 The United States remains the world's largest national venture capital industry, but its share of global venture capital has declined from 80% to 50% over the past 15 years as the venture capital model has been globalized. Yet U.S. data is the gold standard for evaluating venture capital because it is the only data that is not self-reported. It comes from the limited partners, the ones who put up the capital, not the general partners, the ones who invest it. Over the past 40 years, the history of the U.S. venture capital industry is still dominated by the shadow of the great tech-internet.com bubble of the late 1990s. Funds committed to American venture capitalists increased by a factor of 10 in just five years to peak in the year 2000 above $100 billion, and they've only barely reached half as much since then. The surge of funding was motivated by the outstanding returns recorded by venture capitalists in the previous decade. Now, this measure of performance, the public market equivalent, PME, has become the gold standard for all evaluation of private investing. For each fund, when capital is transferred from the limited partners to the fund, invest the same amount in the index, in the NASDAQ index, the index of your choice. When a distribution is made from the fund back to the limited partners, take the same amount from the index. Note that since the year 2000, the weighted average of venture capital returns has approximately equaled the public market. But average venture capital returns are misleading. There's enormous skew in these returns. It means that a power law distribution characterizes venture capital returns, not a normal distribution. And here, note that only the venture capital funds in the top decile of performance, that's the top 10%, outperformed the public market. This is another stylized statistical fact of venture capital in addition to the limited scope of venture capital investments. Now, there's a historical, remarkable fact here. Tom Nicholas of the Harvard Business School published in 2019, venture capital VC and American history going back through risk investing in the United States for 200 years. He shows that the skewed distribution of profit echoes with remarkable precision the distribution of profits from another domain of high risk investing, the whaling industry. My own work from a decade ago documented the skew across 205 funds from data provided by two of the leading limited partners in venture capital. Now note the extraordinary contribution, again, of the top decile, the top 10%, to the statistics for the full sample with internal rates of return reaching more than 200 plus percent. Also note the monotonic progression of returns with time from the 1980-1984 period to 1995-2006. Is this a function of venture capitalists learning how to play the game better or does it relate to evolving market conditions? Harvard data matches the data of Harris and his colleagues and Steve Kaplan and Antoinette Shore in showing correlation of venture capital returns with the public market. Note that the median terminated fund exactly matches the return generated by the NASDAQ index, again using that public market equivalent calculus. Using the top decile, the average fund is just slightly better than the NASDAQ. So it was not learning by doing that improved venture capital returns from 1980, it was the stock market. A broader sophisticated analysis confirms the correlation. First, look at the scales on each of the axes. NASDAQ is on the X, the horizontal axis, and the scale runs from minus 5% to plus 40%. venture capital, the venture capital index is on the Y, the vertical axis, and it runs from minus 20% to plus 120%. Now this equation here is mathematically quadratic because it's capturing the fatness of the tails that would be missed in a normal distribution. The R squared, the correlation is extremely close, 93%. Now third, the right-hand tail of this chart represents the years of the great internet bubble. The NASDAQ public market equivalents that exceed 25% are the venture capital venture years, 1993 to 1997. The one major outlier is vintage year 1998 when the venture return was up 33% and the NASDAQ was only up 8%. In addition to the extreme skew in venture capital returns and their correlation with the public equity market, a further stylized statistical fact is that uniquely among asset classes, venture capital returns are persistent. That means that the performance of fund one from firm A predicts the return for fund two and so on. And this persistence has persisted post 2000. Here's the data. The columns show the quartile distribution of previous fund. The rows show the quartile distribution of the next, the following fund. So almost 50%, 48.5% that were in the top 25%, that's the first quartile in the previous fund, our first quartile in the next fund. If it was random, it would be 25%, 25%. 38% of the firms that were fourth quartile, the lowest 25% in the last fund were also fourth quartile in the next fund. You wonder how they keep raising money. Virtually the identical degree of persistence can be seen post bubble since 2000 as before 2001. That's by the way not true of buyout funds, where virtually all investments, all deals are subject to public auction. Unfortunately, the LBO market has become efficient, unfortunately for investors. But the message here to limited partners is very clear. A blind allocation to venture capital, just allocating a fixed proportion to venture capital runs the major risk of what's known as adverse selection. The funds you want to invest in, the persistently successful ones, don't need your money. The ones who want your money are the ones you want to avoid. Let's consider the state of the US venture capital industry today. There has been a fundamental change over the last 10 years since the global financial crisis characterized by a flood of new investors funding private companies that have taken much longer to go public, if at all, almost 20 years now from the peak of the internet tech bubble of 1998-2000 and 10 years from the global financial crisis. Despite a huge increase in the value of exits in 2019, 80% of which were accounted for by a relatively small number of IPOs, there actually has been a very