Copyright © 2001 by Norman P. Poire. All rights reserved.

Fifty years ago witnessed the debut of the first commercial electronic digital computers. In February of 1951 the British-built computer, Ferranti Mark 1, was delivered to the University of Manchester. The following month the U.S. Census Bureau purchased the UNIVAC I.

These monstrous machines were painfully slow by today’s standards. The eight-ton UNIVAC occupied more than 350 square feet and had a clock rate of only 2.25 megahertz. Faster than anything in its day, it was 26,000 times slower than today’s desktop machines.

The UNIVAC borrowed its architecture from a 1939 prototype machine, the Atanasoff-Berry Computer, or ABC for short. ABC was named after its inventors, Iowa State physics professor John Atanasoff and his graduate assistant Clifford Berry.

Mainframe computers like the Ferranti and UNIVAC dominated the computing landscape for two decades. With a price tag between one and three million dollars they were out of the reach of all but government agencies and very large corporations. In 1952 International Business Machines entered the fray with their IBM 701. Big Blue, as IBM came to be called, soon became the premier mainframe manufacturer.

The invention of the microprocessor by Intel in 1970 changed the industry forever. As recently as 1963 mainframe computers costing 3 million dollars possessed a mere 512,000 bytes of random access memory (RAM). Expressed in year 2000 dollars, these machines cost over 30 million dollars for just one megabyte of RAM.

The microprocessor led to the development of a computer that individual consumers could purchase. The first fully assembled microcomputer arrived in 1977 when Apple marketed its Apple II computer for $1298. The rug had been pulled from beneath IBM as Microsoft, Intel, and Apple moved to the forefront.

Computer price declines spurred by the microprocessor have been truly mind-boggling. With 4000 bytes of RAM, Apple was charging consumers about 900,000 year 2000 dollars for a megabyte of RAM in 1977. This was a remarkable improvement over the mainframes of the 1960s. But no one witnessing the birth of the microcomputer could have predicted what lied ahead. Today Apple’s flagship computer, the Imac, can be had at an unbelievable price of only $6 dollars per megabyte of RAM!

By 1994, half of all U.S. households owned desktop microcomputers in their homes. The rest of the country still needed convincing that this new technology was worth the price tag. Not everyone was enamored with the idea of electronically cataloguing recipes, tracking household expenses, or playing video games.

The following year two innovations would change their collective minds. In 1995 Microsoft introduced Windows 95. It had taken over a decade, but now the dominant operating system in the world had become as user friendly as the superior one it was imitating, Apple’s menu driven operating system. Suddenly consumers could own a computer that was inexpensive, compatible with virtually all available software applications, AND easy to use.

More importantly, 1995 was the year that the Internet became available to the public. The Internet was born in 1969 as the ARPANET, a government sponsored network of universities and government agencies. Twenty-six years later, consumer online services (AOL, CompuServe, Prodigy) began providing access to the Internet following the 1994 release of Netscape’s free Internet browser.

From that time forward, anyone without a microcomputer would be an outsider, cut off from the rest of the world. No email, no personalized news, no finger-tipped research. The microcomputer had progressed from expensive luxury to affordable necessity.

The computer is but the latest in a series of Growth Innovations — technological breakthroughs that grow into massive industries and dominate the economic landscape for a period averaging 55 to 56 years. Since the Industrial Revolution began in 1769 with the patent of James Watt’s steam engine, there have been three prior recognizable growth industries.

England gave birth to the first cotton-spinning mill in 1771, thus introducing the world to the factory system. The blossoming textile industry saw little serious competition from other nations, including the agrarian U.S. who exported cotton to the English mills. The boom years for textiles lasted from 1800 to around 1853, when the railroad surfaced as the succeeding Growth Innovation.

The passenger railroad was invented in England in 1825 and service began by 1830. 1830 also saw service begin in the U.S. when the Baltimore & Ohio (B&O) Railroad began running passengers between Baltimore and Ellicot City.

