翻译过来是幂律定律，是今年出的新书，一本当代VC史的最好作品，没有之一，值得从业者反复读几遍，收获非常大。按惯例做些摘要，同时再补充些个人评论和体会。
What he didn’t mention was that Khosla was also a preacher of the Valley’s most bracing creed: the belief that most social problems can be ameliorated by technological solutions, if only inventors can be goaded to be sufficiently ambitious. “All progress depends upon the unreasonable man,” the “creatively maladjusted,” Khosla declared, borrowing eclectically from George Bernard Shaw and Martin Luther King Jr. “Most people think improbable ideas are unimportant,” he loved to add, “but the only thing that’s important is something that’s improbable.” Khosla wanted radical dreams, the bolder and more improbable the better.”
Joining the storied venture partnership of Kleiner Perkins, Khosla discovered his true métier. His unreasonable impatience—his determination that anything might be possible and everything should work his way—made him one part tyrant, two parts visionary.
VC是需要有特质的人的，也需要特别的思路。科斯拉Khosla其实就很有代表性，相信大部分的社会问题可以通过技术和商业来解决，那些看似不可能的事情才是重要的，或者说突破后才是价值巨大的。那些看似疯狂的梦想，越是大胆、越是看起来不可能，就越有价值。
He noted matter-of-factly that it’s a trillion-and-a-half-dollar global market being served by prehistoric technology.
要寻找全球万里级市场中依然在使用古老的史前技术的赛道，这就是十分有机会的赛道。
Which is better: to try and fail, or to fail to try? Reasonable people—well-adjusted people, people without hubris or naïveté—routinely fail in life’s important missions by not even attempting them; the way Khosla saw things.
大多数人是没有勇气去尝试，能通过尝试验证后知道失败是十分难能可贵的。
This sort of skewed distribution is sometimes referred to as the 80/20 rule.In reality, there is nothing magical about the numbers 80 or 20: it could be that just 10 percent of the people hold 80 percent of the wealth, or perhaps 90 percent of it. But whatever the precise numbers, all these distributions are examples of the power law, so called because the winners advance at an accelerating, exponential rate, so that they explode upward far more rapidly than in a linear progression.
Horsley Bridge is an investment company with stakes in venture funds that backed 7,000 startups between 1985 and 2014. A small subset of these deals, accounting for just 5 percent of the total capital deployed, generated fully 60 percent of all the Horsley Bridge returns during this period.
幂律分布其实是VC生意的市值，不只是80/20，而是5%的好项目获得了60%的回报，是大满贯游戏。
Khosla made extraordinary claims for his work. Venture capital was not merely a business; it was a mindset, a philosophy, a theory of progress. Seven hundred million people enjoyed the lifestyle that seven billion wanted, he liked to say. Bold innovators goaded by even bolder venture capitalists offered the best shot at satisfying human aspirations.
VC不只是一个生意，也是一种能带来社会进步的方法、思维方式和哲学，通过大胆支持勇敢的创新者来更好满足人类的需求和欲望。
The revolutions that will matter—the big disruptions that create wealth for inventors and anxiety for workers, or that scramble the geopolitical balance and alter human relations—cannot be predicted based on extrapolations of past data, precisely because such revolutions are so thoroughly disruptive. The future can be discovered by means of iterative, venture-backed experiments. It cannot be predicted.
技术革命和破坏性创新是无法基于过去的数据被预测出来的，而是要通过持续迭代来试验出来的。所以风险投资很大程度上其实不是什么理论，而是一门实实在在的实践科学，一件创造未来的事情。
Experts may be the most likely source of incremental advances, but radical rethinks tend to come from outsiders. “If I’m building a health-care company, I don’t want a health-care CEO,” Khosla says. I want somebody really smart to rethink the assumptions from the ground up.” After all, he continues, retail innovation did not come from Walmart; it came from Amazon. Media innovation did not come from Time magazine or CBS; it came from YouTube and Twitter and Facebook. Space innovation did not come from Boeing and Lockheed; it came from Elon Musk’s SpaceX. Next-generation cars did not come from GM and Volkswagen; they came from another Musk company, Tesla. “I can’t think of a single, major innovation coming from experts in the last thirty, forty years,” Khosla exclaims.