substantial decline in the number of initial public offerings, public market entries for venture capital backed ventures since 1980-2000. Then the average was 120 per year. In 2018, there were only 88 venture capital backed IPOs and only 80 in 2019. The context needs attention. We're now into the 10th year of zero real risk-free rates of interest. Plus, enormous concentration in the investment banking industry. The size of initial public offerings can be expected to rise as the number of investment banks competing for them falls and each of those investment banks becomes larger and needs a bigger deal in order to answer the telephone. In addition, an act of Congress, cutely known as the jumpstart our businesses, otherwise known as the jobs act, allowed private companies to have more shareholders without having to register with the Securities Exchange Commission. In addition, FOMO on the next fang, fear of missing out on the next Facebook, Apple, Netflix, Google, institutional investors frustrated by the lack of relatively risk-free returns have begun to play venture capitalists. The result is the unicorn bubble. Now let's start with the past two years. In 2019, as I showed in slide four, only $50 billion was raised by venture capital firms. But in 2019, $133 billion was invested in venture capital deals. Where did the $83 billion come from? Unconventional capital. This private equity, mutual funds, hedge funds, sovereign wealth funds, the National Venture Capital Association calls them tourists. They have driven a mega increase in mega rounds. In 2018 and 2019, there was a huge increase in investment rounds, the number of investment rounds greater than $100 million. About 200 such rounds per year with a total value of about $60 billion. With pre-money valuations for the late runs averaging more than $300 million, this indeed is the unicorn bubble, enabled by the unprecedented and persistent decline under non-inflationary conditions of real risk-free rates of interest. All over the world, rates have declined to where the rate realized is lower than the very low rate of inflation. The net present value of expected cash flows increases exponentially with the decline in the rate at which they are discounted. And the further out those cash flows are, the greater the increase in relative value. Companies which offer any potential for replicating the super growth trajectories of the digital giants will benefit most. As those tech leaders have come to dominate the stock market indices, the impact on investors in private companies is reinforced. But my analysis of the unicorn bubble from five years ago remains relevant. Investors are paying premium prices for unregistered illiquid securities. This is an extraordinary anomaly. There should always be a discount for illiquidity. Since there is no trading market in the securities, the investor is locked in without the implicit option when buying listed liquid shares to sell to an even more optimistic investor, also usually known as a greater fool. Now volume has grown through 2019. The volume of these deals, of these investments, these unicorn investments, as the premium has increased. So that's the signature of a bubble. It happened in the late 1990s when the price goes up instead of choking off demand increases. Now every bubble has a plausible story at birth and in fact, in one key dimension, this time is different in the fundamentals on the supply side of the venture industry. Very low cost to develop new software-based services, open source software, cloud rentals, friction-free deployment and usage, potential scale global, unbounded. This is an example of that aggressive exploration of new economic space created by technological transformation which we'll examine in lecture six. On the demand side, institutional investors are hungry for yield. As I said, we're in the 10th year of zero to negative real risk-free interest rates. The supply of IPOs remains limited, very constrained access to those that take place and that means that we've moved from a world where historically what really mattered for venture capital returns was IPO exits. Back to the work I did with Michael McKenzie some 10 years ago and I go back to this because no one else has actually directly examined the relationship between venture capital returns and the state of the IPO market. This goes behind the public market equivalent to relate venture capital fund performance to the IPO market directly, not just the overall performance of the public equity market. We mapped the quarterly cash flows of each fund, each of our 205 funds in and out with respect to the quarter in which each fund received funds and distributed funds and we gave them a cash flow weighted score. On average, through this period, there were 30 IPOs per quarter, 120 a year. The standard deviation was 10 so that meant that if there were less than 20 IPOs, we gave that quarter a 1. It was a poor IPO market. In the range of 20 to 40 around the mean, we gave it a 2, a normal market. More than 40 IPOs, well that was a 3, that's a hot market, and more than 40 IPOs with more than half of the company's going public, still not profitable, that's a bubble. The market conditions score for each quarter is defined by what the exit conditions were minus what the entry conditions were. And when we look at venture capital performance relative to the market conditions score, it's ambiguous. The median for neutral conditions actually shows higher return than for favorable conditions. But the exit conditions alone, they appear to be highly significant in driving venture capital returns. The funds that managed to distribute and sell their positions in hot IPO market time were the ones with the best return. Now recall that the number of IPOs since 2000 has fallen steeply. And acquisition by established buyers now dominates venture capital exits. In this context, providing capital to private companies, the Soft Bank Vision Fund became the poster child for the unicorn bubble. A fund of $100 billion plus managed by investment bankers with no venture capital experience. Bankers who in fact had played key roles in rendering Deutsche Bank one of the least-solving global financial institutions over the previous decade. Remarkable