996, investment climbed to a record $10.1 billion across more than 2,100 deals, with technology businesses receiving about 60 percent of the money. By 1997, total venture investment had risen to roughly $12.8 billion. Those numbers mattered because venture capital is fuel. Give an ambitious technology company enough fuel and it can hire engineers, buy equipment, advertise nationally, and attack a market before established competitors understand what is happening.

Yahoo! Shows What We Were Looking For

At Sequoia, one of the companies demonstrating this new opportunity was Yahoo!. Jerry Yang and David Filo had begun organizing websites while at Stanford because the Web was expanding so rapidly that people needed help finding things. Yahoo! was founded in 1994, partnered with Sequoia in 1995, and went public in 1996. That progression tells you how quickly the Internet was compressing the traditional business timetable. A useful idea created by graduate students could become a venture-backed company and then a publicly traded corporation in only a few years. Investors looked at Yahoo! and began wondering how many other entirely new categories were waiting to be created.

Internet Investment Explodes

The change became unmistakable by 1997. Private venture funding for Internet-related businesses had been only about $134 million in 1995. Two years later it reached approximately $1.88 billion—roughly fourteen times as much. During the second quarter of 1997 alone, Internet companies attracted more than $500 million in venture investment for the first time, while technology companies captured about 70 percent of all venture money invested during that quarter. Silicon Valley received more than $1 billion. The Internet had gone from an interesting investment category to one of the hottest places in American finance.

Every Entrepreneur Suddenly Had a Plan

Once investors demonstrate that they are willing to finance something, entrepreneurs become remarkably good at producing proposals for it. Conferences filled with people calling themselves Internet entrepreneurs. In June 1997, hundreds gathered at VentureNet '97 in California to pitch Internet business plans and meet potential investors. Some ideas were excellent. Others depended primarily on the assumption that Internet traffic would somehow become money. That distinction was important. A website was not automatically a company, and attracting visitors was not automatically a business model. Venture capitalists were supposed to determine the difference.

Investors Began Competing with Each Other

Something else was happening that worried experienced investors. Normally, entrepreneurs compete for capital. During a boom, capital begins competing for entrepreneurs. A venture firm could reject a questionable Internet deal only to discover that another investor had funded it immediately. Investors had more money to put to work, institutions were committing additional capital to venture funds, and successful public offerings made the possible rewards look enormous. By 1997, foreign corporations were also moving into Silicon Valley venture investing because they wanted access to Internet and multimedia startups. When too much money is chasing a limited number of exceptional companies, standards have a tendency to weaken.

The Internet Made Speed Valuable

There was a legitimate reason investors accepted enormous risks. Internet markets appeared capable of developing very quickly. If a company became the place where people searched, shopped, communicated, or exchanged information, being early could create a substantial advantage. Venture money allowed startups to grow before slower corporations could respond. In 1997, investors were even putting tens of millions of dollars into companies such as Juniper Networks before they had released their principal products because the demand for faster Internet infrastructure looked potentially enormous.

That is what venture capital is supposed to do: finance the future before the future becomes obvious.

The problem begins when investors stop distinguishing between financing the future and simply financing excitement.

1996–1998 — Internet Companies Become Household Names - Told by Paul Allen

By 1996, the Internet was no longer something Americans only heard about from computer enthusiasts. People were beginning to recognize company names, type Web addresses into browsers, and discover that the Internet could help them find information, buy products, follow sports, send messages, and even sell things to strangers. What interested me most was that the network itself was beginning to disappear into the background. People were no longer fascinated simply because they were online. They were beginning to care about what they could do once they got there.

Yahoo! Helps People Find Their Way

One of the clearest examples was Yahoo!. Jerry Yang and David Filo had started organizing websites while they were graduate students at Stanford, and Yahoo! grew into a directory and portal that helped people navigate an Internet that was expanding faster than anyone could organize it. The company began selling advertising in 1995 and went public on April 12, 1996. For many new users, Yahoo! became something like the front desk of the Internet. You could arrive without knowing exactly where you wanted to go and begin exploring from there.

Amazon Tries to Replace the Bookshelf

Then there was Amazon. Jeff Bezos had begun selling books online in 1995, and by 1996 and 1997 the company was demonstrating just how quickly an Internet business could grow. Amazon's sales increased from $15.7 million in 1996 to $147.8 million in 1997, while its customer accounts rose from about 180,000 to more than 1.5 million. That was remarkable because Amazon was asking customers to change a familiar habit. Instead of walking into a bookstore, browsing shelves, and carrying a book to the counter, people could search a computer screen and have the book delivered to their homes. Once consumers became comfortable doing that with books, the obvious question was: what else might they buy online?

AOL Puts the Internet in the Living Room

America Online played a different role. AOL did not simply offer access to the Internet; it tried to make online life understandable to ordinary families. Its software, chat rooms, email, news, entertainment, and famously widespread installation discs helped bring millions of people online. The important change was cultural. A computer connected to a network was becoming an ordinary household appliance. You no longer needed to be an engineer or university researcher to communicate online. Parents, teenagers, small-business owners, sports fans, and shoppers were beginning to join the same digital world.

eBay Lets Strangers Become Merchants

Another extraordinary experiment began as AuctionWeb in 1995 and became eBay. Pierre Omidyar created a marketplace where individuals could offer goods directly to other individuals. By the end of 1998, eBay reported roughly two million registered users, and that September the company went public. What fascinated me was not simply the auction technology. The Internet was creating trust between people who had never met. Someone in California could offer an object for sale, and someone hundreds or thousands of miles away could bid on it. The marketplace did not need to own the merchandise. It needed to connect the buyer and seller.