Like all growth industries in the U.S., the expansion of railroad service followed a distinctive pattern: invention was followed decades later by two sustaining technologies. These innovations worked together to extend market penetration to virtually 100 percent of consumers.

The first sustaining technology, the Integration Technology, shows up approximately 28 years after the primary invention and launches the Growth Innovation into its high growth phase. Prior to that time, the Growth Innovation can only find willing buyers among affluent clientele like government agencies, large corporations, or the super-rich. The Integration Technology drastically reduces the cost of production and alters the product or service in a way that makes it more appealing to individual middle-class consumers.

When in 1855 Henry Bessemer perfected his new steel-making process, the price of steel was drastically reduced and the railroad Integration Technology was in place. The natural competitive advantage of steam power over horse power was in long distance travel. With the Bessemer Process a reality, a huge network of rail quickly followed making railroad service a practical reality.

Practical as it was, most people considered railway travel a necessary evil. Passenger cars were dark, dirty, crowded, and uncomfortable. After 28 years in its integration phase, when a growth industry reaches approximately 50 percent market penetration, Domination Technologies appear to make the innovation attractive to the remaining sector of the population.

In the 1880s electric lighting first appeared in passenger cars, followed by steam heat and the adoption of the safer air brake. Before the decade ended, all railroads were running on standard gauge tracks of 4 feet 8 1/2 inches, which allowed passengers to travel anywhere in the country without changing railroads. Railway travel had become an enjoyable experience.

By 1913, the railroad and its supporting industries dominated the economy. After that year, railroad growth slowed dramatically. Miles of railroad in operation in the U.S. peaked and remained constant for two decades. As this growth industry approached its 112^th birthday in 1937, miles of railroad in operation began to decline. Railroads had reached the saturation point.

This process of invention-adaptation-integration-domination-saturation is represented schematically by the S-curve in Figure 1. The process was repeated for each of the United State’s Growth Innovations as summarized in Table 1.

Figure 1. Growth Innovation S-curve

Table 1. U.S. Growth Innovations

Throughout modern history there have been Critical Innovations so profound that they launched entire technological revolutions. The steam engine, responsible for igniting the Industrial Revolution, is one example. Others include the telescope (Scientific Revolution), the printing press (Information Revolution), and the electronic digital computer (Second Information Revolution).

The printing press brought ideas together in a way they never had before and made people skeptical of long-held beliefs. Old ideas were revised, synthesized with newer ideas, or discarded completely. The Information Revolution it spawned ushered in great social change.

We call this dynamic period The Renaissance, a time when the old God-centered world gave way to a more Human-centered existence. This period of relative open-mindedness and learning paved the way for great scientific exploration that in time blossomed into an unprecedented technology boom with its commensurate economic prosperity.

The 500-year period that stretches from the invention of the printing press in 1440 to the invention of the computer in 1939 is the Age of Physics. It was a time when humans gained the confidence that the world was within their control — a time when the human mind knew no bounds.

Science in the Age of Physics was devoted to harnessing nature in order to serve humanity. The World According To Newton operated like a giant clock — a complex, yet mathematically simple and predictable machine. Rarely were scientists and engineers concerned that their improvements might somehow come back to haunt them. Nature, it seemed, was far too robust and stable to be threatened in any fundamental way by the trifling of human beings.

In a similar manner, the Second Information Revolution is mingling ideas in a new way, dramatically changing the way we view the world. Our Human-centered existence is slowly giving way to a more Nature-centered world. It is a time when we are questioning the kinds of technologies that the Age of Physics produced.

Science in the Age of Biology will attempt to bring about change that recognizes human limits and the vulnerabilities of nature. The World According To Darwin is in a constant state of flux. It becomes imperative that we understand how systems interact with each other so that when we use technology to solve a problem we do not summon the Law of Unintended Consequences.

For every action in Newton’s paradigm there is a predictable reaction. For every action in Darwin’s paradigm, we cannot be sure what reaction to expect or whether it will work for or against us.

