Disrupting Venture Capital

Abstract

Global venture capital is struggling after a decade of poor industry returns. In this environment, relatively recent innovations such as disruptive modeling, OpenIPO, and revenue-based funding indicate how venture capital may continue to evolve in promising new directions for investors and entrepreneurs alike. © 2013 Wiley Periodicals, Inc.

In the last quarter of 2009, the venture capital industry's ten-year performance turned negative. No longer propped up by dot-com bubble returns, the prior decade reported cumulative trailing returns of an estimated—0.92% (a sobering fall from 25% to 30% in 2008) (Cambridge Associates, 2009; Kedrosky, 2009). Many explanations have been offered for this decline such as complaints that funds have grown too large or that exit markets are shrinking. Historically salient industries such as high-tech have been blamed for maturing and offering less growth (Kedrosky, 2009). Fresh MBAs have been accused of over infiltrating venture capital funds and proving poor substitutes for seasoned veterans (Miller, 2009). Each theory is tied to a common theme: the venture capital (VC) business model hasn't scaled well. It isn't growing efficiently.

The Venture Capital Assembly Line

To better understand the VC industry's problems, it can help to view VC as a glorified assembly line. Venture capitalists are often regarded as a mystical priesthood who practice an elusive, intuitive high art. However, like an assembly line, venture capitalists take inputs, add value, and produce outputs. They buy businesses, add capital and expertise, and then sell those businesses for a higher price (see Figure 1). Seen this way, to better help the VC business model scale, we might examine its inputs and outputs to see what's constraining the system as a whole.

Figure 1. Venture Capital Assembly Line

The Input Bottleneck

An assembly line is only as good as its inputs; if there isn't enough material coming in the front end, we can expect very little to go out the back. In the case of VC, the industry faces three primary input bottlenecks: (1) a rarity of businesses deemed “fundable,” (2) difficulty in replicating the intuition of successful general partners (GPs) to select investments, and (3) labor intensity and high transaction costs for each deal.

• 1.
  Supply of businesses. While there's no shortage of ventures looking for cash, only a small percentage are likely to grow large enough, fast enough, to warrant VC investment. Within the tiny percentage of promising deals, fewer still warrant investment because VCs can fund only businesses that are likely to be acquired or have an initial public offering (IPO) of their stock within a relatively short-term horizon (five to ten years). VCs get paid only if there is an “exit,” and exits can be hard to find.
• 2.
  Intuition of general partners. GPs tend to be seasoned executives with decades of experience in small and large firms alike. Most hold impressive academic credentials and have successfully grown their own start-ups, run billion-dollar businesses, or excelled on Wall Street. Such elite intuition is hard to mass produce.
• 3.
  Labor intensity and transaction costs. Relying primarily on the subtle intuition of GPs, there are current limits to the number of ventures a venture capitalist can diligently identify for investment in any given year—at least without compromising on quality. As a benchmark, the average number of investments per VC firm in the United States was between a mere four and six deals a year in 2007–2008. On average, there were also around two VC principals for every deal that was done (PricewaterhouseCoopers/National Venture Capital Association, 2010) (see Figure 2).

Figure 2. Deals and Manpower. Source: PriceWaterhouseCoopers/NVCA, 2010.

The Output Bottleneck

Even if input bottlenecks were cleared, venture capitalists still face output bottlenecks. Venture capitalists get paid only if their portfolio ventures are ultimately acquired or have an IPO, yet most venture capitalists can't dictate the number of exits that emerge from their portfolios, the size of those exits, or their timing. Top VC output bottlenecks include (1) lags between investments and exits, and (2) lags between invested capital and exit sums (see Figure 3).

• 1.
  Lags between investments and exits. Since marketplace exits are not directly tied to the quantity of investments in any one VC portfolio, divergence between the number of portfolio investments (inputs) and market appetites for exits (outputs) drags down overall returns. This problem intensifies when exits decline, such as following the recessions in 2001 and 2007.
• 2.
  Lags between invested capital and exit sums. Compounding the issue, not only can the number of exits create lag in a venture capitalist's portfolio, but there can also be gaps between the size of investments and their exits. Putting too much money into a venture (relative to the amount that comes out) is an obvious recipe for disaster.