Media Begins Moving onto the Screen

I was watching the same transformation through Starwave. We had been asking a simple question: what information becomes more valuable when it can constantly change? Sports was an obvious answer. Scores, statistics, injuries, standings, and breaking news do not belong naturally in yesterday's newspaper. Starwave and ESPN launched the site that became ESPN.com in 1995, and by 1998 it was receiving about 1.2 million visitors a day. Suddenly a fan did not have to wait for television highlights or the next morning's paper. The screen on the desk was becoming a source of continuously updated media.

The Internet Starts Creating Brands

This was the real transformation between 1996 and 1998. Yahoo!, Amazon, AOL, eBay, Netscape, and other Internet companies were no longer simply technology businesses known within Silicon Valley. Their names were appearing in advertisements, newspapers, television reports, investment discussions, and ordinary conversations. Some companies became verbs, destinations, or habits. Businesses began spending heavily to make sure consumers remembered their names because being first in someone's mind could be almost as valuable as being first with the technology.

Something Larger Was Taking Shape

I had watched the personal-computer revolution begin with machines that most people considered toys for hobbyists. Then software turned those machines into tools millions of people needed. The Internet was following a similar path, only faster. Between 1996 and 1998, companies proved that the network could support media, shopping, advertising, auctions, communications, and entirely new kinds of businesses.

My Name is Andrew Grove: Survive the Crisis, Then Build the Future

I was born András István Gróf in Budapest, Hungary, in 1936. My childhood taught me very early that stability is temporary. I was a Jewish child during the Holocaust, when survival depended upon hiding and adapting to circumstances that could change without warning. After the war, Hungary fell under Communist rule, and in 1956 the Hungarian Revolution erupted against Soviet control. When Soviet forces crushed the uprising, I decided to leave. At twenty years old, I escaped across the border into Austria and eventually made my way to the United States. I arrived with little money, limited English, and one important understanding: circumstances can change faster than you think.

Becoming an Engineer in America

America gave me an opportunity, but opportunity is useless unless you work. I enrolled at the City College of New York, studied chemical engineering, and graduated in 1960. I later earned my doctorate from the University of California, Berkeley. Engineering appealed to me because facts matter. A semiconductor either performs or it does not. A manufacturing process either produces acceptable results or it fails. You cannot manage technology through wishful thinking. You identify the problem, measure it, understand it, and improve it. That attitude would eventually shape the way I managed companies as well.

Fairchild and the Semiconductor Revolution

After graduate school, I joined Fairchild Semiconductor, one of the companies helping create what would become Silicon Valley. Semiconductor technology was advancing quickly, and I became deeply involved in manufacturing and research. In 1968, Robert Noyce and Gordon Moore left Fairchild to create Intel, and I joined them shortly afterward as the company's first employee after the founders. Intel began as a semiconductor company focused largely on memory chips. We were small, ambitious, and operating in an industry where yesterday's technological advantage could disappear tomorrow.

When Intel Had to Change

The most important decisions in business are often made when the comfortable choice is no longer the correct one. By the early 1980s, Japanese semiconductor manufacturers were competing aggressively in memory chips, and Intel's traditional business was under enormous pressure. Gordon Moore and I eventually confronted a difficult question: if new management took over Intel, what would they do? The answer was uncomfortable—they would probably get out of memory. So we did. Intel shifted its resources toward microprocessors. It was painful, but companies that refuse to abandon yesterday's success can disappear with it.

Running Intel Like Survival Mattered

I became Intel's president in 1979 and CEO in 1987. I believed management required discipline, measurement, debate, and a willingness to confront unpleasant information before it became a catastrophe. Employees were expected to challenge ideas, even when those ideas came from senior executives. I called this constructive confrontation. A company where everyone agrees with the boss may feel comfortable, but comfort is dangerous in technology. Markets change. Competitors improve. Customers develop new expectations. The moment an organization believes success is permanent is often the moment its decline begins.

The Microprocessor Becomes the Engine

During the 1980s and 1990s, personal computers spread into businesses, schools, and homes, and Intel's microprocessors became central to that expansion. The famous "Intel Inside" campaign helped ordinary consumers recognize a component they previously would never have considered when purchasing a computer. As PCs became more powerful, software became more sophisticated, and eventually the Internet created another enormous wave of demand. Websites may have received the attention, but underneath the dot-com boom were millions of computers, servers, networking systems, and processors performing the work. The Internet economy still depended upon physical machines.

Only the Paranoid Survive

I often spoke about what I called strategic inflection points—moments when the basic rules of a business change. The Internet was one of those moments. During the 1990s, entrepreneurs and investors were racing to build Internet companies, and many believed the growth would continue almost without limit. Some were building extraordinary businesses. Others were simply following enthusiasm. My instinct was always to ask what fundamental change was occurring underneath the excitement. A bubble may disappear, but a genuine technological transformation can continue long after the speculation ends.