The Growth Innovation that succeeds the computer will conform to the Age of Biology. That means it will likely be a "green technology" — one that is compatible with the environment and perhaps leaves the world a little cleaner than when it arrived.

To conform to the pattern of previous cycles, it must produce an important invention approximately 55 or so years after the appearance of the computer — 1994 give or take five years. It also must hold the promise of being a commercially viable product within 12 years, or by 2006. Finally, it must have the potential of penetrating close to 100 percent of the entire consumer market.

Two promising fields that are fertile ground for future growth industries are genetic engineering and artificial intelligence. Clearly these are both biologically centered fields -- genetic engineering is machine-like manipulation of living cells while artificial intelligence seeks to endow machines with human-like traits.

Exciting new inventions in both fields have appeared in time to conform to the Growth Innovation archetype. In 1994, the MIT Artificial Intelligence Lab produced their first humanoid robot, Cog. As they describe the project, "Cog is a single hardware platform which seeks to bring together each of the many subfields of Artificial Intelligence into one unified, coherent, functional whole."

It is not too far-fetched to believe that some day human beings will be waited on hand-and-foot by a race of slave humanoid robots. However, these machines have not advanced to the point where a commercially viable model will be available over the next decade.

Dolly the Sheep, the first ever mammal cloned from an adult female, was born in 1996. Along with her was born a major controversy over the ethics of cloning that is not likely to be resolved any time soon. Genetic engineering is a Nature-centered technology built on the Man-centered philosophy that mankind will benefit by rearranging Nature’s handiwork.

Neither of these innovations is as strongly positioned, then, as the fuel cell vehicle. Fuel cells convert chemical energy into electricity with efficiencies as high as 70 percent (internal combustion engines top out at 30 percent). They can be used to supply power to a variety of products — from cell phones to electric automobiles.

When supplied with hydrogen as a fuel, the fuel cell’s only waste products are heat and water vapor, making them a perfect fit for the Nature-centered Age of Biology. However they can be made to function with a variety of fuels, including gasoline.

A British scientist, Sir William R. Grove originally invented the fuel cell in 1839. It would be over a century before NASA would develop the technology for use in its Project Gemini program in 1965. It would take three more decades for the fuel cell to find its way into a consumer product.

In May of 1996, Daimler-Benz unveiled the first fully functional light duty passenger vehicle powered by a fuel cell. Daimler and their Canadian fuel cell partner, Ballard Power Systems, demonstrated that a pollution-free passenger car could be built without compromise to performance, comfort, range or safety. Right on cue, the NECAR 2 (short for New Electric CAR) appeared fifty-seven years after the invention of the computer. Fuel cell vehicles are expected to become a commercial reality by 2004.

The fuel cell, along with microturbines and solar power, will play a key role in the emerging field of distributed power generation. Unlike conventional power generation facilities that require years to build, smaller power units placed nearer to the demand source can come on line much more quickly.

Although there have been no large power plants built in nearly two decades in California, their energy crisis is not simply a supply problem. In the October 2000 publication, TechStrat Insights, the Merrill Lynch Technology Group headed by Steven Milunovich points out that our "microprocessor-based economy is driving a rapidly escalating need for high-quality power."

They add: "The pervasiveness of computing places high demands on the electric grid. The answer is distributed generation through new power technologies." In their opinion, Canada’s Ballard Power Systems (BLDP) and Connecticut’s FuelCell Energy (FCEL) are the two best positioned companies for accomplishing this task.

In his 1995 book, How To Buy Technology Stocks, Forbes technology columnist Michael Gianturco provides some helpful insights on narrowing down the field of technology candidates.

His guidelines include choosing companies with P/E ratios between 14 and 28 that carry little or no debt. He advises us to ignore dividends and points out that company book value is less relevant for many technology firms due to a high level of intellectual property (particularly at software firms).

Fuel cell companies at this time have no earnings so his P/E guideline is not yet applicable but his other suggestions are useful. His most important insight, however, is that the only company worth investing in is one that has secured some kind of monopoly power. Monopoly power can take many forms.