大的创新很少是行业内的专家能够搞出来的，往往是外行跨行搞出来的。这件事已经被反复证明过了，但依然难以令人相信。这其实是VC行业的基础假设之一：用新技术把传统行业重新做一遍。
If the future is best discovered by means of maverick moon shots, another insight follows. Thanks to the work of the Nobel laureate Ronald Coase, the economics profession has long recognized two great institutions of modern capitalism: markets, which coordinate activity via price signals and arm’s-length contracts; and corporations, which do so by assembling large teams led by top-down managers. But economists have focused less on the middle ground that Khosla inhabits: the venture-capital networks that lie somewhere between markets and corporations. And yet networks of venture capitalists deserve closer attention. Venture capitalists have achieved this disproportionate impact because they combine the strengths of the corporation with the strengths of the market. They channel capital, talented employees, and large customers to promising startups; in this way, they replicate the team formation, resources, and strategic vision to be found in corporations.
这是对VC最有趣的理论解释了，科斯的制度经济学把市场和企业作为现代资本主义的两个基础，市场基于价格信号来协调，而企业则是上下级制度来协调，科斯研究的是企业的边界。但Khsola认为VC则集合了企业和市场的双重优势，既可以通过市场机制来连接人才、募集和分配资金、满足需求，又能通过创设企业来组织团队、资源实现市场的需求。这个观察是很到位的，这其实就是VC的神奇之处。
再换个角度看，往往失败的VC就是在这二者中产生了偏离，KPCB在互联网时代All in Cleantech的失败其实是忽略了市场的需求，把宝押到了政府补贴上；今天不少国内引导基金的过分返投要求实际也是在偏离市场规律，让基金投资到要素不足的市场，这其实非常危险；另一个偏离就是外行进来搞VC，不懂得企业运作的规律，这种偏离也十分危险。VC实际还是个蛮高专业度的事情，既要熟悉市场，又要懂得企业，从业务到财务到管理，不容易。
Most financiers allocate scarce capital based on quantitative analysis. Venture capitalists meet people, charm people, and seldom bother with spreadsheets. Most financiers value companies by projecting their cash flows. Venture capitalists frequently back startups before they have cash flows to analyze.
读完这本书才发现所谓成长性投资其实是新生事物，传统的VC投资就是早期，现金流产生之前的创造性工作和眼光，创造这个从0到1的过程，不靠量化分析。今天我们的VC投资洋洋洒洒写报告、做预测，并非错误，但万不可因为写材料就认为阅人、识人的判断次要了，不能主次颠倒。写材料是为了事后求证、复盘和反思，以及完善自己，不是为了投资项目。项目判断的本质依然在人，不在材料。
1 ARD & Fairchild Rock as FA
The first personal computer was the Altair, created in New Mexico. The first precursor of the worldwide web, the network-management software Gopher, was from Minnesota. The first browser was developed by Marc Andreessen at the University of Illinois. The first search engine, Archie, was invented by Alan Emtage at McGill University in Montreal. The first internet-based social-networking site was SixDegrees.com, launched by Andrew Weinreich in New York City. The first smartphone was the Simon Personal Communicator, developed by Frank Canova at IBM’s lab in Boca Raton, Florida. No single geography—not even Silicon Valley—dominates invention. And yet all these breakthrough products have one thing in common. When it came to turning ideas into blockbuster products, the Valley was the place where the magic happened.
谁第一个发明不重要，重要的是成为第一个家喻户晓的产品。硅谷的神奇之处就是虽没有产生这些发明，却借着这些发明产生了一个个家喻户晓的产品，以及他们背后的一批伟大的公司。这是个非常重要的历史经验，如何从好的发明走向伟大的产品？这就是值得被反复思考的专业问题。
As part of his efforts, he launched a $5 million fund to safeguard the spirit of free enterprise by providing investment capital for entrepreneurs. But after five years in operation, J. H. Whitney & Company had backed just eighteen ventures; his successes included an early maker of the building material perlite and Vacuum Foods, the producer of Minute Maid orange juice. In his first five years, moreover, Whitney outperformed the much safer S&P 500 by a relatively modest margin. Indeed, on the risk-adjusted basis that financiers use to measure themselves, the fund could not justify its existence.