Figure 3. Venture Capital Exits. Source: Thompson Reuters/NVCA (2010).

To illustrate the limiting economics of such output bottlenecks, consider a venture capitalist who strives to deliver 50% average annual returns on a seven-year investment. That venture capitalist must generate approximately 11-fold cash-on-cash returns (see Figure 4). With acquisitions accounting for 89% of exits (and the average merger-and-acquisition [M&AJ transaction totaling around $53 million in 2011) (Mendell & Herman, 2012), it defies probabilities for the venture capitalist to invest more than approximately $5 million in any one deal, regardless of the venture's uniqueness, promise, or needs. Investing more is a bet against the odds.

Figure 4. Demanding Returns

While mathematically straightforward, venture capitalists habitually overfund deals in the spirit of irrational exuberance, extreme risk tolerance, or unhealthy pressure to put large amounts of capital to work. Bigger funds want to make bigger bets. It feels more efficient to invest $10 million in 10 ventures rather than $1 million in 100 ventures. The grim reality is that beyond a certain threshold bigger bets decrease likely percentage returns.

Addressing the Input Bottleneck

How can input and output bottlenecks be alleviated in venture capital? To begin with, the solution is unlikely to come from doing “more of the same.” Something will have to change. In the case of input bottlenecks, helpful innovations must alleviate the scarcity of investable businesses, the supply of qualified GP intuition, or the labor intensity of doing deals. For example, if the right tools existed, a greater use of quantitative and statistical models might allow VCs or even less skilled individuals to more effectively separate good investments from bad.

“Disruptive modeling” is one such candidate tool. Disruptive modeling grew from “Disruption Theory” as developed by Professor Clayton Christensen of the Harvard Business School (Christensen, 2003). Rooted in Disruption Theory, disruptive modeling was pioneered by Christensen and Thomas Thurston to predict the survival or failure of businesses using empirical research and statistical analyses of how businesses in select categories of circumstance tend to fare. Disruptive modeling uses the “outside view” rather than the “inside view.”

Venture capitalists have historically favored the “inside view” for predicting venture survival. Inside views use information that is close at hand to evaluate a venture by focusing on specific attributes such as the quality of a team or technological trends. Such decisions are usually weighted in intuition and instinct rather than by explicit metrics. When asked how VCs pick a good investment, the answer is often akin to Potter Stewart's definition of obscenity: “I know it when I see it.” As such, the inside view is readily susceptible to cognitive bias, subjectivity, misplaced heuristics, and fallacious perceptions—especially in the domain of business. For example, a range of studies analyzing the predictive accuracy of expert decision makers have found manager and investor decision accuracy to rank below that of livestock judges, photo interpreters, grain inspectors, soil judges, nurses, physicians, test pilots, astronomers, weather forecasters, and many other professions (Shanteau, 2001).

In contrast, the “outside view” asks if there are similar circumstances that can provide a statistical basis for a decision. Rather than seeing every problem as unique, the outside view wants to know if others have faced similar problems before and, if so, what happened (Kahneman, 2011; Grove & Meehl, 1996; Mauboussin, 2009). The field of psychology is perhaps a worthy analogy to venture investing due to its multivariate, amorphous, dynamic nature. Psychology is divided between clinical methodologies (relying on human judgment and subjective analysis) and mechanical methodologies (relying on statistics, algorithms, and other objective tools). More than 136 studies have tested the efficacy of both methods going as far back as the early 1900s, almost invariably concluding that mechanical methods are more consistent, accurate, and yield higher-quality results (Grove & Meehl, 1996). Similarly, using historical data from 628 previous cases, professors Andrew Martin and Kevin Quinn developed an algorithm to predict US Supreme Court rulings with 75% accuracy (Martin & Quinn, 2004). Sometimes statistics outperform intuition.