Building the Physical Internet Economy - Told by Andrew Grove

When most people looked at the Internet in the late 1990s, they saw websites. I saw machines. Every search, email, online purchase, photograph, and webpage had to be processed somewhere, stored somewhere, and carried across physical equipment. The Internet might have appeared almost magical on a computer screen, but there was nothing magical about the infrastructure underneath it. It required microprocessors, servers, routers, switches, telephone lines, fiber-optic cables, and enormous investments in computing capacity. Between 1997 and 1999, America was not simply building websites. We were constructing the machinery of a new economy.

The Internet Needed More Powerful Machines

At Intel, we understood that connecting computers would increase the amount of work those computers were expected to perform. In 1997, we introduced the Pentium II processor, and we were developing versions specifically suited for servers and workstations as well as desktop computers. Internet and intranet servers needed speed, reliability, and the ability to handle increasing amounts of information. By 1999, Intel's Pentium III and Xeon families were pushing further into those markets. Industry estimates cited by Intel indicated that approximately 80 percent of Internet and intranet servers shipped in 1998 used Intel architecture. The Internet revolution was creating demand not only for websites but for the computers hidden behind those websites.

The Router Becomes Part of the New Industrial Machinery

Processors could perform calculations, but networks also needed equipment capable of directing information to the correct destination. That was where companies such as Cisco became essential. Routers examined digital information and helped determine where it should travel next across interconnected networks. Cisco reported selling its one-millionth router in 1997. Its annual revenue climbed from about $6.5 billion that year to $12.2 billion by 1999. Those numbers tell you something important about the Internet boom. The businesses selling products on the Web received much of the attention, but enormous companies were also being built by supplying the equipment that allowed those websites to exist.

America Begins Laying Digital Highways

Then came bandwidth. You could build the fastest computer in the world, but if information traveled through an inadequate connection, the network would still feel slow. Telecommunications companies therefore began pouring billions of dollars into fiber-optic systems and Internet backbones. Level 3 announced plans to invest roughly $3 billion in broadband deployment, including about 15,000 miles of fiber-optic cable. UUNET announced a $300 million expansion of its backbone infrastructure in 1997. The Federal Communications Commission reported that available Internet backbone bandwidth, which only a few years earlier had doubled about once a year, was by this period doubling approximately every four to six months.

Think of what was happening. Beneath streets, alongside highways, between cities, and inside office buildings, companies were installing the physical pathways through which the Internet economy would travel. These were the railroads and highways of the digital age.

Behind Every Website Was a Room Full of Equipment

When an Internet company attracted thousands and then millions of visitors, somebody had to keep the service running. Companies needed racks of servers, storage systems, backup equipment, network connections, electrical power, cooling, and technicians capable of keeping everything operational. Every successful Internet company created demand farther down the supply chain. More customers meant more traffic. More traffic meant more servers. More servers meant more processors, networking equipment, storage, electricity, and bandwidth.

This was one reason I considered the Internet a strategic change rather than a temporary fashion. Intel's own 1998 annual report explicitly connected future growth in computing with increased Internet usage and expanding Internet products. The Web was beginning to create its own cycle of demand: better computers encouraged richer Internet services, and richer Internet services encouraged people and companies to purchase better computers.

Intel Becomes an Internet Business Too

The transformation was not confined to Internet startups. Established corporations were beginning to move their operations onto networks as well. Intel offered a good example. In 1997, essentially none of our business with customers was being conducted through the Internet. In 1998, more than 20 percent was. By 1999, we estimated that more than 40 percent of Intel's business would be transacted that way—representing roughly $12 billion to $13 billion in revenue. That was not an experiment anymore. The Internet was entering purchasing, supply chains, customer service, manufacturing relationships, and corporate management.

The Invisible Side of the Dot-Com Boom

By 1999, people could look at Yahoo!, Amazon, eBay, and hundreds of new Internet companies and conclude that the dot-com boom was about websites. That was only the visible layer. Underneath them stood semiconductor factories, server manufacturers, networking companies, telecommunications carriers, software developers, and thousands of miles of new fiber. A 1999 University of Texas study sponsored by Cisco estimated that the broader Internet economy was generating about $507 billion in annual revenue and supporting approximately 2.3 million American jobs.

That physical investment would prove important when the excitement eventually faded. Websites could disappear overnight. A failed company could shut its doors. Stock prices could collapse. But much of the infrastructure being constructed remained behind. The processors, fiber, routers, servers, technical knowledge, and networks did not suddenly vanish when investors changed their minds.

My Name is Sir John Templeton: I Looked Where Others Were Afraid to Look

I was born John Marks Templeton in Winchester, Tennessee, in 1912. My parents taught me thrift, independence, and the importance of making careful use of what I had. During the Great Depression, I attended Yale University and supported myself while completing my education. I graduated in 1934 and then went to Oxford as a Rhodes Scholar, earning a degree in law. Those years impressed upon me a lesson that would guide my investing career: scarcity forces you to distinguish between what is fashionable and what is truly valuable.