1946年惠特尼的伟大尝试的结果并不理想。5年翻倍算不上差的业绩，同期是45%的通胀以及75%的标普回报。
In April 1946, the Rockefeller family launched a parallel effort to Whitney’s, aiming to solve the generally acknowledged lack of finance for new firms. “What we want to do is the opposite of the old system of holding back capital until a field or an idea is proved completely safe,” declared Laurance Rockefeller, the prime mover, adding, “We are putting money into many underdeveloped areas.” In 1961, Barron’s reported that Rockefeller Brothers had returned $40 million on the $9 million invested during fifteen years in business.The S&P 500 had shot up by 600 percent over this period.
洛克菲勒家族也是从1946年开始尝试VC，完全是从解决社会问题、慈善的角度开始的，这种伟大的社会责任感真是现代社会和当下中国无比缺乏的东西。但就这点上讲，咱们的差距何止百年。15年的试验9m增值到40M，4.4x的回报放今天也不算差，但同期的股市回报是6x，相比之下就有点黯然了。
To a modern venture capitalist, this pitch would have been instantly attractive: the founders came from a cutting-edge research lab, and they proposed to commercialize a technology that was already proven. But in the financial climate of the 1950s, even the most compelling scientists had a hard time raising money, and Doriot exploited this circumstance to the maximum, making Digital’s founders an offer that by later standards would have counted as an insult. ARD would provide a $70,000 investment and a $30,000 loan in return for 70 percent of the company: it was a “take it or leave it” offer. Lacking any alternative, the MIT professors accepted; nor did they protest when Doriot managed to push his stake up to 77 percent. Having seized so much of the equity, Doriot was in a position to profit massively when the professors succeeded. By the time ARD closed in 1972, it had reaped some $380 million from its bet on Digital—or $2.3 billion in today’s money. It was a bonanza that accounted for perhaps 80 percent of all the gains that ARD generated over a quarter of a century.
ARD投资的DEC项目是当代VC史上当之无愧的开端，即便是顶尖的科学家团队和可靠的商业化前景面前，7w美元投资/70%股份+3w美元贷款的交易结构估计会惊掉不少现代人的下巴，但这就是最真实的历史真相。当然最值得称道的莫过于当初这10w美元投资在基金结束时回报了3.8亿美元，4000x的回报。换个角度看这个回报是ARD整个25年历史上全部回报的80%，这就是幂律定律的神奇之处。
Kleiner’s wife, Rose, typed up the letter, dated June 14, 1957. Then she mailed it to the New York firm of Hayden, Stone. The broker who had served Kleiner’s father was preparing to retire from Hayden, so he passed the letter on to a young MBA named Arthur Rock. And yet, as luck would have it, Rock was the perfect person to receive Kleiner’s letter. Two years previously he had assembled a funding package for General Transistor, the first independent manufacturer of germanium semiconductors, which were to be used in hearing aids. On June 20, 1957, Rock placed a long-distance call to Kleiner, assuring him of his interest. The next day he wrote Kleiner a letter, urging him to keep his team together until they could meet face-to-face. The following week Rock flew to San Francisco, accompanied by a glad-handing Hayden partner named Alfred “Bud” Coyle.
