Unicorns’ Credit Limit
It’s conventional wisdom that late-stage venture valuations are too high.
If conventional wisdom is wrong, the increasingly popular explanation is that valuations don’t matter as much as we think. Sam Altman argued this yesterday, writing that “people are misunderstanding [late-stage] financial instruments as equity” when really “they’re much more like debt.” Such valuations, Michael Moritz previously described in the FT, are “illusory” because they’re “debt in all but name.” When looked at that way, the story goes, bubbles don’t look so soapy.
But when companies’ successive rounds are increasingly debt-like, it’s concerning, not reassuring. It’s not a good sign when companies raise them.
Billion-dollar headline valuations are illusory because they are complicated by terms in many rounds that grant privileges (convertible structures, liquidation preferences, ratchets) over each round before it. It often happens in institutional rounds, not just the last among them. This fact alone causes valuations to inflate over time because it makes it more likely that new investors will value a company more richly than old ones: they become the first to get paid in a downside scenario, and so have less to lose in bidding up the price. “Debt-like” terms don’t prevent this from happening. They cause it.
Real debt is typically inappropriate for startups. It 1) has a defined time horizon, 2) is secured by collateral, 3) pays a coupon or is issued at a discount, and 4) returns principal. Companies that traditionally have high debt-to-equity ratios raise debt because their businesses are capital intensive and asset-heavy. Tech companies, even the biggest unicorns, have unpredictable cash flows and are asset-light. Raising debt would thus be prohibitively expensive. Sure enough, “debt-like” rounds mostly haven’t involved real debt, yet.
What we ought to be concerned about, then, is when unicorns reach the point that they are forced to raise expensive debt or debt-like rounds anyway. That’s what late-stage rounds are beginning to resemble as preferences get stacked atop preferences. The terms themselves aren’t novel; their use today is. Once staples of the venture capital industry, the recent flood of VC funding shifted negotiating leverage from investors toward entrepreneurs, and onerous terms were mostly competed away. They’re fashionable again now among late-stage companies because accepting them 1) allows companies to bargain for higher headline valuations in exchange, and 2) forestalls down-rounds by allowing companies to maintain valuations they no longer deserve. Fidelity and T. Rowe Price know exactly what they’re doing even as others wail about the valuations they accept. They’re structuring their investments not to optimize for price, but for companies’ vanity.
Consider the distressed capital structure of Jawbone, which closed a $300mm loan from Blackrock this spring. As Dan Primack noted at the time, the structure was designed to allow the company to avoid repricing its last valuation. That was a magic trick worthy of mythical horses. Employees continue thinking their options are worth more than they are. Valuation-centric PR about the company’s success continues looking better than it is. VCs see their holdings marked down, but having few other options to recover value anyway, prolong saving face.
Illusory as they may be, stretched headline valuations matter a lot, because they succeed for awhile at pulling off these tricks and say a lot about the companies performing them. It is entirely appropriate for the largest unicorns to eventually transition from relying on venture equity to more traditional combinations of debt and equity as they can better evaluate the costs of each. But when they are forced to take on expensive debt, or issue what is effectively debt, simply to prop up a deteriorating headline valuation, then we ought to be very concerned indeed. About disclosure, maybe, even more than valuation.
Needless to say, adding debt (or debt-like terms) makes it more likely that early equity investors take it on the chin — regardless of whether the rounds they participated in were themselves reasonably priced. That means there’s a valuation problem not just in late-stage rounds, but one that trickles down to all rounds. It ultimately makes it that much harder for ordinary investors to find value in common shares when a company finally does go public.
Reclassifying late-stage rounds as debt doesn’t make it “hard to say where exactly the bubble is.” It makes it easy.
Thanks to Brad Holden, Jasjit Singh, and David Sack for reading drafts of this note. The views I express here are my own and not necessarily the views and opinions of LOYAL3.
 Moritz comes to the opposite conclusion of Altman, calling many unicorns “the flimsiest of edifices” despite their debt-like structuring of late-stage rounds.
 The most competitive financings are usually the simplest. Snapchat’s last round, for example, sold only common shares.
 Later investors underwrite their returns to more modest levels, which makes them less price sensitive as well.
 There are notable exceptions, including a $750mm credit facility WeWork secured this week — although it is, as startups go, a relatively asset-heavy business, with leases and buildings. Some other of the most developed late-stage companies, like Uber, also have such facilities.
 Jawbone is a hardware company, and so debt financing could be appropriate, assuming Blackrock finds collateral value in piles of unsold UP bands. Their call, not mine!