Note: While reading a book whenever I come across something interesting, I highlight it on my Kindle. Later I turn those highlights into a blogpost. It is not a complete summary of the book. These are my notes which I intend to go back to later. Let’s start!

The guy who told me he looked at only the second derivative of industrial production was onto something. Because information is distributed in milliseconds, there is no time advantage anymore. You have to be ahead of news. You have to look not at change but at how fast change is changing. Your old calculus teacher would remind you, the first derivative is speed. The second derivative is acceleration. There are so many barriers to change. Some are technical, some are based on silly government regulations. But who cares—barriers are barriers. As long as you can find a barrier to invest against, you can make money when the barrier breaks, when change accelerates.

There are so many barriers to change. Some are technical, some are based on silly government regulations. But who cares—barriers are barriers. As long as you can find a barrier to invest against, you can make money when the barrier breaks, when change accelerates. Most hedge fund guys take the other side of a trade when they know something no one else does, i.e., investing because others don’t know. That’s their edge. When you think long-term, the edge is really investing because others can’t know. I suppose others could know if they thought hard enough, but, oddly, no one does. It’s not the actual declining cost of power or transportation, chips or bandwidth, etc., that is hard to figure out. It’s the change they enable. These concepts are hard to grasp. It’s that nasty second derivative stuff. It’s not the amount of change, but the change in the rate of change. Confused? Me too. It’s why there are few physicists and why people quit science after high school. It’s just too hard to think about. From the side of the highway, you can’t tell which cars are going a constant speed and which are accelerating. But from inside the car, you can feel the seat press against you when you gun it. That’s probably the first lesson I learned: to do well, you’ve got to be in the car, not on the sidelines watching. But in reality, second derivative stuff is quite simple. If you can figure out what is getting cheaper, year by year, you can start to imagine the change to the status quo into the future. And, I think, you only have to be close. That’s the beauty of Silicon Valley. It is about only one thing—change. Maybe I could find pockets of change. Then I wouldn’t just be looking for a 30% move in the value of construction stocks after a hurricane; instead, I’d be trying to anticipate huge moves of 500% or 1000% that take place after a barrier collapses.

So, what is it? Are there still barriers? How can I make sense of the sea of economic data, earnings announcements and noise on the Street? What is it that others can’t know because second derivatives are hard to conceptualize? Could I find a grand unifying theory of how the world works? If so, while others are sweating out and trading around short-term changes, I could sit back and think out the big moves. Or die trying.

“Lots of money in Hong Kong, but no investors.” “Well put. Anyway, I met with a guy that I knew back in my days at Paine Webber, a Bill Kaye.” “Haven’t heard of him.” “He runs some Asia investment fund and seems to have it easy. I guess I’ve been depressed about how hard it is since I talked with him.” “Why is that?” Mr. Zed asked. “Well, he had it all figured out. He said that I had the hard job, figuring out the future, and that in Asia, it was just the Industrial Revolution movie playing over again, and you could just find countries and companies in different stages of development and place them in the Industrial Revolution timeline and figure out whether to invest. So, why do I always pick the hard stuff to do, instead of just sitting in some marble building and eating Peking duck all day, picking stocks from a script from some already played-out movie.” “Well, you’ve got the easy job.” “How’s that?” I asked. “Because you can find the next Industrial Revolution.” “I don’t understand.” “You will. Go find that movie and play it. Find the parts you need, and then make your own. Then you will tell me how you will invest.” “I thought I stopped doing homework when I graduated college.” “Surely you have figured out that you never graduate.” “Never?” “Get back to me, you have some work to do. In the meantime, I’m going to wire $5 million into your fund tomorrow morning. But your hard work is just starting, not ending.”

In 1763, a technician named James Watt was employed at Glasgow University. His task was to maintain—more like fix—a Newcomen steam engine that the university owned. It was, as techies like to say, a POS, a piece of shit. It was a terrible kludge, literally held together by wet rope. It broke all the time. So like all good engineers, Watt took it apart to figure out how it worked. It was nothing more than a giant cylinder with a plug, or piston, inside of it. A furnace boiled water and pumped steam into a cylinder. These were low-pressure steam engines, also called atmospheric engines. High-pressure engines kept blowing up, killing off everyone involved. Low-pressure workers were survivors. Unlike your car engine, the steam in an atmospheric engine didn’t push a piston in a cylinder. Instead, steam filled the cylinder, which was then doused with cold water to rapidly cool it. When the steam turned to water, it created a vacuum, hopefully a strong enough vacuum to suck the piston down. This action then raised a rod that lifted a plunger of sorts, which tried to suck water up out of the mine. Ingenious for 1706, it huffed and puffed and barely had the power comparable to a horse or two. Since the engine broke down all the time, fixing it was a full employment act for technicians like James Watt. Plus, someone constantly had to seal the cylinder to make a strong vacuum and prevent steam from leaking out of the craggy-edged cylinder. Wet hemp, Jamaica’s finest, was the sealant of the day. The professor in charge of Watt at Glasgow University, Dr. Joseph Black, was teaching courses, theorizing about a concept known as latent heat, starting back in 1761. Another professor at U of G around the same time was Adam Smith (of the invisible hand). In fact, Smith and Black were good friends. Latent heat is the reason a watched pot never boils or why you put ice cubes in soda. Latent heat means you can add heat to a pot of water, but it won’t boil and give off steam until the entire pot of water is at 212° Fahrenheit. And no matter how much heat is applied by the hot sun at a baseball game, all the ice has to melt before a soda increases in temperature, right before the kid behind you spills it on your shoes.

As a favor to Professor Black, Watt ran a series of experiments measuring temperature and pressure and proved a prevailing theory that steam contained “latent heat.” In doing so, Watt figured he knew why Newcomen’s steam engine was all wet. Watt theorized that the cylinder had to stay as hot as possible, boiling hot, so new steam added to it would stay steam and not condense too soon. To create the vacuum, Newcomen had been splashing cold water in and on the cylinder. On each stroke of the engine, lots of steam was needed just to reheat the cylinder, so the engine was running at 25% efficiency. And then a lightbulb went off in Watt’s brain: Watt added a simple improvement to the Newcomen design, creating a separate chamber outside of the cylinder. This “condenser” was kept underwater, as cool as possible, and the steam condensed into it while the cylinder stayed hot for the next cycle. The power of the engine doubled to 3 or 4 horsepower. Between 1763 and 1767, Watt went into debt up to his eyeballs to perfect his new design. His engine leaked like Newcomen’s because the cylinders were not “true” and the condenser was not efficient. Improvements were small. In 1767, John Roebuck, the owner of a Scottish iron foundry and a part-time venture capitalist, assumed Watt’s debts of 1,000 pounds and gave him fresh money to improve his design. In exchange, Roebuck received two-thirds ownership of any patents. Patents were an important part of English law to protect property owners, even if the property was just ideas. In 1769, Watt was granted a patent for his steam engine design by Parliament, which had recently taken over the patent-issuing duty from the king. Parliament was run by property owners, who, not surprisingly, were all for upholding property rights. Almost simultaneously in 1769, old John Roebuck went bust, collapsing under the burden of his own debt. His foundry lacked a good source of power, which is why he was so interested in Watt’s inventions. Roebuck went bust because he couldn’t turn a profit with horses running his shops.