Buying When Everyone Else Was Selling

I began working on Wall Street in 1938, just as the world was approaching another terrible war. When war broke out in Europe in 1939, fear dominated the markets. I saw something different. I borrowed money and purchased 100 shares in each of 104 companies whose stocks were selling for one dollar or less, including companies already in bankruptcy. Only four ultimately became worthless. I was not predicting that war was good. I was recognizing that investors had allowed fear to push many prices below reasonable value. The greatest bargains are rarely found where everyone feels comfortable.

Searching the Entire World

I never understood why an investor should limit himself to companies located in his own country. If the purpose of investing is to find exceptional opportunities at attractive prices, then one ought to search wherever those opportunities exist. In 1954, I established the Templeton Growth Fund and increasingly invested internationally, searching Europe, Asia, and other markets that many American investors ignored. The principle was simple: do not ask where everyone else is investing. Ask where the best value can be found.

Maximum Pessimism

Over time, people came to associate me with investing at the "point of maximum pessimism." That does not mean buying something simply because everyone dislikes it. A bad company can always become worse. It means doing your homework and then looking for situations where fear has become greater than the facts justify. When investors are euphoric, bargains become scarce. When investors are frightened, opportunities frequently appear. Successful investing therefore requires something emotionally difficult: you must sometimes be willing to stand apart from the crowd.

When the Internet Became the Crowd

By the late 1990s, I had seen enough market cycles to recognize another extreme forming. Internet companies were transforming business, and I certainly did not doubt the importance of the technology. But investors had begun treating almost any company associated with the Internet as though extraordinary future profits were guaranteed. That is dangerous. A wonderful technology does not automatically make every company using it a wonderful investment. During the dot-com mania, I took positions against a number of highly valued technology stocks, particularly around the expiration of restrictions that had prevented company insiders from selling their shares. My concern was not whether the Internet would survive. It was whether investors were paying sensible prices for Internet companies.

The Bubble Breaks

In 2000, enthusiasm began colliding with economics. Companies that had been valued at extraordinary levels suddenly had to answer ordinary questions: Where are the profits? How much cash are you losing? What prevents a competitor from replacing you? The dot-com collapse destroyed enormous amounts of paper wealth, but the Internet continued changing society. That distinction is important. Investors frequently make the mistake of believing that because a new technology will transform the world, every business attached to that technology must be worth almost any price. History repeatedly demonstrates otherwise.

Investing Beyond Money

As I grew older, money itself interested me less than what could be accomplished with it. I established the Templeton Prize in 1972 and the John Templeton Foundation in 1987. That same year, Queen Elizabeth II made me a Knight Bachelor for my philanthropic work. I became increasingly interested in science, spirituality, human character, and questions about how little humanity actually understands. Investing had taught me intellectual humility. Markets punish people who believe they know everything, and life often does the same.

“Get Big Fast” Replaces Traditional Business Thinking - Told by Don Valentine

By 1998, Silicon Valley had begun rewriting one of the oldest rules in business. Traditionally, you built a product, found customers, controlled expenses, generated profits, and expanded as those profits allowed. Internet companies were beginning to reverse that order. Raise enormous amounts of money. Spend it. Acquire customers. Expand nationally. Establish the brand. Capture the market before somebody else does. Profit could come later. Investors had a phrase for the philosophy taking hold: get big fast.

Amazon Shows Everyone the Strategy

Amazon became the clearest example. Jeff Bezos understood that online retail might eventually produce a handful of dominant companies, and he was determined to establish Amazon before competitors could catch him. The company had adopted "Get Big Fast" as a motto years earlier, but by 1998 the strategy was visible in the numbers. Amazon's sales soared from about $148 million in 1997 to $610 million in 1998. Its customer accounts jumped from 1.5 million to more than 6.2 million. But its net loss also climbed to approximately $124.5 million. The company was deliberately reinvesting money into warehouses, technology, advertising, new products, and expansion instead of concentrating on immediate profits.

To an old venture capitalist, that was not automatically irrational. If a company was genuinely building an enormous market position, spending money could be exactly the right thing to do. The problem came when everyone copied the spending without possessing the business underneath it.

Market Share Became More Important Than Profit

Internet executives began talking constantly about market share, traffic, registered users, page views, and "eyeballs." The reasoning was straightforward. If millions of customers became accustomed to using your website first, competitors might have great difficulty taking them away. A company could spend heavily today, dominate its category tomorrow, and eventually raise prices, sell advertising, reduce marketing expenses, or introduce additional products.

That logic worked particularly well when a business benefited from scale or network effects. A marketplace becomes more useful when more buyers and sellers participate. A portal attracts advertisers when millions of people visit it. An online retailer gains purchasing power and customer information as it grows. But those principles were increasingly applied to companies where the supposed advantage was little more than being first to spend the most money.

The Race for Customers Gets Expensive

By 1999, Internet companies seemed to be everywhere because they were paying enormous amounts to make sure you noticed them. During just the first nine months of 1999, online companies spent approximately $1.4 billion advertising through traditional television, newspapers, magazines, radio, and billboards—more than twice the roughly $649 million they had spent during all of 1998. Some newly funded startups were putting enormous portions of their investment capital directly into advertising.