历史固然有很多巧合，但还是一遍遍阐明“机会总是留给那些有准备的人”这样简单的道理。1957年仙童半导体故事的神奇之处在于这个时间，当时的通信和出行方式下，公司6月14号创始人老婆写好了BP寄出来到纽约，横跨美国估计顾上要不少时间，Rock收到后6月20号就回了电话，表达了意向，这个速度和习惯放到今天都可以谈的上神奇，21日还再次回信给创始人要求见面，下一周估计6月底就飞到了旧金山，还带上了合伙人…虽然有Rock刚搞过一个二极管公司的融资有关，但如此神奇的执行力，不值得我们好好学习吗？半个月的时间而已，能行动的如此迅速，这样的人，不成就一点事情都难。不只是当年FA的榜样，也是今天投资人的榜样，特别是当年机会遍地尚且如此，何况高度内卷的今天呢。
The researchers said they needed $750,000 to get their business started. Rock and Coyle countered that they should have at least $1 million. The Wall Streeters were projecting more confidence than was really justified: finding north of $1 million to launch an untested collective would not be straightforward.But the financiers’ bravado served to win over doubters. With the promise of a seven-figure funding package, whatever resistance the researchers had been feeling began to melt away.There were no more reasons not to go forward. Bud Coyle pulled out ten crisp dollar bills and proposed that every man present should sign each one. The bills would be “their contracts with each other,” Coyle said. It was a premonition of the trust-based contracts—seemingly informal, yet founded, literally, on money—that were to mark the Valley in the years to come. After approaching thirty-five potential investors, Rock had failed to raise a single cent. Then Bud Coyle suggested Sherman Fairchild, a playboy with an inherited fortune who was a self-described “putterer” and science enthusiast.
仙童的融资故事也非常有意思，创始人要75w，FA说需要100，找遍了35个投资人，却没有融到1分钱。现实虽然骨感。但当初FA的自信是极其重要的，和创始人完成了互相忽悠、互相相信的过程，10美元钞票上签字的事情也颇具仪式感，把创始人团队凝聚了起来。苍天不负有心人，看似纨绔的富二代成就了这一切，反而实现了可能上一代都无法实现的价值。
Each of the eight founders was invited to put up $500 in return for 100 shares in the startup. The men scraped together the money, but not without difficulty; $500 was two or three weeks’ salary, and Noyce had to call his parents to ask if his grandmother could lend him the funds.For its part, Hayden, Stone bought 225 shares at the same price as the founders, and 300 more shares were set aside in a kitty to help recruit senior managers; despite Noyce’s charisma, Fairchild saw him as merely the interim chief. Each founder was therefore left with just under 10 percent of the company, with the prospect that this share would fall to 7.5 percent when new management materialized. Meanwhile, Fairchild’s company, Fairchild Camera and Instrument, was stumping up just about all of the initial capital—some $1.4 million, dwarfing the $5,125 put up by the scientists and Hayden. But because Fairchild’s money came in the form of a loan rather than equity, the founders’ ownership was not being diluted. The Fairchild loan was actually not really a loan: it came bundled with an option to purchase all of the new company’s stock for $3 million.
仙童当年的结构还是非常经典的，融资主要是债权人投资了140w，附带300w收购全部股份的期权。股份上1325股，每股5美元，8个创始人合计持股60%，给新人预留的期权大约23%，FA按照创始人价格认购了17%，完全一个新创的公司，大家进行了种子轮投资，公司估值才5000左右，300w的未来收购价相当于种子轮估值的600x，对于团队以及投资人来说，其实给的收购价不低的，很合理。
In 1959 Fairchild took in orders worth some $6.5 million, thirteen times more than the previous year. The young company’s after-tax earnings already came to about $2 million, and given its huge operating margins, it had every reason to expect windfall profits as its sales volumes ramped up.The news was so excellent, in fact, that Fairchild Camera and Instrument decided to exercise its option, paying the agreed $3 million for all of Fairchild Semiconductor’s stock. For Noyce and his co-founders, it was a bittersweet moment. The Traitorous Eight each received $300,000, fully six hundred times what they had invested two years previously; the bonanza amounted to around thirty years’ salary. But at the same time, Fairchild Camera was doing even better: it was paying a price-earnings multiple of about 1.5 for a spectacular growth firm.