Another manufacturer stepped in. Matthew Boulton was born into the stamping business (buttons and buckles—Puritans loved them). His dad acquired the Sarehole Mill in an area known as Hall Green, where they rolled their own sheet metal to be stamped. Later, J. R. R. Tolkien grew up next door, and the area provided the background for his Middle Earth. Perhaps Boulton’s mill sparked Tolkien’s distaste for the industrial age. Boulton struck out on his own in 1762 as a manufacturer of luxury goods, a “piecer.” He created the Soho Manufactory out of a water mill a couple of miles outside Birmingham. He was constantly on the lookout for ideas and processes that could improve his 1,000-employee, three-story shop. Back in 1768, James Watt had stopped by to check out Boulton’s manufactory. They discussed Watt’s new engine, now up to 5 or so horsepower, as well as its potential uses in the factory and even in driving carriages. It was a fateful meeting, because a year later, Watt needed to raise some money fast to buy out the now-seared Roebuck. When Watt came back begging, Boulton agreed to buy out Roebuck’s two-thirds interest in the patent. More importantly, Boulton agreed to fund the continued research by Watt into making his external condenser steam engine work. The Newcomen design was still selling, despite all its flaws, but the market wanted more powerful engines. Watt’s biggest problem was getting materials and labor to construct his engine accurately. The cylinder was key, and its “trueness,” how accurately round the cylinder was, directly affected its power. Watt was excited when he constructed a cylinder that was within ? of an inch, the thickness of your finger, of being a “true cylinder.” In 1775, he went to London, where he had a few, shall we say, influential friends introduce legislation in Parliament to extend Watt’s patent, which was set to expire in 1783. The bill passed, and the newly formed Boulton & Watt Company owned the patent on atmospheric steam engines for 25 years, until 1800.   So now it’s 1774, and the king desperately needed those Wilkinson cannons. Wilkinson got the order because he had a secret weapon for making cannons. The Iron Master had a nifty precision-boring tool—a monster lathe. This tool cut cannon barrels true, making highly effective cannons with ever-so-narrow windage. No smelt-it jokes, please—windage is the gap between a cannon’s barrel and the cannon ball. The smaller the windage, the greater the distance the gunpowder’s blast propels the cannonball, rather than leaking out past it. Wilkinson never patented the tool, he just used it in his own shop, which like everyone else’s, had recently moved from river’s edge to the coalfields. High-grade coal or coke was plentiful up in the hills of Staffordshire and contained almost no sulfur, so the resulting iron was sturdier, but he lost his source of power, the river. Wilkinson solved one problem but ended up with another. Coke burned hotter than charcoal. He needed to crank 15-foot-high bellows to blow enough air to heat up the coke to an intense enough heat. His boring tool also needed a source of power to turn. It required teams of horses, which were expensive to feed, let alone clean up after. James Watt’s steam engines were in the area, pumping water out of coal mines, and Wilkinson thought he could use one to crank his bellows instead of horses. So, Wilkinson tried one. Success? Nope. Instant failure. There was barely any power from Watt’s engine to pump the bellows. So Wilkinson took the steam engine apart and probably started laughing. Watt’s cylinder was awful—as jagged as England’s shoreline. Even wrapped with wet hemp, it leaked steam with every stroke, robbing the engine of most of its power. While Watt was proud of his ? of an inch from true cylinders, Wilkinson had his lathe and knew he could make Watt’s cylinders truer. Wilkinson recast Watt’s leaky cylinder using his top-secret precision boring tool and found it generated four to five times more power, enough to run his bellows. This meant 25-to 40-horsepower engines, up from 5 to 8. The difference for miners and millers was staggering. Being a reasonable businessman, he told Boulton and Watt that he could improve their crappy little steam engine by a factor of five, in exchange for the exclusive rights to supply precision cylinders to B&W.

Deal. As an investor, I was getting more and more intrigued by this little tale. It had market demand (flooded mines), technology (Watt’s condenser and Wilkinson’s precise cylinders), capital (Boulton’s money), intellectual property rights (Parliament’s patent) and a ready workforce (ex-farmers). I was ready to invest—all that was missing was a business model. It was Matthew Boulton who came up with one. Boulton and Watt didn’t actually sell steam engines. No one could afford one. Most of the early customers were Cornish mines. Beyond Parliament-sponsored joint-stock companies, the stock market and banking were not quite developed, especially for risky businesses. Limited liability for corporations wouldn’t be the law until 1860. Miners lived day to day. They used a cost book system of accounting (I slept through accounting too). At the end of each quarter, all the partners in the mine would meet at the counting-house to go over the numbers and split any profits. These count dinners were giant drunk fests, each mine vying for the prize of offering the most potent punch. At the end of the night, the mine companies were drained of cash and the miners drained of brain cells. So instead of selling steam engines, Boulton just traveled around to mines (and later mills and factories) and simply asked how many horses they owned. Boulton and Watt would then install a steam engine and charge one-third of the annual cost of each horse it replaced over the life of the patent, that is, until 1800. Back then, a horse cost about 15 pounds per year, and I have seen figures for the parts cost of their steam engine of 200–300 pounds to build a 4-hp engine. A 50-hp engine cost around 1,200 pounds. Not having four eating and shitting horses around meant saving 60 pounds a year. Boulton and Watt would charge 20 pounds a year. If the 4-horsepower engine cost, say 200 pounds, B&W started turning a profit after ten years and they had a 25-year patent. But they could charge 250 pounds a year for a 50-horsepower engine and turn a profit in less than six years. It was in their best interest to install more powerful engines—they just needed to find something beyond pumping out flooded mines to drive demand for horsepower. A barrier lowering the cost of power had just been busted down, but I didn’t see the gusher. Something sounded familiar. It was from my conversation with Mr. Zed. Cheap microprocessors and cheap memory made computers cheaper, and they sell by the boatload. I started to wonder whether this same scale is what drove jolly old England as well. Would I have invested in Boulton & Watt? Not yet. It wasn’t clearly a home run. What was I missing?

“Good morning, Andy. Is this too early?” “OK, power got cheaper,” I said, cutting to the chase the next time Zed called to check up on me. “So?” Mr. Zed asked. “So Watt’s steam engine meant horsepower got cheaper for England than the rest of the world.” “And who used it?” “Miners.” “So the British became the world’s miners?” “Well, no.” “What did they become?” “Ironworkers?” “Are you asking me or telling me?” “Both?” This was not going well. “Does Silicon Valley provide the greatest silicon to the world?” “No.” “So, find out what sucked up that horsepower. Why did they need so much of it? Find the scale.” “But where?” I asked. “Everywhere. Underwear.” And then he hung up. I think he’d seen this movie before too.

Mr. Zed got me thinking—would I have invested in the steam engine business? Maybe. The 25-year patent was nice, the business model fairly unique. But pumping water from mines? Where is the monster market in that? By the time the Boulton & Watt patent expired in 1800, they had 500 steam engines up and pumping. England was on its way to being an empire. What I found the most interesting was the drop in the cost of power. Something besides iron and cannons were using this cheaper power. Everywhere, underwear?

