This year Uber and Lyft went public to mixed receptions, and WeWork tried to and could not. We will say something more direct than usual: the late-stage private market has been pricing growth without rigor, and 2019 was the year the public markets refused to ratify those prices. This is healthy.
On What WeWork Asked Us About Diligence
The WeWork S-1 was filed in August, withdrawn in September, and was followed by a recapitalization that valued the company at a fraction of its prior private mark. The substantive failure was not the document; the document merely surfaced what existed. The substantive failure was a multi-year process in which sophisticated investors funded a company at valuations that the underlying business could not support, and through governance structures that no public-market regulator would have permitted.
We had no exposure to WeWork. We do not write about it for self-congratulation. We write about it because the failure pattern is generalizable, and because the discipline that prevented us from participating is the same discipline that has cost us paper performance every year of the post-2015 vintage. The pattern, simply stated, was that capital was committed at scale to a business whose unit economics had never survived rigorous scrutiny, against a narrative that compensated for the absence of the rigor. The pattern is not unique to WeWork. We have seen versions of it in three other companies adjacent to our portfolio in 2019 alone, and we have declined to participate in all three. None of the three have yet failed publicly. We expect at least two of them to before 2022.
The harder lesson, and the more uncomfortable one, is that diligence at scale is structurally difficult. WeWork's investors were not, by any reasonable measure, lazy or unsophisticated. They had read the S-1 equivalents — the private offering memoranda, the audit packages, the operational reviews — for several rounds. They had still funded. The mechanism that overrode their diligence was, in our reading, the social-proof effect of seeing other capable investors participate. We have built explicit guardrails against this effect in our own decision process; we have not yet seen evidence that the industry as a whole has.
On the Late-Stage Premium That Has Quietly Disappeared
Late-stage rounds in the fourth quarter priced, on average, at thirty percent below the marks of the same companies' previous rounds. We participated in two such recapitalizations. The terms were materially better than the companies' prior rounds had been; the founders, in both cases, accepted the new economics because the alternative was running out of capital, and they are now operating with a cap-table discipline that will likely benefit them over a five-year horizon.
The recapitalizations were not punitive. They were corrective. The previous rounds had assumed execution that the companies had not delivered; the new rounds price the actual execution rather than the projected execution. In our experience, companies that go through this kind of correction emerge with healthier shareholder bases and more focused management teams. The cohort of companies that will benefit most from the late-2019 corrections will, we predict, produce some of the strongest 2024-2025 outcomes.
On the Companies in Our Portfolio That Should Wait
Several of our portfolio companies, this year, asked us whether they should accelerate IPO timing to take advantage of what they perceived as a closing window. We told all of them to wait. The window that closed in 2019 was not the IPO window; it was the late-stage growth-at-any-cost window. The IPO window for companies with durable unit economics has not closed and will not close. The companies that should worry are the ones that do not have durable unit economics.
The hardest of these conversations was with a portfolio company whose late-stage round had been negotiated at a price that would have required, for the IPO to clear, a stronger market than 2019 has provided. The founder believed that a 2019 IPO would crystallize the late-stage valuation; we disagreed. We argued, instead, that holding through 2020 would allow the company to reset expectations and enter public markets at a lower price but on healthier terms. The conversation was prolonged. The founder eventually agreed. We expect the resulting outcome to be materially better than the alternative would have produced.
On the Sectors That Will Matter Through the Next Cycle
The sectors we expect to produce the most consequential companies through the next cycle are, in roughly this order: applied AI, biotechnology with delivery breakthroughs, climate-related industrial technology, and certain enterprise-infrastructure categories that benefit from continued cloud penetration. Three of these are deep tech. None of them are well-suited to capital that needs to exit in five years.
The fourth — enterprise-infrastructure — is the category most under-discussed in the venture press, partly because it is less photogenic than the others, and partly because the companies that will dominate it have not yet emerged into public visibility. We have made three commitments in this category in 2019. We expect them to be among our better 2019 vintage commitments.
A Closing Note
The healthiest moments in our industry are the ones in which the marginal capital learns that it has been wrong. This was one of those years. We expect more good vintages to follow. The 2020 letter will be writing about how we deployed against the corrected pricing.
The Partners
Winzheng Family Investment Fund · December 2019