That spending was not necessarily about selling enough merchandise that month to cover the advertisement. It was about planting a company name inside the customer's head. Investors were financing businesses that hoped brand recognition would become an asset valuable enough to justify the losses incurred creating it.

I had spent decades evaluating companies by asking about markets. But a market is not the same as an audience. Ten million people can visit your website and still leave you with a terrible business if it costs more to attract and serve those customers than you will ever earn from them.

Amazon Keeps Accelerating

Amazon demonstrated just how dramatic the growth-first model could become. During 1999, its sales reached approximately $1.64 billion, up 169 percent from the previous year, and its customer accounts approached 17 million. Yet Amazon also reported a pro forma net loss of roughly $390 million for the year. The company continued expanding into electronics, toys, auctions, international markets, and additional product categories.

There was an important distinction, however. Amazon was losing extraordinary amounts of money, but it was also building warehouses, technology, customer relationships, distribution capabilities, and a brand customers were actually using. Many entrepreneurs saw Amazon's losses but failed to understand the machinery being constructed behind them. They copied the red ink without necessarily copying the underlying value.

Wall Street Begins Rewarding Speed

Once public investors demonstrated that they would assign enormous valuations to companies growing rapidly despite continuing losses, management incentives changed. Executives learned that increasing revenue by 100 or 200 percent might impress investors more than producing a modest profit. Venture capitalists knew that if a startup could demonstrate enough growth, an IPO might provide another enormous infusion of capital.

That created a powerful cycle. Venture investors supplied money. Startups spent it to acquire customers. Rapid customer growth increased valuations. Higher valuations attracted more investment. More investment financed more expansion. As long as stock prices continued rising, the strategy appeared brilliant.

The Difference Between Investing and Financing a Race

There was nothing inherently wrong with growing quickly. Great companies should attack large opportunities aggressively when the opportunity exists. But by late 1999, "get big fast" was becoming something more dangerous: an excuse for ignoring ordinary economic discipline. Companies could claim that losses were evidence of ambition rather than evidence of a flawed business model. Spending became strategy. Growth itself became proof of success. That should make an investor uncomfortable.

The Internet really was changing business. Amazon really was demonstrating that online commerce could grow at astonishing speed. But a technological revolution does not repeal arithmetic. Eventually every company has to create more economic value than it consumes.

Between 1998 and 1999, Silicon Valley temporarily began behaving as though that reckoning could be postponed indefinitely. Entrepreneurs were no longer simply asking, "Can we build a profitable company?" They were asking, "Can we become enormous before anyone asks us about the profits?"

For a while, Wall Street answered yes.

IPO Mania, Day Trading, and Dot-Com Millionaires - Told by Sir John Templeton

I had spent more than sixty years studying markets by the time the Internet boom reached its most extraordinary stage. I had seen depression, war, inflation, recessions, bull markets, and crashes. Each period looked different, but human nature changed remarkably little. By 1999, investors had discovered a genuinely revolutionary technology—and then began convincing themselves that almost any price was reasonable if the company had some connection to the Internet. That was when I became especially interested. Great inventions can transform civilization. They can also produce terrible investments when enthusiasm outruns value.

1999 Becomes the Year of the IPO

An initial public offering had traditionally been an important but uncertain step in a company's development. In 1999, IPOs began resembling public celebrations. Research later calculated that the average first-day return for U.S. IPOs during 1999–2000 reached approximately 65 percent, compared with about 15 percent from 1990 through 1998. About one-quarter of the companies going public in 1999 doubled in price on their first trading day. Investors were not patiently buying businesses they expected to own for twenty years. Many were desperately trying to obtain shares before somebody else drove the price higher.

Then Came VA Linux

If you want to understand the atmosphere, consider December 9, 1999. VA Linux Systems offered shares to investors at $30 each. When trading began, the stock opened at $299 and eventually closed at $239.25—a first-day increase of 698 percent. The company had suddenly achieved a market value approaching $10 billion. When people watch fortunes appear in a single afternoon, something happens psychologically. The question stops being, "Is this company worth this price?" and becomes, "How do I get into the next one before everyone else?"

The Stock Market Moves into the Home

The Internet was also changing who could participate. Online brokerage accounts allowed ordinary investors to research stocks, view prices, and place trades from their own computers. The number of online accounts rose from about 3.7 million in 1997 to 7.3 million in 1998 and 9.7 million in 1999. By the first half of 1999, online transactions accounted for approximately 37 percent of retail stock trades. This was a remarkable democratization of investing. But making an investment easier to execute does not make the investment easier to understand.

Day Trading Turns Investing into a Daily Contest

Some people stopped thinking like long-term owners altogether. They became day traders, buying and selling shares within minutes or hours while attempting to profit from small movements in price. By September 1999, the Securities and Exchange Commission was publicly warning Americans about the risks. Regulators estimated fewer than 7,000 people were operating as full-time day traders, although millions more were trading online and many engaged in shorter-term speculation. Some traders borrowed money to increase their positions, which could magnify gains but could just as easily magnify losses. Technology had shortened the distance between an investor's impulse and the execution of a trade to a few clicks.