With steam-powered mills and looms and a steady flow of clean cotton, England now had the economic engine it needed. Import raw materials like cotton, run them through industrial machinery run by steam-powered engines and export finished goods like yarn and cloth and textiles. That was and still is the definition of an industrial economy. The key is that finished goods can be sold at prices so much cheaper than handmade goods. Back then, it changed the way the world dressed, with less itchy clothes at that. Even Don Valentine would have called that a monster market. I suppose this is no different from the way the world was changed by electricity at the turn of the last century or by radio in the 1920s or television in the 1950s, or maybe automobiles or washing machines or refrigeration or air conditioning. Industry supplied the product more and more cheaply, and an entire consumer economy was built around these cheap, revolutionary products. Heck, entire economies evolved to supply this stuff—Japan with consumer electronics and then cars, Taiwan or China with all sorts of manufacturing and assembly.

Industrialization was not some master plan to remove workers from their century-old tasks. Instead it was a complete reengineering of life based on the ability to provide daily staples at much lower costs. Getting everyone together in one steam engine–driven manufactory produced higher-quality and lower-cost textiles than anything that could be done at home by old spinsters. As long as England could keep prices for their cloth going down, they would both create new markets and keep competition away. Growth created by and protected by its own declining price elasticity. Hmm. This is something I want to invest in.

My pitch was simple. “Imagine it’s 1982 and we are about to head into an era of massive wealth creation based on the proliferation of technology whose cost is dropping fast and ever higher-performing chips. If you knew then what you know now, how would you invest differently? “Well, we are at the top of another rickety roller coaster, this one based on cheap bandwidth and ever-faster wires and fiber optics. It’s going to be a wild ride, you can’t see where it ends up, lots of people are going to toss their cookies, but lots of wealth is going to be created. So, how will you invest differently?” It was as if I had the word risk tattooed to my forehead. I got a lot of nice comments after the talk but no real bites.

Maybe the alphabet analogy is a good one, but after thinking about it, I think Doug Engelbart is a modern King James. I know that sounds odd, but so does scaling human knowledge. Think about it—the King James Bible took religion out of the hands of the high priests and put it into the hands of the people. An entire generation became literate just to be able to read the King James version of the Bible. These same people went on to work in manufactories of the Industrial Revolution, where literacy came in handy to speed up training of workers. Doug Engelbart’s augmentation effort took computers out of the hands of the high priests of the corporate information technology centers. Some call it democratization, but I like to look at it as a literacy move. Since 1969, an entire generation has become computer literate, which eases training for jobs in a knowledge and intellectual property economy. King Doug indeed. Computers were not just for boring accounting functions (the Cornish mines could have used one). They really could augment the human race and increase efficiency and productivity by replacing costly repetitive human functions. That was real scale, like steam engines scaling labor. Bing-bing-bing. On December 5, 1968, the world changed. Of course, you couldn’t invest quite yet. A few more things needed to be invented, like microprocessors.

Digital watches were hot in the early 1970s. Intel bought a company named Microma in 1973. Their watches with cool red seven-segment LED displays were selling for over $100. But as volume increased, Intel was able to make the integrated circuits inside the watches much more cheaply. Even though they were driving costs down the learning curve, no one quite understood the impact of lower costs on the end markets. As cheaper models were introduced, Microma was stuck with an inventory of expensive watches that wouldn’t sell. As others sold the necessary chips to new watchmakers, the market was flooded with cheap, under $20 watches. Intel quietly sold Microma in 1978, swearing never to forget this valuable lesson. Texas Instruments had the same problem, taking a bath in an early version of personal computers and again and again in calculators. They should have just sold the chips that went into calculators to everyone, instead of the calculators themselves. I once asked Gordon Moore about the whole Microma experience. He quickly pulled up his sleeve and pointed to a Microma watch on his wrist and told me he wore it often to remind himself to never be that stupid again. Intel’s lesson: make the intellectual property, not the end product.

“You what?” I asked incredulously. “We don’t have a fab,” Bernie Vonderschmitt told me again. No other CEO of a chip company had ever uttered those words to me before. “I guess I don’t understand that.” “What’s not to understand? We don’t have a wafer fab or any factory. We don’t need one. We don’t want one.” “Can you do that?” I was still confused. “We’re doing it.” “But Jerry Sanders says ‘real men have fabs.’” Sanders was the flamboyant silver-haired CEO of chip maker Advanced Micro Devices. He had hired Ray Charles to play at the company Christmas party in 1984, right before a nasty two-year recession hit the industry. Credibility was not Sanders’s middle name. He spent billions on wafer fabs and was questioning anyone else’s manhood who couldn’t similarly spend billions. “Yeah, we may not be real men, but we’ll be rich men. Jerry will figure this out someday. He’s a little slow on the uptake. Plus, AMD owns a piece of us.” “But every other successful chip company in the Valley owns a fab.” “Good for them, but now they are drowning with them. They have to keep feeding the beast.” Bernie was right about that. A fab cost $100 million minimum and lasted maybe three years if you were lucky, before you had to build another more state-of-the-art one. It’s what kept investment bankers in business, funding all that spending. “A couple of years back,” Bernie continued, “we took on venture capital and had a tough choice to make. We knew we had a unique and patented chip design, our field-programmable gate arrays. But we had a lot of work to do before we could get them to market. This meant a lot of designers and a lot of programmers.” “OK, everyone has this problem.”

“OK, everyone has this problem.” “More so with us. We sell an architecture and have to keep it current.” “So, either you keep your design current or your fab current?” “Right. I could have gone back to other venture capitalists and raised the $100 million I needed to build a fab, but they would have owned the company, not me and my team. So we looked around to see if anyone had a fab they weren’t using completely and maybe wanted to rent us some space.” “Did you find one around here?” “No, we found TSMC instead.” “Who?” “The Taiwan Semiconductor Manufacturing Company. All they do is run a fab and sell finished wafers to us at a fixed price. We can put anything we want on those wafers, same price. So we crank out more designs, and these guys will manufacture them all.” “This is a big change.” “It’s really no different than the publishing business. You don’t have to own a printing press to write a book or publish it. Someone else prints the pages and binds them into a book at a fixed price, no matter what the book is about.” “Can you make the same margins as Cypress or LSI Logic? They own their own fabs, so they must have some advantage.” “Actually, they make 45% gross margins in a good year, and we’ll make 55% in a bad year.” Gross margins were the percentage of profits after expensing just the materials used, not the research and development or sales costs. This was new—same business, no factory, but higher margins? Xilinx could afford to spend more on their designs than manufacturing. If this works, this thing is going to be a home run. “No fabs is a pretty bold step. Will investors let you do this?” “That’s why we picked Morgan Stanley to take us public, Andy. You’ve got your work cut out for you.” “Oh, great. We are supposed to just show up and collect our fees on IPOs.” “Not this time. This won’t be easy. Now get to work.”   It turned out to be easier than I thought. Investors on the road show needed a little coaching on what it was like to be a “fabless” chip company, that it wasn’t an oxymoron. What put everyone over the top were the company’s profit margins, that “55% in a bad year” thing. Investors “get” profits—it is their common language. My problem was in believing that someone would just make chips for Xilinx or any of the other fab-less chip companies that quickly followed. My annual December trip to the Far East was coming up, so I scheduled a side trip to Taipei. It was a bit more than a day trip to Kansas City.