更为神奇的是2年后仙童半导体就实现了200w美元的税前利润，仅利润就是当时公司价值的400x了，所以虽然300w的期权价格退出实现了看似惊人的600x回报/2年时间，但也会有点遗憾。其实对当初的各方都是多赢局面了，更为难得的是，大家神奇地把资源聚到一起，团队、FA和资金，实现了新创造。
What he had done, however, was to demonstrate that liberation capital was about much more than keeping a team together in the place where its members happened to own houses. Liberation capital was about unlocking human talent. It was about sharpening incentives. It was about forging a new kind of applied science and a new commercial culture.
自由资本的神奇力量，资本的神奇力量，让科学家的天才充分大会，让各方得到充分激励，既是新的应用科学，也是新的商业文化。太有道理了。字里行间都能读出作者就此伟大发现的至高赞誉，只是相对于正批评“资本无序扩张”这种狗屁官话的国内当下，不免唏嘘。多么令人惊掉下巴的对比啊，一边从事实和历史上总结出来资本的伟大之处，在于其自由扩张；而另一边正在无知地批判这种扩张的自由。相信时间和时代最终会给出正确答案。
2 SBIC，Davis & Rock
If liberation capital launched the Traitorous Eight and Fairchild Semiconductor, the following decade brought two further advances that forged the modern venture-capital profession. First, technology investors embraced the idea of an equity-only, time-limited fund, rejecting various rival formats. Second, technology investors devised a new kind of risk management, suited to the peculiarities of venture portfolios.
Most SBICs fared worse than that. More even than the restrictions on how they invested their capital, their fatal flaw turned out to be the apparently generous terms on which they raised it. A large loan from the government sounded attractive, but the loan had to be serviced. Even with a subsidized 5 percent interest rate, this obligation had a crippling consequence: it forced SBICs to invest in startups that paid dividends. By 1966, only 3.5 percent of SBIC portfolio companies were engaged in applied science, undermining the original purpose of the SBIC program.
SBIC的结果和设立的初衷实际是事与愿违的，政策导致符合SBIC的投资必须有支付股息的能力，这简直就是开玩笑。国内的场景虽然不是完全一致，但一些地方以过度返投为条件的结果与此类似，一定会导致逆向选择，最终事与愿违。
In October 10, 1961, Davis & Rock filed its Certificate of Partnership. The two general partners—one shy and laconic, the other sunny and garrulous—rented an anonymous office on the sixteenth floor of the grand, brick-fronted Russ Building on San Francisco’s Montgomery Street. On a door at the end of a long corridor, a small sign said only “1635.” Backing low-profile startups would be a low-profile activity.Rock and his partner articulated an approach to risk management that would resonate with future venture capitalists: they promised to make concentrated bets on a dozen or so companies. Although this would entail obvious perils, these would be tolerable for two reasons. First, by buying just under half of a firm’s equity, the Davis & Rock partnership would get a seat on the board and a say in its strategy: in the absence of diversification, a venture capitalist could manage his risk by exercising a measure of control over his assets. Second, Davis and Rock insisted that they would invest only in ambitious, high-growth companies—ones whose value might jump at least tenfold in five to seven years. Venture investing was necessarily speculative, he explained, and most startups would fail; therefore, the winners would have to win big enough to make a success of the portfolio.“Trying to play it safe in small companies is, to my mind, self-defeating,” Davis insisted.Although they did not use the term, Davis and Rock were acknowledging the logic of the power law. The best way to manage risk was to embrace it fearlessly.
1961年Arther Rock的VC基金是现代历史上非常有意义的尝试：两个互补的合伙人担任GP，旧金山16楼低调的开始，以保持低调地投资低调的创业公司为目标。首创控制风险的方式就是要集中下注到12个左右的公司，持股比例不过半，但要有董事会席位、有影响力；投资目标是预期5-7年能获利10倍以上的捕鲸类型投资，胜就大胜。不做看似有安全性的小量投资，这其实是自我毁灭。控制风险的最好方式就是勇敢地拥抱风险，确保从回报看冒险是相当值得的。这一系列的思路和做法在今天看起来也非常有收获，我们在很多事情上实际迷茫了好几年，而今读起来依然是茅塞顿开，这就是历史的价值。
In the early 1960s, when the partners were laying out this vision, academia was turning finance into a quantitative science. But the way Davis and Rock saw things, the art of venture investing was necessarily subjective. Judgments about technology startups would “come from either ‘the seat of the pants’ or the ‘top of the hat,’” as Rock once wrote to Davis. Quantitative investment metrics such as the price-earnings ratio would be irrelevant, because the most promising ventures were likely to have no earnings whatever at the point when they sought capital. Likewise, they would lack the physical assets—the buildings, machines, inventory, and vehicles—that constituted “book value” at mature companies: thus another standard metric used in public markets would be meaningless. In sum, venture capitalists would have to bet on startups without the reassuring yardsticks used by other financiers. They would have no choice but to practice finance without finance.