A Nation of Paper Millionaires

Inside technology companies, another extraordinary phenomenon was occurring. Founders, executives, programmers, and early employees frequently owned shares or stock options. When an IPO sent a company's valuation soaring, people who had been ordinary employees could suddenly calculate personal fortunes worth millions of dollars—at least on paper. But many insiders could not immediately sell their stock. IPO agreements commonly imposed lockup periods, often lasting roughly six months, before insiders could sell. That distinction between wealth on a computer screen and cash in the bank would soon become very important. Research on Internet IPOs later found substantial price pressure surrounding the expiration of those lockups and subsequent insider selling.

The Nasdaq Climbs Almost Straight Up

By the end of 1999, the excitement had become visible in the broader market. The technology-heavy Nasdaq Composite rose approximately 86 percent during that single year, its greatest annual percentage increase to that point. Yet the Federal Reserve noted that the gains were far from evenly distributed: more than half of the companies in the S&P 500 actually declined during 1999. Technology was carrying the excitement. When one part of a market begins behaving as though ordinary standards no longer apply, an investor should become more careful, not less.

I Began Looking for the Exit

I did not believe the Internet was a fad. That was precisely the point. Electricity had transformed civilization too, but that did not mean every company formed during the electrical revolution became a wonderful investment. By early 2000, I believed many technology shares had reached extraordinary valuations. I therefore began shorting selected technology stocks around the expiration of their IPO lockup periods, reasoning that insiders finally permitted to sell might provide additional selling pressure. Accounts of my strategy describe positions in 84 Nasdaq stocks, averaging roughly $2.2 million each.

Corporate America Rushes onto the Internet - Told by Andrew Grove

By 1999, I was telling business leaders something many of them did not particularly want to hear: soon, there would be no meaningful distinction between an "Internet company" and an ordinary company. Businesses would either learn to use the Internet in their operations or risk becoming less competitive. The Internet was not simply creating new companies such as Amazon and Yahoo!. It was beginning to force America's old companies—manufacturers, banks, retailers, airlines, brokers, and industrial corporations—to reconsider how they bought, sold, communicated, and competed.

A Website Was No Longer Enough

During the first years of the Web, many corporations treated the Internet almost like an electronic brochure. They built a website, displayed their address and telephone number, perhaps added some information about their products, and considered the job finished. By 1999, that attitude was becoming obsolete. The real competitive advantage came from integrating the Internet into the business itself. Customers could place orders electronically. Suppliers could receive purchasing information faster. Companies could track inventory, communicate with distributors, provide technical support, and exchange information without relying on piles of paper, telephone calls, and fax machines. The Internet was beginning to affect the entire organization.

Intel Becomes Its Own Experiment

At Intel, we decided that if we were going to talk about the Internet economy, we had better use it ourselves. In 1997, almost none of Intel's customer business was conducted over the Internet. By 1998, more than 20 percent was. For 1999, I estimated that more than 40 percent of our revenues—roughly $12 billion to $15 billion worth of business—would be transacted that way. We were not selling books to consumers. Much of this involved companies purchasing processors and other products from us. That taught me something important: the largest Internet economy might not ultimately consist of consumers buying things from websites. It could be businesses buying from other businesses.

IBM Calls It “E-Business”

Intel was hardly alone. IBM had begun aggressively promoting the idea of "e-business," arguing that the Internet would transform established corporations rather than merely create new dot-com startups. In 1998, IBM reported selling more than $3 billion of products and services over the Internet. The company also expected Internet-based purchasing to save it nearly a quarter-billion dollars during 1999. By that year, IBM said it was providing e-business services to more than 10,000 customers. Corporate America was beginning to understand that the Internet could reduce costs as well as generate sales.

Factories Were Going Online Too

People sometimes imagine electronic commerce as someone sitting at home ordering a book. That was only a small part of what was happening. The U.S. Census Bureau later found that business-to-business electronic commerce dominated Internet-related commercial activity in 1999. Manufacturing plants reported approximately $231 billion in online purchases of materials that year, while electronic orders accounted for a substantial portion of manufacturing shipments. Factories were ordering components electronically, coordinating production through networks, and connecting suppliers directly to purchasing systems. The Internet was entering places where consumers never saw it.

Old Industries Suddenly Had New Competitors

The urgency came from competition. A traditional brokerage firm now had to worry about online brokers allowing customers to trade from home. A bookstore had to think about Amazon. A travel agency had to consider customers purchasing airline tickets online. Banks were developing online account services. Manufacturers were connecting suppliers electronically. IBM warned in its 1998 annual report that companies in banking, retailing, health care, and other industries were preparing for competitors that might seem to appear almost from nowhere. Suddenly, a company that had dominated an industry for decades could face a competitor without hundreds of stores or branches.

The Internet Was Becoming Part of the Economy

By 1999, studies were already attempting to measure something people had barely known existed a few years earlier: an "Internet economy." One University of Texas study estimated that Internet-related businesses were producing about $507 billion in revenue and supporting approximately 2.3 million American jobs. It also estimated that e-commerce had grown 127 percent between the first quarter of 1998 and the first quarter of 1999. The precise definitions could be debated, but the direction could not. Corporations were moving business processes onto networks at extraordinary speed.