 TSMC was located in Hsinchu City, a bumpy hour-plus cab ride from Taipei. I had never been to Taiwan before, but I had a whopping 18 hours to figure out how this fab-less stuff worked—no time to see the sights. Hsinchu City looked a lot like Parsippany, New Jersey: a bunch of industrial parks, a couple of malls and one medium-size hotel. What was hard to figure out was that the place was teeming with activity, entrepreneurs run wild. I checked into the hotel and dumped my bags in the room. Before heading out, I rummaged through the nightstand until I found the phone book. My friend, Hank Zona, used to nod his head in someone’s direction and whisper, “That guy has more chins than a Chinese phone book.” I never knew what that meant, but now I do—14 pages’ worth. The address was simply Science-based Industrial Park. The taxi driver never heard of TSMC. This was 1991. By 2001, he probably worked for TSMC. I was met in the lobby by Dr. Morris Chang. “OK, let’s go then. You’ll need a bunny suit.” I had been on enough tours to know the routine. I put the white booties over my shoes and slid on pants, a jacket and a surgical cap to tuck all my hair into. I was spared a mask over my face. We stepped through the blowers, and blasts of air came from the floor and side and sucked every tiny particle off us into filters in the ceiling. Marilyn Monroe would have enjoyed this room.

“This is Fab 1. The building used to be a textile factory. About a thousand sewing machines tucked side to side in here. I think you Americans called them sweatshops.” “What happened to them?” I asked. “They moved to Malaysia a few years ago. Labor is a lot cheaper there. This is a Class 10 fab, no more than 10 particulates per square foot. Each little particle kills chips, since they are larger than the wires being written onto the chip. In this fab, we can do 10,000 six-inch wafer starts a week, for now. We’ll double that soon and have already broken ground on Fab 2 and are spec’ing out Fab 3,” Dr. Chang boasted. “Wow, that makes you a medium-size player in Silicon Valley,” I told him. “We’ll do more wafer starts than the entire Silicon Valley before long.” His boasting quickly turned into bragging. I wonder if he knew Jerry Sanders. “How do you pull this off?” “It’s real simple. We sell wafers at a fixed price, around 800 bucks. You put what you want on the wafers. We don’t care. Sell them for what you want. We make money at 800 bucks.” “How much?” “If we run this fab close to capacity, we can make $250, maybe $280.” “So, around 30% gross margins.”

“Thirty-five percent on a good day. We don’t change our process. It is more or less the same for everybody. Our yields are very high.” “But even the worst companies in the Valley make 40% or 45% gross margins.” “Yeah, but they waste all that money on their fabs. We worry about that stuff for them. I have process engineers, not design engineers. If the other companies focus on design, they can make 50–60% margins. Why wouldn’t they use us?” Why indeed. “And you can survive on 35% margins?” “We do enough volume, we can survive on a lot less. We don’t pay $100,000 a year to PhDs. We also don’t have to pay that much to technicians. But it’s all relative. Our job applicants stretch around the block because we pay more than toy companies.” We left the fab, stripped off the bunny suits and headed to a conference room. “Any questions?” Dr. Chang asked. “Well, we are starting to see more fab-less chip companies in the Valley, and I want to figure out if they are for real or just some passing fad.” “Fad? Hardly. It’s the future. You guys shouldn’t make stuff. Let us. Go tell your companies we are here.” “But aren’t the Japanese renting out their spare fab capacity?” Chang started laughing hysterically. “That’s funny.” “What is funny?” “Mr. Kessler, have you ever negotiated with the Japanese?” “No, can’t say I have.”

“Well, here is how it goes. You fly in, and they meet you for dinner. Sushi, shabu-shabu, whatever that silly stuff is they eat. Then you go to a geisha bar and drink lots of American whiskey. The next day, they are bright-eyed and your head hurts. Then they take you on hours and hours of tours of their facilities. Then you get put in a conference room with some smiling low-level drones, and they just nod as you talk. You ask them for some specifics on price, and they look at one another, shake their heads and do that sucking sound. I can’t stand that sucking sound, lots of air rushing past their teeth. It is like, what do you Americans say, fingernails on a chalkboard.” I was starting to enjoy this. “Then it’s back to your hotel and you meet them for another dinner and more Asahi Super Dry and more geisha bars and two fingers of Jack Daniel’s. The next day, you come in and ask for someone in charge instead of those smilers. You’ve got a flight at seven and need to check out and get your bags. You figured some deal would be wrapped up by now. They say, ‘So solly, plesident of division vely busy, no can meet till rater.’ Some more air sucking and you are ready to hit someone. Finally, some gray-haired guy comes in with contracts, saying, ‘You sign here.’ You can’t figure out what the hell the contracts say, the numbers don’t add up, but you were told by your boss not to come home without a contract. Another night of ugly geishas and Wild Turkey would kill you, so you sign, what the hell, it’s not your money.” “Someone must be patient enough to get decent contracts,” I said. I didn’t want to interrupt, this guy was on a roll. “Maybe, but remember, those Japanese companies like Toshiba and Hitachi and Matsushita make chips too. They are your competitors. Why use them? The day demand picks up, you’ll sit around waiting for your wafers to show up while they are shipping their own chips like mad.” “What about the Koreans?” I asked.

“They are just little Japan. There are four big companies there, and they do what their banks tell them to do. Plus, we outnetwork them. Taiwan sends lots of engineering students to U.S. schools, and they stay there and work in Silicon Valley. There are few Koreans at U.S. engineering schools. Korean-Americans, yes, that’s different. No natives. So when someone wants to do business, we know who to deal with.” “And Singapore? They have Chartered Semiconductor, which I think does the same thing as you.” “They are OK, we compete. But you can get caned for jaywalking there. I’d be careful even going to Singapore.” “But why would companies trust you with their precious designs?” “That is our biggest hurdle. But we don’t compete with them. The publishing model works. We are the book makers, they write the words.” Someone has been spreading around this metaphor, but it seems to work. “Won’t design companies pop up around here to take advantage of your fabs?” “We hope so, but it will be slow. We still make a lot of clothes and toys and shoes. You know what is down the block? Animators. Actually, it is just a bunch of artists with ink and paint. They make, what is that show you have? Oh yes, The Simpsons.” “Really?” “Yes. I have watched this cartoon when I visit the U.S., and I often wonder if the animators who work here understand any of the jokes.” “D’oh,” I said reflexively. “Pardon?” “Nothing. Many of those jokes get past me too,” I had to admit. “We have a long way to go before we move beyond manufacturing. Silicon Valley will be the center of innovation for many years because they know what the market wants. We know how to make it for them. Pretty good relationship.”

And there it was—one of the top Taiwanese companies making chips for 35% profit margins “on a good day” for an American company that ships their designs over and sells the chips they get back for 55% profit margins “on a bad day.” And I’ll bet Fox Television makes more on one episode of The Simpsons than they pay Taiwanese artists to draw for them. The stock market teaches you the hard way—it’s all in the margin.