institution. Now, here's a genuine slide from the pitch deck for what was supposed to be the second Soft Bank Vision Fund. That was before the fiasco over WeWork, the We company, and a number of other embarrassments. It seems that Soft Bank's investment team has retreated to transactions closer to their core competence, complex option structures of the sort that they executed back in their Deutsche Bank days. But the Vision Fund is not the only venture capital fund exposed to unsustainable valuations. This is a chart that shows the returns reported by venture capital firms in aggregate relative to what has actually been distributed in securities or cash back to their limited partners. Reported returns have been inflated by marked to illiquid next round valuations, not marked to market liquid transactions. This chart quantifies the unrealized portion of reported returns. The yellow line is the value actually distributed as a percent of the original capital of the fund. The orange line is the value actually distributed as a percent of the value reported. So as of late 2019, funds raised in 2013, now seven years ago, had distributed slightly more than a third of paid in capital, but less than 20% of reported value. And the 2015 vintage, raised five years ago when now should be at a mature point, had distributed only 14% of paid in capital and less than 10% of reported value. Now 2019 was a banner year for the value of venture capital exits. There were a number of very large unicorn IPOs. And the COVID-19 pandemic promises to extend the unprecedented unicorn bubble as the Federal Reserve commits to lower for longer and a wave of unicorns has filed for initial public offerings led by Airbnb and Palantir. Now reflecting back on this phenomenon, the unicorn bubble, it's challenged what I learned during my 35 years sabbatical as a venture capitalist. My first theorem of venture capital is what I refer to as cash in control. The venture capitalist's joint hedge against the radical uncertainty of funding startups at the technological frontier. The cash means unequivocal access when something bad happens to enough cash to buy the time to find out what's going on and enough control to do something about it. Bear the chief executive officer, sell the project for what we can get, repurpose the technology, restart the venture. Venture capitalists and follow on investors today certainly have the cash, but too often they lack control. And the consequences of surrendering control to founding entrepreneurs while giving them unprecedented amounts of investment dollars has been predictable. With the unicorn startups funding their growth by selling illiquid securities to unconventional investors and extreme valuations, my second fundamental theorem also appeared to be irrelevant. Corporate happiness is positive cash flow. When customers pay more in cash than it costs to deliver the service or product, they are both demonstrating the economic value of the venture and they're liberating it from dependence on external funding. The economic impact of the pandemic has demonstrated the relevance of the cash flow constraint on growth. Now from the beginning, the unicorn bubble has been exposed in two ways. For the time being, the pandemic has ensured that the unicorn bubble is still insulated from the macro threat of rising interest rates. But the micro threat has become real on a case by case basis as business models come under scrutiny thanks to the full disclosure required by U.S. security laws, a testimonial to the wisdom of their authors some 75 years ago. While venture capital has become institutionalized as a pillar of the innovation economy and the unicorn bubble dramatizes beyond reason the potential for value creation, at least transiently, a very recent authoritative review by two leading scholars at the Harvard Business School, Josh Lerner and Ramana Nanda, identifies three strategic sources of concern. First, the disproportionate role of a few deep-pocketed investors. The top 50 investors they write, or about 5% of the total of 1,000 venture capital firms raised half of the total capital raised during these years of the unicorn bubble. Next, they highlight the very narrow sociological and geographic traits of venture capitalists themselves and they put it this way, the non-representative nature of the decision-makers at these firms is important because of the growing evidence that a lack of diversity among venture capitalists has an impact on what businesses get funded. And finally, they raise the question embedded in my first theorem of venture capital, the cash and control issue. The fundamental engine driving the venture capital industry has shifted from the narrow focus on a hot IPO market to the systemic consequences of the secular decline in real risk-free interest rates to zero or negative levels. Once capital was pulled into venture capital by the outsized returns generated relative to the public market, but that performance has not been seen since 2000 for the industry as a whole. Now capital is being pushed to VC by the meager returns available elsewhere. Even as the scale of the venture capital industry has reached new heights, approaching $500 billion assets under management, the compensation model has remained the same, 2 in 20, 2% annual management fee plus 20% share of investment profits taxed to capital gain rates. Learner and Nanda point out the perverse incentives that are thereby created. Venture capitalists have incentives to raise larger funds with lower likely returns. There is a clear correlation. The larger the fund, the more likely it is to underperform the industry as a whole and the public markets. Invest rapidly. Put the money to work in order to be able to raise another fund with less due diligence, less vetting of each deal. And choose safer investments since the compensation from the 2% on very large funds can actually approach the compensation from the returns, the share in the profits the fund generated. Collectively, these incentives produce individually rational but socially perverse results. And they may in turn, over time, serve to undermine the extraordinary role that has been played by professional venture capital in the American innovation economy.