The Dangerous Part Was the Rush

Whenever a major technological shift occurs, companies face two opposite dangers. Move too slowly and a competitor can pass you. Move too quickly without understanding what you are doing and you can waste enormous amounts of money. During 1999 and 2000, fear of being left behind became nearly as powerful as enthusiasm for the Internet itself. Executives were told they needed an Internet strategy, an e-commerce division, online purchasing, new servers, new software, and new partnerships. Some investments fundamentally improved companies. Others were made because executives were terrified of appearing old-fashioned.

The Bubble Bursts - Told by Sir John Templeton

By March 2000, investors had spent years being rewarded for optimism. Technology shares rose, Internet companies went public at astonishing valuations, and people who had never studied a balance sheet were watching stock prices from their desks and homes. I had seen enough markets to understand the danger. When almost everyone agrees that prices will continue rising, the great risk is not missing the next advance. It is discovering how much of that optimism was already contained in the price.

March 10: The Peak Arrives

On March 10, 2000, the Nasdaq Composite closed at a record 5,048.62. During 1999 alone, the index had gained roughly 86 percent. Companies without profits had been valued in the billions, and investors had become accustomed to technology stocks climbing simply because they were associated with the Internet. Then the direction began to change. There was no single morning when everyone suddenly announced that the bubble was over. Markets rarely behave so politely. Selling appeared, prices weakened, companies disappointed investors, and confidence gradually began feeding upon itself in reverse.

When Growth Was No Longer Enough

During the boom, a company could report losses and investors might celebrate because its sales, customers, or website traffic were growing. After the mood changed, those same losses became frightening. Investors began asking questions they should have been asking all along. How much cash does the company have? When will it become profitable? Can it survive without another round of financing? Technology and Internet valuations were reassessed sharply. The Federal Reserve later reported that some of the most dramatic declines during 2000 occurred in technology, telecommunications, and Internet shares as investors reconsidered the exceptionally high prices they had been willing to pay.

The Money Begins to Disappear

This was particularly dangerous for young Internet companies because many had been designed around a simple assumption: more money would always be available. Venture capital financed expansion, advertising, employees, offices, and warehouses while companies raced toward profitability. Once investors became cautious, that financing became much harder to obtain. Companies that had burned millions of dollars every month suddenly discovered they could not simply raise another round. By the end of 2000, more than 100 Internet companies had reportedly closed, and dot-com firms had announced more than 41,000 job cuts. Pets.com, Furniture.com, Garden.com, and MotherNature.com were among the names that disappeared.

A Fortune on Paper Can Vanish

The Nasdaq finished 2000 at 2,470.52, down approximately 39 percent for the year and more than 50 percent below its March record. The Federal Reserve estimated that falling stock prices during 2000 erased more than $1.75 trillion from American household wealth. Think about what that meant for someone whose company shares had made him a millionaire six months earlier. If those shares had not been sold, much of that fortune might exist only as a memory. A market teaches the difference between price and value very efficiently once prices begin falling.

The Collapse Continues into 2001

Those hoping the damage would end with the calendar year were disappointed. During the first quarter of 2001, the Nasdaq fell to more than 60 percent below its March 2000 record. Technology-company profits weakened, telecommunications businesses struggled, investment slowed, and layoffs continued. By January 2001 alone, dot-com companies announced nearly 13,000 additional job cuts; by the end of that year, more than 100,000 people had reportedly lost jobs at Internet-related businesses and hundreds of dot-com companies had shut down or entered bankruptcy.

I Had Been Betting Against the Mania

Before the collapse, I had taken short positions in a basket of highly valued technology companies whose insiders were approaching the expiration of IPO lockup periods. My reasoning was not that technology would fail. I believed prices had simply become detached from reasonable expectations. I concentrated on companies whose shares had risen dramatically above their offering prices and expected that some insiders would sell once legally permitted to do so. I did not succeed on every position; no investor does. But the strategy reflected an old principle: when enthusiasm becomes nearly universal, one should examine very carefully what everyone else has stopped questioning.

The Internet Had Not Failed

The most important lesson came afterward. The collapse did not prove that the Internet was useless. Quite the opposite. People continued sending email, shopping online, searching for information, connecting businesses, and building digital services. The mistake had been confusing a revolutionary technology with the assumption that every company associated with it deserved an extraordinary valuation.

What Survived—and What the Dot-Com Crash Taught America - Told by Sir John Templeton and Paul Allen

By 2001, the celebrations had ended. Technology stocks had fallen dramatically, startups were disappearing, and thousands of workers who had believed they were building the future were suddenly carrying boxes out of their offices. But the Internet itself had not disappeared. Sir John Templeton and I would have looked at the wreckage from very different directions. He would have asked what the businesses were actually worth. I would have asked what technology remained after the investors went home.

The Crash Did Not End with the Year 2000

Templeton: People sometimes imagine a financial bubble bursting like a balloon—one dramatic moment followed by silence. Markets seldom work that way. The decline continued throughout 2001. The Nasdaq lost another 20 percent during the year, and falling stock prices through the first three quarters of 2001 erased an estimated $3.5 trillion of American household wealth. The economy weakened as well, and the September 11 attacks added another terrible shock to an economy that was already struggling. The important lesson was becoming clear: a stock certificate does not promise wealth simply because the company represents an exciting technology.