Finding great long-term investments turned out to be tougher than just finding interesting pieces of intellectual property that scale. I had to find a bulletproof way for the intellectual property to be delivered too. It couldn’t be just the price customers would pay; that could be fleeting. Yahoo was getting $50 per thousand page views. TV and radio were lucky to get $1. I didn’t think they were going to meet in the middle. I sat through a meeting with a private company that does some incredibly esoteric software to decide which data get priority on corporate networks, which are a security risk and which Web sites should be banned (sweetandsaucypizzagirls.com was one example). They had a lot of interest in their policy software, but no one wanted to pay for it. “Cisco offered $200,000 for it,” the CEO of the software company told me. “Per year? Per server?” I asked. “Nope, just $200,000. They wanted to own it, source code and all. You can’t get paid for software anymore.” “So what are you doing?” “We just implemented the same code on a chip. Turns out Cisco will pay us $50 for each chip, so we are designed into one of their switches that ships 10,000 units a month.” “And how much do you make?” I asked. “Well, it’s our design, so we can charge what we want. Don’t tell anyone, but they cost us $15 to make out of Taiwan. It makes a difference how you get paid.”     How true. It was all about how you were getting paid. Drug companies sell little patent-protected pills that cost pennies for $10 a dose. Hollywood sells copyright-protected silvery DVD disks for $19.99 that cost a buck at most to manufacture. But then again, music is copyright protected but fits in a five-megabyte file and can move across the Internet in a few seconds. The CD is broken. Same with a lot of software; it’s tough to get paid for it. It’s not just intellectual property—it’s how it’s packaged and how it’s protected.

I was warming up, taking a few shots and getting my creaky body stretched before the games started. For the last five years, I’ve been playing pickup basketball on weekends on a tennis court behind Jay Hoag’s house in Palo Alto. Jay, an old friend who used to run money in New York, is one of the top fund managers in the area, investing in both public and private companies. In fact, Fred and I modeled our fund a little like Jay’s, but different enough so that we really didn’t compete and so I could continue to play basketball at his house. A core group of successful venture capitalists and company execs showed up each week. A few were like me—white guys who can’t jump. It was hard for me because we were still trying to get our fund to a level of respectability, and here was a group of hitters in the Valley. The usual question was, “So, how much money you got?” It wasn’t so much asset envy as a desire to, well, to jump with the best of them. Because of this, I tried to avoid business talk, but I would overhear a few juicy conversations. This morning, two VCs were talking about a company trying to raise money. “Did you meet with those guys from Cybersource?” This piqued my interest because I happened to be a small investor in the company, which was really two companies, one that sold software online and another that did fraud detection for credit card transactions. “Yeah, they came in yesterday. They’ve been making the rounds.” “What did you think?” “I’m not sure.” “Me neither. My concern is that nobody’s in the deal.” This was a not-so-veiled reference that no big-name venture capital firm was an investor in the company. Without that Good Housekeeping Seal of Approval, it’s tough for a new player to step up and put in money. One of the VCs turned to me and asked, “Andy, aren’t you in this deal?” My face probably turned red, but I answered, “Yup, but I agree with you—nobody’s in this deal.” “Sorry, no offense.” “None taken,” I said. “Well, in that case, I don’t like to say nobody’s in the deal. I prefer to say that there are a bunch of nobodies in the deal.” I should have been offended at that, but oddly I wasn’t. This was white-guy trash talk—you gotta take it to be able to dish it out. Plus, we were struggling and not destined to be top-tier venture capitalists with an A-list of portfolio companies and fancy offices on Sand Hill Road. But you don’t have to be a so-called player to make money in Silicon Valley, just smart enough to find a few good companies and avoid the fashionable ones. Like public investors, VCs are notorious momentum investors—each one often investing in two or three companies in the same market, hoping something works. This was a big reason for the boom-bust cycle in the valley: VCs fund 30 plus companies in each hot segment, and after a run-up and boom in taking companies public, a shakeout is almost inevitable.

“It’s really over, isn’t it?” I said. “See those guys over there?” Nick asked. “You mean those guys with the Ashleigh Banfield glasses? Are those the Janus guys we always used to see?” “Yup, just a new look. Until six months ago, those guys used to never break a sweat. Money flowed in like water, so they would just buy more of what they owned. Helped their performance numbers. What did they care? Gemstar—Gemscar. It was just their investors’ money—they got paid either way. They rode ’em up and they’re going to ride ’em down.” “You’re fouler than usual,” I said. “Doesn’t bother me, I’ll take either side of this thing. I’m just not sure who the next sucker is to come in and buy this garbage. The Asians are broke, and the Europeans don’t buy anything without a dividend, so it’s just the dumb-ass American public pouring into the Januses of the world. I’ll take the other side of that trade every day of the week.” “You don’t think fundamentals will hold up?” “Not a chance—it was all funny money. AOL buying TimeWarner is just a farce. Even if business stays good, paying 100 times earnings and 20 times sales means they have to stay great for five or more years. How many tech cycles do you remember that have lasted that long? Any?” “Nope.” “The stock of the greatest company in the world is crap if every investor already thinks it is the greatest company in the world.” Yup, sometimes you gotta wait for the next waterfall.

What startles me is that those who generate wealth in Silicon Valley run at 100 miles per hour. They don’t own anything of value like a textile mill or an auto factory. They own a process, the ability to constantly update their products and take advantage of that waterfall, some massive price declines and then move on to the next product or process.

The world had changed. The mighty economies of Japan and Korea and Thailand are not taking over. Their output of cars and laptops and VCRs and DVD players and memory chips and computer monitors and sneakers is booming, but something’s wrong. They have giant factories with lots of lower-wage workers who wind wire, screw screws, bolt bolts, wrap plastic and stick in power cords. But that’s not what anyone pays for anymore. Their economies achieve full employment, sure, but these countries are not economic powerhouses. Not anymore. We were investing in companies with no more than 50–100 workers, most of them highly paid programmers and engineers, whose occupational hazard is coming down off a caffeine buzz and an occasional late night Nerf gun injury. Yet even after the market bubble burst in 2000, these companies would still be worth more than Ssangyong, a company a hundred or a thousand times their size. The stock market values small businesses with high margins over big businesses with low margins. Is that good or bad? Should I even care? Whenever I try to figure out why this is, I keep thinking back and visualizing Mr. Shim, a walking, talking and sweating metaphor for how to invest. Something like “We think, they sweat.” The spoils go to those with high margins.

You can make intellectual property, but real soon it’s worthless. You don’t really own anything tangible, just the ability to move it along, kick the can down the road, as diplomats like to say. Ask economists, and somewhere in their babble they will tell you that the role of an economy is to increase the standard of living of its participants. Did the boom-bust, boom-bust yo-yo I just lived through do that? I think so, but with lots of change. America doesn’t make stuff anymore—we design it. The numbers are fishy, but even after the rocky start in the 1990s, I think this new think/sweat thing will create more wealth for more people, not just in the U.S. but around the world, than the Industrial Revolution ever did.

It’s about expanding minds, not expanding muscles. That is the secret of creating intellectual property. It is also the secret to finding the profits. The cost of humans is a great umbrella. If it costs $10 to sort each one of millions of airline tickets and $2 to sort billions of bank checks by $10–20 per hour human beings, you can charge millions of dollars for equipment. It is easy to profit from automating these tasks. For a while anyway, and then you have to move on and automate other, more difficult human tasks to extract profits. We pay $40 per month for cell phone service because of the cost of the alternative—staying home or using pay phones. People costs drive demand. Intellectual property augments people, which lowers costs, which drives profits that create wealth, although not always in obvious ways. Napster begat Elantec, and Netscape begat Cisco. Not every big market makes for great investments. I always go back to the way I positioned IPOs to figure this out. Monster markets are great, but they have to be coupled with an unfair competitive advantage and a business model to leverage that advantage. Is that too much to ask for? Anyone who bought MP3.com envisioned only the monster market (if that). Investing will never be that easy—it’s all in the details of the business and what other investors think.

