The Unicorn Puzzle

Rüdiger Fahlenbrach is Swiss Finance Institute professor at Ecole Polytechnique Fédérale de Lausanne (EPFL) College of Management. This post is based on a recent paper by Professor Fahlenbrach, Professor René M. Stulz, Daria Davydova, and Leandro Sanz.

Unicorns are private companies with pro forma valuations of at least $1 billion. In our paper “The Unicorn Puzzle,” recently posted on SSRN and available here, we investigate the puzzle of why controlling shareholders of certain startups find the unicorn status more valuable than being a public firm and the closely related question of why the number of unicorns increased so much recently. Our key findings are that unicorns differ from other VC-funded firms in that they rely more on organizational capital as well as network effects and the internet. Unicorn status enables startups to access new sources of capital. With this capital, they can invest more in organizational intangible assets with less expropriation risk than if they were public. As a result, they are more likely to capture the economies of scale that make their business model valuable.

We create a new unicorn database. Our sample consists of 639 U.S. unicorns. We have 427 active unicorns at the end of our sample period and observe 212 unicorn exits. Our sample covers all U.S. unicorns since the beginning of the 2000s until the end of the third quarter of 2021. We document the evolution of the number of unicorns and find that the number increases at an accelerating pace over our sample period. Even though 2021 has the highest number of unicorn exits, unicorn births outpace exits and the number of unicorns in existence increases in 2021. The increase is surprising because one would have expected the high valuations of 2021 to represent a unique opportunity for startups to enter public markets rather than seek to attain unicorn status.

Unicorns have reached a size that is much larger than the size of the typical IPO firm. To evaluate why founders find it valuable for their startups to stay private even though they have a much larger size than the typical IPO, we assess how the benefits and costs of being public may differ for unicorns from those of other startups. Since the unicorn phenomenon is a new phenomenon that did not exist before the 2000s, it has to be that either (1) the net benefit of being public (benefit minus cost of being public) became negative for many existing firms with private valuations of $1 billion or more, or (2) a new type of firms emerged for which the net benefit of being public is negative even at private valuations exceeding $1 billion. We show that both forces are at play. Funding has become increasingly available for firms with a valuation of at least $1 billion, which has decreased the funding and liquidity benefits of being public for these firms. In addition, a new type of firm that relies more on organizational capital and network effects has emerged. These firms are highly valuable if they succeed at capturing the benefits associated with the organizational capital and network effects that are central to their business plan, but they may not succeed in their efforts to build sufficient organizational capital and create network effects if they have to do so as public firms. These two effects constitute our proposed explanation for the unicorn phenomenon.

In the unicorn era, the funding and liquidity benefits associated with being a public firm are lower because of greater access to private funding. In the pre-unicorn world, private firms were funded by venture capitalists with a well-defined investment horizon and strict regulatory limitations on their ability to raise funds, which resulted in a limited supply of capital and pressure on startups to go public. Since then, as one study puts it, “advances in the ease of capital raising in private markets have made it possible for firms to remain private indefinitely”. We show that VC firms become less likely to participate in post-unicorn rounds, but asset managers are more likely to do so. In addition, we document that these alternative sources of finance are willing to provide liquidity to founders and employees and, therefore, also reduce the liquidity motive of being public.

The second explanation is the emergence of a new type of firm. For these firms, the cost of being public is high at high valuations because they are still developing intangible assets and are not sufficiently established yet. Disclosure costs are high for firms that invest heavily in intangible assets because the non-rivalry property of such assets makes it harder for firms to exert property rights on intangible assets. While some intangible assets such as R&D expenditures can result in patents that protect the property rights of the innovating firm, other intangible assets are much more difficult to protect, leading to high costs of public disclosure. We argue that organizational capital, broadly defined as spending on advertising, information technology, human capital, and customer relations, is more important for unicorns than R&D intangible assets.

Our database allows us to measure unicorn frequency among startups that have received at least $50 million of VC funding. If organizational capital is important, only firms with specific characteristics benefit from unicorn status. We conjecture and find that most industries do not have firms with those characteristics. Indeed, we find strong evidence of very high industry concentration in unicorns.

We show that 59% of unicorns have a business model that relies on the internet for distribution, where network and scale effects are particularly important. Further, using a new measure of the importance of network effects in a startup’s business model, we find that a firm for which network effects are important is more likely to be a unicorn. We would also expect that agglomeration externalities would be stronger for startups building intangible capital. Therefore, it is not surprising that Silicon Valley, which specializes in funding and developing technology that exploits network and scale effects, has an extremely high share of unicorns.

We next investigate the role of fund inflows in the birth and exit of unicorns. A perennial difficulty in examining the role of fund inflows on valuations is that inflows could be high because valuations are high rather than inflows causing high valuations. We obtain a source of exogenous variation in inflows using the investment plan at the inception of the Vision Fund of SoftBank. We find that exogenous variation in the supply of funding increases the number of unicorn births, consistent with a decrease in the net benefit of being public as the funding motive for an IPO becomes less important.

During the period of elevated valuations, the exit rate, defined as the number of unicorn exits in a quarter over the number of unicorns in existence at the end of the previous quarter, is high but not exceptionally so. Startups in industries with more intangibles exit later, but there is no evidence that startups in industries with higher valuations exit faster. Further, unicorns do not appear to exit faster when market valuations are high or when first-day IPO returns are high. These results are challenging for a window of opportunity explanation of the unicorn phenomenon.

Our research implies that the unicorn phenomenon is here to stay because it enables private firms to create more value than if they went public as they are building their organizational capital. These efficiency gains from building organizational capital privately could not be obtained in the absence of ample capital for private firms that have high valuations. Further research should explore the implications of decreases in available funding for startups that rely heavily on organizational capital. Our explanation for unicorns would indicate that having less funding available makes it less likely that startups relying on organizational capital will succeed in capturing the economies of scale and the network effects that make them especially valuable.

Both comments and trackbacks are currently closed.