Allen: And yet, John, when I looked around, the technology had not gone backward. Nobody disconnected the Internet because technology stocks fell. Nobody decided email had been a mistake. Companies did not dig fiber-optic cables out of the ground or throw away their servers. People continued going online, and corporations continued connecting their operations. That distinction mattered. Investors had overestimated how quickly some businesses would become valuable, but they had not imagined the underlying technological transformation.

The Weak Companies Disappear

Templeton: Exactly. A downturn performs a rather unpleasant form of selection. During prosperous times, easy money can keep a weak business alive because someone is always willing to finance another year of losses. Once capital becomes scarce, the company must survive on customers, cash, and economics. The damage was substantial. In information-technology-producing industries alone, the Bureau of Labor Statistics recorded more than 196,000 separations associated with extended mass layoffs during 2001 across computer hardware, software and computer services, communications equipment, and communications services. When financing disappeared, businesses that had confused fundraising with success discovered the difference very quickly.

Allen: But failure also released programmers, engineers, equipment, ideas, and experience back into the economy. That happens after major technological booms. The company may vanish while the knowledge remains. Thousands of people had learned how to build websites, operate servers, write Internet software, manage online customers, and move enormous quantities of data. The dot-com era had trained an entire generation in technologies that other companies could now use.

Amazon Finally Answers the Profit Question

Templeton: Amazon was an especially interesting test because critics had spent years asking whether it could ever earn money. On January 22, 2002, Amazon announced its first quarterly GAAP profit: $5 million on fourth-quarter 2001 sales of $1.12 billion. Five million dollars was modest compared with the enormous losses the company had accumulated, but symbolically it mattered. Amazon was demonstrating that an Internet company could move from rapid expansion toward an economically sustainable business. The lesson was not that every dot-com deserved patience. It was that investors had to distinguish between companies losing money while constructing something valuable and companies losing money because customers simply did not value what they offered.

Allen: And by 2002 Amazon's annual sales had reached approximately $3.93 billion. That tells you why dismissing the Internet because of the crash would have been as foolish as believing every Internet startup during the boom. Millions of customers had actually changed their behavior. They were buying online. Once people discovered that ordering merchandise through a computer could be convenient, that habit did not disappear because the Nasdaq declined.

eBay Proves That Network Effects Were Real

Templeton: eBay demonstrated another kind of durable business. The company had approximately two million registered users at the end of 1998. By the end of 2002, it reported approximately 62 million registered users and nearly 28 million active users. That is what investors should have been searching for during the mania: evidence that a growing network actually created additional usefulness. More sellers attracted more buyers; more buyers attracted more sellers. The growth had an economic reason behind it.

Allen: That is an important distinction. Some dot-com companies believed a large number of visitors automatically constituted a business. eBay had created a system people actively used to conduct transactions. Amazon had built logistics, customer relationships, and technology. Search engines were proving that enormous amounts of online information required better ways to organize it. The survivors were beginning to show which Internet ideas represented lasting changes in behavior rather than temporary experiments.

The Infrastructure Survives the Investors

Allen: There was another survivor most people never saw: the infrastructure. During the boom, telecommunications companies had spent heavily on fiber, networking equipment, data centers, and computing capacity. Some of those companies themselves would encounter severe financial trouble because far too much capacity had been built too quickly. But the fiber did not cease functioning because its owner's stock price collapsed. The United States emerged from the boom with substantially more digital infrastructure than it possessed before it began. The financial return on some investments was disastrous; the technological usefulness of the equipment could be very different.

Templeton: That is one of history's more interesting ironies. Investors can lose money financing assets that later prove extremely useful to society. Railroads provided earlier examples. Too many lines might be constructed, companies might fail, and shareholders might suffer—but the tracks remained. The Internet boom demonstrated something similar. One should never confuse the eventual importance of an invention with the price an investor should pay while enthusiasm is at its height.

What America Learned

Templeton: If I were teaching a young investor about this period, I would tell him never to accept the phrase "this time is different" without considerable examination. New technology can change the economy while basic principles of investing remain intact. Revenue matters. Cash matters. Competition matters. Price matters. Most importantly, the price you pay for an asset influences the return you can receive from it. The Internet was revolutionary in 1999. That fact did not make every Internet stock reasonably priced.

Allen: And I would tell the inventor something complementary: do not mistake a financial crash for the death of an idea. The personal-computer industry experienced failures. The Internet experienced failures. Every meaningful technological revolution produces experiments that do not work. By 2002, the frenzy had been stripped away, but the Web, electronic commerce, digital media, networking, and broadband were still advancing. The next generation of companies would build upon infrastructure and knowledge created during the boom.

Templeton: Then perhaps that is the final lesson. Optimism was not the mistake.

Allen: Unquestioning optimism was.

Templeton: Quite so. The dot-com boom taught America that a revolutionary technology can be real while a financial bubble surrounding it is also real.

Allen: And the crash taught us something equally important: when the bubble disappears, look carefully at what is still standing. That is often where the future has been hiding all along.