Mr. Zed: My grand unifying theory of why the U.S. is still the best place to invest. Rejected in D.C., but here it is:      I just love American customs and tradition. My favorite? Well, years ago, I got stuck in Detroit Rock City, on the night before Halloween, so-called Devil’s Night. The Motor City was glowing. Quite literally. It turns out that Detroit’s quaint tradition revolves around arsonists who randomly select parked cars to burn. Well, not randomly, 99.9% of the time, it is a Toyota or a BMW or other foreign-made cars (any Yugos left?) that get torched. Xenophobia? Patriotism? Loud message to job-stealing foreigners? How about just amazingly backward? These industrial-age “hoodlums” ought to be lighting up a few Fords and GMC trucks if they want to create more local jobs.      You see, car manufacturing used to be a great business. In 1962, General Motors made just over half of all cars sold in the U.S. market and had profits of $1.5 billion, the equivalent of $9 billion today. Now they are struggling to sell one in four cars and work on wafer-thin margins. But if foreigners want to bend sheet metal and tighten bolts for less than $20 per hour and still make barely 1% of sales as profit, why should we stop them? Instead of providing capital to U.S. automakers, Wall Street, as it always does, focuses on businesses that make higher returns. It’s creating its own world order.      History does rhyme. In the 20th century, the resource-rich U.S. took over leadership of the industrial era. But now it is grappling with a painful transition to an economic system based on intellectual property rather than factories. We are all used to an industrial America, the Dow Jones Industrial Average and all that. But now all of those great traditions sparked by Watt and Wilkinson are coming to an end.      There is a classic Wall Street adage that says, “Money sloshes around the globe, seeking its highest return.” It’s been sloshing away from Detroit for decades. But where is it sloshing to? That’s where I want to be.

In the U.S. the Industrial Revolution is dead. Kiss it goodbye. Heating, stirring, mixing, stamping and bolting have all been played out. These are no longer things that make America great. Instead it’s, uh, thinking. I know what you’re thinking, “Why didn’t I think of that?” It’s what people who think they are smart call intellectual property. This IP is the Silicon Valley model.      But how did we get here? The biggest change since World War II is that design and manufacture are no longer linked. Computers and communications move designs around in nanoseconds, while industrial-era factories locate near cheap labor.      A computer on a chip, an operating system, a wireless packet switching network—these are all highly intelligent properties that come from mind instead of matter. But so is a Nike swoosh, an anti-intelligent Adam Sandler movie or Baywatch rerun, a pill to stop the runs from eating a Happy Meal, Vanilla Diet Coke and a venti double decaf blended caramel macchiato with a twist at Starbucks.      The 225-year-old Industrial Revolution never did burn out but merely faded away into a pit of profitlessness. On the flip side, intellectual property is highly profitable, which makes Wall Street squeal with delight. Money would flow uphill, if it had to, to fund these high-margin intellectual enterprises. Money sloshes to margin!      The stock market sorts all this stuff out. A stock is nothing more than the current value of a company’s future profits. Capital sloshing around, seeking its highest return, naturally funds highly profitable companies.      General Motors’ days were numbered when Wall Street figured out the company couldn’t dominate as they did back in 1962. GM should have been creating subsidiaries in Japan rather than trying to keep Japanese imports out of the U.S. Wall Street probably would have provided them all the capital they wanted.      As the Pink Panther Inspector Clouseau said when told, “That’s a priceless Steinway baby grand piano,” right after he had smashed it with a knight’s chalice stuck on his arm: “Nuut anymeure.”      Give GM expansion capital? Not anymore. Over time, stock markets are all-knowing and very persuasive. It’s the stock market that is leading the change to the intellectual property economy we are in today, by canceling the credit cards of companies that don’t fit the model. You have to squint to see it happening now, but it will be more and more obvious every day.

The whole Industrial Revolution was based on a premise that if a company designed it, they had to make it to get paid. No more. What began as coal-powered, steam-belching, industrial-factory-based economy has morphed into a highly elastic, horizontally organized, globally distributed, sleek, efficient economic system that a government-knows-best Keynesian would require a cough-syrup–induced hallucination to dream up. Instead, it happened naturally, based on labor-saving technology and disparate worker costs. The strangely named Fabless Semiconductor Association, whose member companies strictly design chips and don’t own factories, has over 400 members, adding more than 100 during the downturn of 2002 alone. These are the high-margin companies in Silicon Valley. Intellectual property now rules.      Everything you know about what makes the world work—sovereign nation states, balance of trade, gold standards, fiat money as well as car torching—is up for grabs.     Simply stated, the world economy now runs on U.S. intellectual property. Santa Clara and Redmond sell chips and software at disgustingly high profits to foreigners who turn them into personal computers at disgustingly low profits. It’s a great trade. We can buy a PC for $799, and they (China, Malaysia, Thailand) employ their unwashed masses. Our masses are already washed, heck, coiffed. In the U.S., the typical autoworker makes over $70,000. At an average manufacturing salary in the U.S. of $20 an hour, we can’t afford to bond integrated circuits, wind inductors for read-write disk drive heads or stuff circuit boards. That’s buck-an-hour or buck-a-day work. Exploitive? If we weren’t paying a buck an hour, many third-world workers would be shoveling shiitake in their gardens for a buck a month. In fact, our appetite for multi-gigahertz PCs to surf for cheap prices and Britney videos is what creates jobs and a rising a middle class in the rest of the world. And slowly but surely, that middle class will consume our intellectual wares, as they can afford it.      That’s the thing about intellectual property. It usually has a huge up-front cost to create and almost zero marginal cost to sell. One more copy of Windows or one more arthritis pill costs practically nothing. The next one sold is pure profit. It’s in the best interest of Microsoft and every high-margin company in the U.S. to have a growing middle class worldwide to buy more of their stuff. Memo to Karl Marx: You don’t exploit for profit anymore. Instead, you expand and enhance. Quite a change.

Whole economies are based on a disengagement of design and manufacturing—of intellectual property and industry. And just in time, or progress in the U.S. would have come to a screeching halt, and you and I would be waiting in line to load our punch cards into IBM time-sharing computers.      The British milked a lead in steam engines into a hundred-year empire, imported raw materials, ran them though foul coal-burning, steam-engine–run manufactories and then exported value-added finished goods. Their profit was a huge markup above the cost of coal. Homespun textiles, metalworks and pottery made in the rest of the world were more expensive and of lower quality. It wasn’t even fair. The British wiped up and built their empire.      Now it’s the opposite, which is why the Industrial Revolution is finally dead. A very bright star in Sunnyvale, California, e-mails the design for her chip to Taiwan, where it is manufactured and then shipped to Shanghai to be assembled into a Toshiba laptop. They keep a few for themselves and sell the rest back to us for margins so low that we can’t afford to do it ourselves. The U.S. exports high-margin intellectual property, which others add sweat to and sell back to us at tiny markups. It’s one of those funny win-wins that drives economists crazy (crazier?).     You may ask, “Uh, someone said something about my needing to drive a Beemer?”      OK, OK. It takes lots of dollars to buy chips and software and Viagra and McNuggets and Jerry Lewis movies. No problem, we just have to buy something from other countries, to put dollars in their hands.      In the early 19th century, the British Empire was almost stillborn. In 1815, the long inflationary Napoleonic Wars ended, and sure enough, wheat prices collapsed. Farmers began turning in their pitchforks and moving to the cities in droves to work in manufactories. Steam engines were driving textile mills to allow the British to sell cheap and comfortable clothing to the world. In fact, one particular machine, Samuel Crompton’s spinning mule, hooked up to a Boulton & Watt steam engine, would repeatedly stretch and wind cotton thread and yarn until it was as “smooth as silk,” like Kessler Whiskey. No one making clothes at home anywhere could match these mills for either cost or quality in terms of smoothness.