60年代的时代背景也很有意思，当时的金融届正往量化方向发展。而早期投资这个事核心是主观判断的艺术，既没有计算P/E的财务数据基础，也没有有型资产的P/B基础，妄图使用财务指标衡量早期公司的价值是徒劳的。所以VC只能是没有金融工具的金融实践了。
Having discarded conventional investment metrics, the partners needed something else to go by. They found it in judgments about people, never mind that these could sound like a soft basis on which to commit capital. The central principle of the venture business, Davis once explained boldly, could be summed up in four words: “Back the Right People.” For his part, Rock made a habit of skipping over the financial projections in business plans and flipping to the back, where the founders’ résumés were presented. “The single most important factor in the long run for any company is, of course, management,” Rock told the Harvard Business School Club of San Francisco in 1962. “However, I believe that in the applied science industry this is especially true.” The only asset of tech startups, and the only possible reason to invest in them, was human talent, or what Rock liked to call “intellectual book value.” “If you are buying intellectual book value, then you’d better place a great deal of emphasis on the people who you hope will capitalize on their intellect,” Rock lectured.
团队是早期投资唯一重要的因素，Davis和Rock居然在60年代就把这个事情想透了，早期投资的核心就是投资正确的创始人和团队，这也是创业公司的最核心资产——团队智力资产。这就是VC投资的核心。
Rock in particular believed that his intuitions about people gave him an edge as an investor. His shy outsider’s temperament made him an expert listener, and he would meet promising company founders multiple times before committing to back them. His method was to pose open-ended questions—Whom did they admire? What mistakes had they learned from?—and then wait patiently for the entrepreneurs to fill the vacuum created by his silence. Self-contradiction, wishful thinking, a fondness for ingratiation at the expense of honesty: these were the clues that Rock should pass on an investment. Intelligent consistency, gritty realism, fiery determination: these were the signs that he should seize the opportunity. I believe so strongly in people that I think talking to the individual is much more important than finding out too much about what they want to do.
Rock绝对是个阅人的高手，而这件事正好又是早期VC的核心能力之一，需要会倾听，反复和创始人交流，通过开放性问题的探讨来识别创始人的个性，寻找稀缺的那些保持思路一致、务实、有定力的人。
On June 30, 1968, Davis and Rock wound up their partnership. Thanks overwhelmingly to SDS, but also to a defense contractor called Teledyne, their initial fund of $3.4 million was now worth almost $77 million, an extraordinary return of 22.6x; it was a performance that easily eclipsed Warren Buffett in this period, as well as that of the inventor of the “hedged fund,” Alfred Winslow Jones. Adding together their share of the fund’s appreciation plus the gain on their personal $100,000 stakes, each partner walked off with almost $10 million—or $74 million in today’s money.
Davis & Rock这期基金的运作是相当成功的，7年不到基金的整体回报达到了23倍，从340w增值到了7700w，两个创始人分别拿走了1000w，对于当初的10w投资来说增值了更是100x！按今天的汇率算，大约是1.2亿人民币的基金，实现了30亿左右的整体回报。有趣的是这期基金真正的大明星SDS，他们投入了不到26w，今天约1000w人民币，占基金规模的8%，实现了230x的回报，变成了23亿，是整个基金回报的78%。这就是幂律的神奇之处。其他组合也实现了大约5x的回报，其实是不错了，仅仅7年而已。