Through similarly “outside view” processes, disruptive modeling is yielding compelling results. For example, disruptive modeling has demonstrated an ability to predict the survival or failure of ventures with more than 85% gross accuracy and statistical significance with over 99% confidence. Controlled experiments involving more than 300 MBA students at schools including the Harvard Business School and the Massachusetts Institute of Technology have confirmed disruptive modeling's ability to improve a population's business survival or failure predictions. It has also allowed less skilled individuals with fewer resources to begin predicting with accuracy what was historically limited to an elite few with greater skill and resources. Perhaps the elite intuition of seasoned venture capitalists may one day be mass produced after all?

To illustrate disruptive modeling, one of its categorization schemes finds that new entrants (as opposed to industry incumbents) have very low probabilities of survival when they position their offerings as having higher performance than significant competitors. In other words, if a new entrant offers better performance than its incumbent rivals, that new entrant is disproportionally likely to fail. As anecdotal examples, in the 1970s IBM was a new entrant in the office photocopier industry when it positioned its technology as superior to industry incumbent Xerox (Xerox was less than one fifth of IBM's size). In the late 1990s Entrada Networks developed high-tech storage technology that was positioned as superior to that of industry leaders such as EMC. In 2006 a start-up, SiCortex, positioned its computer servers as higher speed and more efficient than industry leaders Hewlett-Packard, IBM, and Sun Microsystems. In each case (IBM, Entrada Networks, SiCortex), new entrants were driven from their markets by established incumbents. While these are only a few examples, they highlight a pattern of circumstances that, as it turns out, hold a statistically low probability of success for start-ups overall. Once this pattern has been recognized, the lesson for venture capitalists becomes “when you see this pattern, don't invest.” As such, disruptive modeling is showing how empirical variable isolation and pattern recognition can improve VC decision making in an increasingly reliable way.

Start-ups may never be 100% predictable, but empirical tools such as disruptive modeling offer new promise in VC. Quantitative rigor and the “outside view” may improve VC decision making beyond the limits of VC's intuition-driven past. Good investments may become more effectively identified, decision tools can be more easily taught to others, and investor capital can be more effectively allocated. “Art” may evolve closer to science. Whether through one quantitative methodology or another, VC input bottlenecks may be loosening after all.

Addressing the Output Bottleneck

Regarding output bottlenecks, VCs must innovate to resolve lags between investments and exits or lags between invested capital and exit sums. While a daunting challenge, one example of a promising innovation includes the “OpenIPO” process of investment bank WR Hambrecht & Co. that enables smaller exit events. Historically, underwriting and distributing a company's shares upon IPO has been a costly, labor-intensive process. It has also been one of private backroom dealings and exclusive negotiations between ventures, their investment banks, and large institutional investors. With only a handful of large investment banks (e.g., J. P. Morgan, Goldman Sachs) underwriting the vast majority of IPOs, the relative size requirements a business must meet before being taken seriously for an IPO (around $100 million or larger) has been a daunting barrier for venture capitalists who would like to see more of their portfolio companies go public (Freeman, 2011).

In contrast, the OpenIPO process uses an online auction to set the share price for an IPO. Through an electronic “Dutch” auction, any qualified investor (not just large institutional investors) can make offers to buy shares at specified prices. The auction software (developed by Nobel Laureate William Vickrey) then uses a mathematical model to treat all qualifying bids in a more transparent, impartial way that is similar to how auction prices are set for Treasury bills. The highest bidders win the auction, and all qualified bidders pay the same price per share (which becomes the IPO price). Through this process, there can be significant cost reductions and simplification for ventures wanting to raise capital through smaller IPOs of even $20 million or $30 million. As such, OpenIPO can aid VC output bottlenecks by supplying more accessible exits. Examples of businesses that have utilized this process include Google, MorningStar, Netsuite, Peet's Coffee, Ravenswood Winery, Salon.com, and Overstock.com.

Another promising tool to address output bottlenecks is revenue-based funding (RBF). RBF is a rapidly emerging category of venture funding that allows investors to enjoy significant returns without exits. Rather than “fixing” the exit problem, RBF avoids exits altogether.