But landowners who controlled Parliament, and these farmers, who didn’t know their ashes from their bellows, passed the Corn Laws. These tariffs set minimum prices on agriculture and kept out cheap corn and grain from the Continent (read, France). Workers started starving because they were not making enough in the factory to pay for now expensive bread. But factory owners couldn’t raise wages because they were having trouble selling their manufactured goods overseas. Why? Because the French and Germans were paying for these goods with their wheat and corn, and the British taxed them out of affordability. How stupid—this almost killed the British Empire in its infancy.      With any foresight, the landowners should have dumped their unprofitable farms and invested the proceeds in highly profitable joint stock companies making pottery, shirts and potbelly stoves. England should have gladly bought French wheat and Dutch flowers and German barley and hops so that consumers in these countries could have turned around with the money they received and bought British manufactured goods. There was no substitute. Once you go silky smooth, you never go back. The Corn Laws foolishly lasted until 1846.      So, go ahead, buy that Beemer so that Germans can afford to buy our software.   Wait a second, I hear you screaming, “The U.S. is running trade deficits as deep as the Mariana Trench, $500 billion a year or more.” Relax—it’s just money, and funny money at that. You see, using industrial-era measurement tools, “money out, goods in,” the common perception is that consumers in the U.S. are running these massive trade deficits, mortgaging their future and putting the health and well-being of the U.S. in the hands of devious foreign strangers—to xenophobes, are there any other kind?      Which reminds me of a great story:

An economist and an investor are helplessly lost on a hike through the peaks and valleys of the Dow and NASDAQ mountain ranges, well, the Rockies. The investor sticks a wet finger in the air to find his direction via the prevailing wind. The economist is studying charts and numbers, GDP growth, trade stats, unemployment data, inflation, hours worked, productivity, meticulously compiled by the Commerce Department, the Bureau of Labor Statistics and Bureau of Economic Analysis. “OK, I’ve got this figured out,” the economist yells. “You see that big mountain over there?”

     “Yup,” sighs the investor.      The economist proudly announces, “We’re on top of that one.”      According to government statistics, the U.S. has been running trade deficits, except for a couple of brief surpluses, since 1976. But that’s like the government counting railroad ties to determine the strength of the U.S. economy—it’s the wrong measure.      The cumulative deficit since 1976 is around $4 trillion. Most alarmists and car burners equate that with spending our grandchildren’s future, that the U.S. has spent $4 trillion more than it produced, and our grandchildren will practically be in slavery to pay back our “hog” iness.      If this was just German beer or French wine that we were quaffing in mass quantities and running deficits to do it, I’d be a little nervous. But, no, the intellectual property we export and then reimport in finished goods is the type that makes us more productive. Computers, with our designs but Taiwan-made chips and Malaysia-assembled disk drives, run our corporate back offices. Similarly created digital networks now deliver phone calls, and other systems deliver e-mail cheaper every year.      Looked at another way, foreigners made us $4 trillion of stuff, in exchange for a piece of paper with Ben Franklin’s mug on it. But we get it all back. Charity? Nah. They buy a piece of debt signed by Alan Greenspan, promising to pay them interest and their principal back in 10 years. Or maybe a colorful stock certificate. So, either foreigners are suckers or there is some economic reason for these deficits.      We got most or all of the 4 trill back, chasing our profits. Do they now own the U.S.? Nah. The entire U.S. bond market is $19 trillion, stocks another $15 trillion and growing.

  A not-so-well-kept secret that Wall Street has figured out: creating intellectual property is lucrative. Microsoft sells Windows for $50 with virtually no physical costs, just massive research and development costs. Even at the bottom of the bear market of 2002, Microsoft, with its 40% operating margins, was the second most valuable company in the U.S. GE, a giant hedge fund, was #1.      Money comes out of the woodwork to chase these kinds of returns. Wall Street is an essential conduit. Excess capital at our trading partners, the “beneficiaries” of our drooling gluttony, comes right back and invests in the U.S.      Economists measure GDP, Gross Domestic Product, and live and die by its direction. People who run money look at gross domestic profit, since that’s all that really matters. The Dow trumps GDP—that sigh you just heard was industrial economists fainting into a heap on the floor.      It would take 40 companies the size of Microsoft making 1% operating margins to generate the return that Microsoft does on their intellectual property. That’s a lot of factories in a lot of countries making stuff for us. No wonder the deficit is $4 trillion and Microsoft is worth more than most Continents. And the deficit is probably going to get larger, but as Alfred E. Newman would say, “What, U.S. worry?”    But don’t we owe someone $4 trillion? Sort of, not really—aw, screw it, no. They can get their money back eventually, but wealth beyond those four big ones will be created in the U.S. if we put their money to work effectively. When capital sloshes around and chases high returns, much of it gets invested in our stock market. And unlike a passbook savings account at your local bank, there is no account statement that reads $4 trillion. Bean counters need solid beans to count; if they are smashed and stirred into a soup, no one really knows how many beans there are.      A million dollars invested into a company’s stock over a week or a month can increase the value of that company by a hundred million dollars. It works both ways: a million dollars sold can decrease a stock’s value by that or more as well. Dollars flowing into the U.S. are lost in the bean soup of the market, but their effect is to lower the cost of capital for great companies. The $4 trillion may have increased wealth by $7, 8 or 9 trillion. Who knows? Except to say that it increases wealth.      Economists may insist we are running a trade deficit, but in reality, we are running a margin surplus. In a world where money crosses borders faster than J-Lo goes through boyfriends, profits are all that matter. The margin or profit over cost of goods and services is the only true measure of trade. I will gladly trade you an idea I make $1 profit on for five widgets that you only make a nickel on. And I’ll make that trade every day of the week.