The most obvious monopoly power is granted by the government in the form of patents. A patent allows an inventor the right to market his or her product without competition in order to recoup R&D expenses.

Other forms of monopoly power might be: sole access to a distribution network, a large market share that offers economies of scale, or natural low cost inputs (for example a firm that is built on a river bank with access to "free" water power or one that is built near a university with access to inexpensive research).

Companies can also build a wall around their business by doing something so intellectually challenging it is difficult for others to imitate. Or through marketing they can create a wall of goodwill. Whatever it is, the company you select should have some kind of lasting competitive advantage.

Another important lesson from Gianturco: "In any competition of par products, capital wins. Bigger is better."

With these guidelines in mind, five fuel cell manufacturers stand out. The largest (most heavily capitalized) company by far is Ballard Power, worth around 5 billion dollars. They also have partnering agreements with Daimler-Chrysler and Ford. The next two largest at around 1 billion dollars are FuelCell Energy and Plug Power.

Plug Power enjoys access to one of the best distribution channels in the world thanks to their marketing agreement with General Electric. FuelCell Energy is unique in that it is the only large player that is pursuing the molten carbonate technology. Molten carbonate is more efficient (and more expensive) than the proton exchange membrane (PEM) design that most other companies are using. It is also better suited to large stationary applications, which is playing well into the current energy crisis.

HPower Corporation is the fourth largest fuel cell company and has a ten year marketing agreement with ECO, an association of approximately 250 U.S. rural electric cooperatives. Manhattan Scientifics is one of the few companies targeting the below 1 watt portable market with their MicroFuel Cell^TM. They also have a patented process that produces fuel cells from economical mile-long thin printed sheets (much like printed circuits).

We believe at this time that these five fuel cell companies are best situated to lead us into the next technology boom.

If history is a reliable guide, the time to buy and hold the market leaders of the fuel cell industry is just around the corner.

The following Figure 2 displays the Standard & Poors computer index since 1939 when the computer was first invented. The index is normalized using two methods. First, it is divided by the S&P 500 Composite Index to measure computer stock price performance against the market in general. Then it is also divided by the Standard & Poors automobile index to get a performance relative to the preceding Growth Innovation.

Once computer companies began selling their units in 1951, computer company shares outperformed the general stock market by a factor of 14 over the next two decades. During that period, S&P 500 stocks were paying out cash dividends of around 4 percent. A $1000 investment in the S&P 500 stocks would have been worth $10,505 in 1972 if those dividends had been reinvested in shares.

A $1000 investment in computer shares in 1951, even making a conservative assumption that they were paying NO cash dividends at all, would have been worth a whopping $68,433 by 1972! Timing, as always, is everything as computer shares underperformed the rest of the market once the microprocessor emerged (no doubt because this index includes neither Intel nor Microsoft, never mind Compaq, Dell, et. al.).

The message is the same for both curves. The time to buy is when fuel cell companies begin commercial production. That could be as early as this year for manufacturers of stationary units and 2004 for automotive fuel cells.

Figure 2. Standard & Poors Computer Index Relative to the
S&P 500 Composite Index and to the Standard & Poors Automobile Index

In 2001 we find ourselves perhaps in the first secular bear market in two decades (the last one lasted from 1968 to 1982). There have been two prior capital spending boom/bust periods in the U.S. that are similar to the one we are experiencing today.

The first peaked in 1835 while the railroad was in its infancy. The second peaked in 1929 when the auto industry was in its rapid growth integration period. Figure 3 demonstrates that these Growth Innovation share prices tended to follow the general market down during these severe bear market periods.

The message here is simple. If you believe as I do that U.S. stocks have entered a secular bear market, the best time to buy shares in the fuel cell industry is at the beginning of the next bull market in stocks and after commercial production has begun. Those two events, should they occur about the same time frame, would present us with one of the greatest investment opportunities of a lifetime.

Figure 3. Growth Innovations during Secular Bear Markets

Posted May 2001

Revised June 2001

Revised March 2004

back to the top