Traditional venture capitalists put cash into a business and take stock in exchange; however, an RBF investor typically puts cash into a business in exchange for 3% to 5% of the business's monthly revenue until a predetermined sum (often three to five times greater than the initial investment) is repaid over time. Examples of RBF include “royalty-based financing,” “top-line income rights,” “revenue dividends,” and other variations. RBF is best understood as a class of deal structures, rather than a specific, narrow funding architecture. For example, one RBF deal may call for 5% monthly revenue payments, another structure may involve hybrid debt, or perhaps a third uses royalties with additional equity provisions or warrants. Whatever the adaptation, RBF is defined by repayments primarily oriented around revenue, hence the term revenue-based funding.

To benefit from RBF, ventures need gross margins that are large enough for the business to thrive despite paying a percentage of its revenue to investors. There must also be covenants if the business seeks additional funding, since most businesses can't support more than one revenue-based payment at a time. However, with the right provisions in place, RBF allows investors to profitably fund a vast pool of high-potential ventures that will likely never be acquired or have an IPO. The world is brimming with profitable, rapidly growing businesses that have no likely exits. For example, many consulting, public relations, architectural, information technology (IT) service, software, retail, local manufacturing, and distribution firms with $1 to $10 million in annual revenue enjoy prolonged periods of more than 30% annual growth. In fact, over 80% of the “Inc 500” fastest growing firms in the United States between 1997 and 2007 received zero VC funding (Kedrosky, 2009). Venture capitalists historically haven't been able to invest in such businesses unless a midterm exit was foreseeable. RBF offers another way.

Experimentation with RBF as a venture funding structure began in earnest in the mid- to late 1990s. Today, examples of varied RBF funds include Cowen Healthcare Royalty Partners, DRI Capital, Paul Capital Healthcare, Revenue Capital Management, RevenueLoan LLC, Rockwater Capital, Royalty Pharma, Royal Capital Management, and The Speedy Lemur. A growing number of “traditional” VC firms and angel investors are increasingly using RBF on occasion or planning dedicated RBF funds of their own. In the health care industry alone, RBF investments in 2000 grew from 2 publicly announced deals totaling around $145 million, to 27 RBF deals totaling approximately $3.3 billion in 2007–2008 (Jung & Tamisiea, 2009). Several RBF firms have reported average annual returns of between 17% and 22%, with their initially invested capital often recouped after a mere 18 to 30 months.

RBF holds the potential to provide much needed growth cash to scores of promising businesses that, but for RBF, would have little to no access to growth capital beyond what can be secured from friends and family or through onerous loans by risk-averse banks. Ventures can often tax-deduct a large portion of their RBF repayments; such payments entirely depend on revenue (no revenue, no payment); and entrepreneurs do not have to give up precious ownership of their businesses in exchange for funding. By preserving an entrepreneur's ownership, control, and upside benefits from the venture's success, there are many circumstances where entrepreneurs consider RBF highly alternative. At the very least, for many promising businesses, RBF is vastly superior to “nothing,” which is what they might get from a traditional venture capitalist today.

Conclusion

The VC industry is not doomed, broken, or otherwise destined for failure. Rather, like any business model, it has been limited by critical input and output bottlenecks that have constrained its ability to more effectively scale, adapt, and thrive. At a time when start-ups and economies need investment capital more than perhaps any other period in living memory, innovations such as disruptive modeling, OpenIPO, and revenue-based funding suggest that venture capitalists may yet reach new heights of ingenuity and success. Through these and other innovations, the best may be yet to come for entrepreneurs and venture investors alike.

Biographical Information

Thomas Thurston is CEO of Growth Science. He is also a board member of the Revenue Capital Association, a member of the Forum for Growth and Innovation at the Harvard Business School, and a member of the Oregon State Bar Association (044359). A Thunderbird alumnus, he holds a BA, MBA, and JD, and was honored as a research fellow at the Harvard Business School. He has been a guest lecturer at Harvard and the Massachusetts Institute of Technology.