As long as the U.S. focuses on intellectual property, perhaps we’ll never have to pay the “debt” back. It will just be constantly rolled over and shrink in relative size as foreigners buy our intellectual property and chase our returns. In the long run, the dollar will probably rise in value. But who really cares what the dollar does—it’s just an economic construct. Real wealth is created by profits, not currency printing presses. Currencies trade more on relative interest rates than anything else—this was the secret of the Soros and Tiger successes.As long as the U.S. focuses on intellectual property, perhaps we’ll never have to pay the “debt” back. It will just be constantly rolled over and shrink in relative size as foreigners buy our intellectual property and chase our returns. In the long run, the dollar will probably rise in value. But who really cares what the dollar does—it’s just an economic construct. Real wealth is created by profits, not currency printing presses. Currencies trade more on relative interest rates than anything else—this was the secret of the Soros and Tiger successes.      Prevailing industrial economic thought has trade balancing via exchange rates. If the U.S. is running deficits, the experts say, we should lower the value of our currency to make our goods more competitive on the world market, sell more of them and eventually bring trade back into balance. But that theory is flawed for intellectual property. Money is flowing in, but perhaps our stuff (companies, intellectual property, etc.) is too cheap!      Things are now backward. A PC maker has to buy Windows XP no matter what the exchange rate. If the dollar is rising, then it takes more Chinese yuan and more Thai baht and more euros to buy Windows. Lowering the dollar won’t help Microsoft sell a single additional copy, since there is no substitute. A rising dollar might perversely shrink deficits as foreigners are forced to pony up more of their currency for our intellectual property and make our reward, those leather-clad BMWs cheaper—an upside-down J-curve.      According to the International Monetary Fund, in 1994, foreigners owned 11% of the total value of U.S. long-term securities. At the end of 2001, they owned 18.3%. Their share of stocks went from 7% to 12.4%. But the stock market, even at the bottom of a bear market, more than doubled in that same time period. So even though foreigners own more, the U.S. is wealthier. Isn’t that the right end game? At some point, as low-profit industries move out or close altogether, U.S. margins will be high enough to make the economy cash-flow positive. We won’t need the economist’s phrase “kindness of strangers” anymore. No kindness here—just money chasing returns.      This may sound like voodoo economics, but only to those who see the world through the eyes of the Industrial Revolution. Economists live in Washington and New York, not Silicon Valley. Margins are making economists marginal.      Prevailing industrial economic thought has trade balancing via exchange rates. If the U.S. is running deficits, the experts say, we should lower the value of our currency to make our goods more competitive on the world market, sell more of them and eventually bring trade back into balance. But that theory is flawed for intellectual property. Money is flowing in, but perhaps our stuff (companies, intellectual property, etc.) is too cheap!      Things are now backward. A PC maker has to buy Windows XP no matter what the exchange rate. If the dollar is rising, then it takes more Chinese yuan and more Thai baht and more euros to buy Windows. Lowering the dollar won’t help Microsoft sell a single additional copy, since there is no substitute. A rising dollar might perversely shrink deficits as foreigners are forced to pony up more of their currency for our intellectual property and make our reward, those leather-clad BMWs cheaper—an upside-down J-curve.      According to the International Monetary Fund, in 1994, foreigners owned 11% of the total value of U.S. long-term securities. At the end of 2001, they owned 18.3%. Their share of stocks went from 7% to 12.4%. But the stock market, even at the bottom of a bear market, more than doubled in that same time period. So even though foreigners own more, the U.S. is wealthier. Isn’t that the right end game? At some point, as low-profit industries move out or close altogether, U.S. margins will be high enough to make the economy cash-flow positive. We won’t need the economist’s phrase “kindness of strangers” anymore. No kindness here—just money chasing returns.      This may sound like voodoo economics, but only to those who see the world through the eyes of the Industrial Revolution. Economists live in Washington and New York, not Silicon Valley. Margins are making economists marginal.

 If Toshiba wants to put some liquid crystal display and plastic around an Intel Pentium and Microsoft Windows, and sell it to me as a laptop, let them. In fact, they have to do it—it’s their contribution to the world economy. The entire margin on PCs is around $400, $350 of which is split by Wintel (Microsoft and Intel) and other U.S. companies. If GM made the exact same laptops, they would cost $10,000 a pop, weigh 40 pounds and need a battery change every 3,000 miles.      If Sony wants to sell you a big honker, 60-inch diagonal comb filter DLP or LCOS cable-ready TV for practically no profit, should we stop them? We should instead encourage them. They go to the dark edges of western China, seeking out cheap labor to grow the tubes or assemble the flat panels and then move them to assembly plants all around the South China Sea. Why? Because it keeps Sony employees fully employed. Now you know why they bought a studio in Hollywood: anything to add value. But if you and me buying that Sony or Sharp TV means Japanese children can buy Kentucky Fried Chicken and go to Jim Carrey’s next movie, and upgrade Windows 98 to Windows ME to Windows XP, so be it. Jerry Lewis will tell you—do it for the children.      But never underestimate the ability of policy makers to stick with oxymoronic conventional wisdom. Even if the dollar stays where it is and we go to $6 trillion in trade deficits, our stock and bond markets might be worth $50, $60, $70 trillion, double or triple today’s value, run up by margin chasers. Just about everything about this margin surplus model is upside down.      The modern U.S. has farm bills and textile quotas and on and off steel tariffs like the British Corn Laws, not to mention late-model-car flamers. How dumb. Because of our margin surpluses, big trade deficits are our just dessert. These foreign-made consumer items are our gold. Let them flow. Tariffs, quotas and subsidies will return us to an industrial age. No thanks. Foreigners sweat and toil to make our physical delights, in exchange for our intellectual output. It doesn’t get any better than this.

So, Mr. Zed—there it is. A paradox? Yes! Investing where others can’t know, unless they stand on their heads. That is my edge—someone tell Jack Nash. I know it sounds wrong, counterintuitive, like eating your spinach, but for Americans in an intellectual property era, we all just have to suck it up and toodle around in our Mercedes with Corinthian leather. And drink French wine and watch the biggest-ass Japanese TVs we can afford. And wear clothes made in Ghana. Someone has to help those in other countries to be able to afford our ideas. Whatever dollars we send them will come back like a boomerang to buy our high-margin stuff—the fifth season of The Sopranos on DVD to the next release of code to run Cisco routers, or maybe they’ll just buy our stocks. Who cares—there are great investments to find either way. That is the big waterfall. This works for all intellectual property, but I’m sticking with things that scale. Warning. Even the name intellectual property is a paradox. It wrongly implies you can own something and have wealth. It’s not true. Intellectual property means constantly improving on something to continue to generate high margins—else the market price collapses. None of this stuff is static; it constantly evolves. Microsoft doesn’t sell DOS, their original disk operating system, anymore, even though that was their key piece of intellectual property. It’s worthless today. But their process of invention and innovation, even if it is painfully slow, is enormously valuable because of its 99% gross margins. Their only material costs are CDs and a thin manual. Even after spending billions on research, they make 50% operating margins and can afford to spend more each and every year on research. Nothing static about it. Wealth is a process. If I learned nothing else from running money, that is it.

Running money is supposed to be unemotional. Get conviction on things others don’t or can’t know about and find future returns. But the stock market is also a force of change. It is the embodiment of Adam Smith’s invisible hand. It funds growth and starves dying businesses and pushes progress. Not always, I suppose, but enough. With intellectual property and the margin surplus the U.S. is running, the stock market plays a key role in balancing payments and currencies and how the world now works. In an odd way, being a player in the stock market is a form of activism, albeit indirect and invisible activism, in pushing a model that increases living standards. But it’s hard to be unemotional for long. It’s about ideas that affect people. It’s about finding waterfalls that create massive growth and monster markets and change for the better to the status quo. One hundred thousand or more people running money all day, every day, most staring at their screens, doing trades, providing access to capital, sloshing capital around. Whether we realize it or not, we are all driving progress, not individually, but collectively. Adam Smith’s invisible hand? If you like. That’s what a stock market does. But in this postindustrial, intellectual property world, it’s more like Doug Engelbart’s visible hands, clasped together, rising up, scaling knowledge, augmenting humans.