Debt and startups: A note
Debt isn’t the worst idea for startups, but it’s not as simple as just “coming”
Recently, a blog post Alex Danco called “Debt is Coming” made the rounds in VC Land, arguing that we are on the cusp of seeing more debt in the startup landscape. While this blog post blew up, the general concept has been floating around for a while, perhaps best popularized by Bryce Robert’s debt over equity movement.
Debt is typically seen as a negative signal. It often indicates a lack of availability of more prestigious capital.
On top of that, debt creates “dead weight” that must be paid off, and startups often lose money, meaning that the money will come from equity in a future fundraise. Consequently, Paul Graham has compared debt to an exploding sandwich.
So, traditionally, startups have avoided debt. This is fairly rational, seeing as debt truly isn’t good for early stage companies. But today’s growth companies aren’t exactly startups, so it’s worth asking whether this received wisdom is just another thing that today’s growth companies have inherited from a bygone era when VCs only backed startups. Let’s examine that.
Debt: a taxonomy
Let’s be clear on what debt actually is. Unlike equity, debt is a fixed item on a balance sheet that is eventually paid off and then goes away, unlike equity, which stays forever. As a result, debt is described as nondilutive. Indeed, this is often described as the main benefit of debt over equity. In fact, there are articles aplenty about avoiding venture capital due to dilution pointing to startup outcomes where the valuation was lower but the founder payoff was higher. I tend to think of debt as dilutive since there is still a liability that must be paid eventually. After all, equity and debt are dilutive only because they represent claims on cash flows, thus affecting the value of existing equity; equity just dilutes over an infinite time horizon while debt dilutes over a fixed horizon. Instead, I describe debt as dilution with a risk multiplier: it is underdilutive if the company does well and overdilutive if it doesn’t.
Like everything in life, debt comes with tradeoffs. Debt can be more onerous than preferred equity. A preference overhang is fixed and only comes due at an exit event, though in an exit the value of preferred equity can balloon. A debt is due when it’s due, it can grow over time at a compounding rate, and it is the most senior part of your preference stack, above even your investors. Plus, while there’s a lot of discourse on investors meddling too much, debt comes with a lot of restrictions too called covenants; they are often more restrictive, just less open-ended and with less nagging.
There are also many different kinds of debt, making it too simplistic to just say “debt is coming”: you must ask, “which debt?” Founders are often most familiar with convertible notes which, even though they’re treated like SAFEs, are actually a form of debt.1 The most visible kind of debt is a corporate bond — the same ones you can buy on E*TRADE — which is just a note put out to the market for a fixed period of time. But there are many other kinds of debt, including short-term commercial paper used to fund operations and revolving loan facilities, which are essentially open lines of credit that can be drawn on relatively freely. Some companies also use factored loans, where accounts receivable are the collateral. These are just a few examples. When we say “startups should start looking at debt,” startups should understand that the playing field is vast and the potential uses are many, so they will need to be relatively sophisticated when drawing on these resources.
The key thing about debt is that it is best used to fund operations — or, put another way, for cash timing instead of cash generation. So, if you have a growing business, it’s a great idea to use debt to be capital efficient, basically “pulling forward” future cash flows at an interest rate lower than your return on investment. But even restricting debt to funding operations is no panacea. Operating debt, through instruments as simple as commercial paper and rotating credit facilities and as complex as securitized debt obligations, can easily become addicting and backfire during a credit freeze. Most notoriously, General Electric used its GE Capital division to fund its operations. In fact, prior to the financial crisis, GE was the world’s biggest user of commercial paper, resulting in outstanding debt that was over 10 times its cash on hand and almost putting it out of business in literally weeks when the financial crisis hit. With startups, the risk of an event similar to a credit freeze is much more likely — all that needs to happen is for bankers to fall out of love with a vertical, for instance, or for public markets to enter correction territory during a particularly sensitive time for a company.
Should startups use debt?
Traditionally, startups haven’t had great access to debt markets, partially due to a lack of good collateral, partially due to risk aversion, and partially because debt is often a bad fit for startups that are principally concerned with growth.2 Yes, you can use debt to fund growth, but this is extremely risky because if your bet doesn’t pan out you will need to draw on other cash flows to pay back your creditor. A startup without failed growth initiatives is a failure in itself; a startup without a wildly successful experiment to pay for all the losses is, unfortunately, common, and with a fixed debt obligation would be game over. This also makes it extremely difficult to raise another round of financing because neither creditor nor investor likes to give companies money just to pay down liabilities. It therefore seems unlikely that a Brex for debt will arise, or that such a company can be successful (setting aside that the “credit” in “credit card” means debt).
However, some later stage companies have already started raising some debt and rolling it into their round, like Kabbage in 2019. In fact, just this past week, Astranis announced just such a round, and there was more debt than equity! But early stage startups will still have to manage the perception that debt is a negative signal about the company because early stage startups have only development and growth to fund, neither of which are a good fit for debt over equity.3 Debt shouldn’t be a substitute for venture capital, but it is fair to say that some later-stage companies have been using venture capital as debt. This is a mistake that should go away just as a matter of good business.
Who will issue this debt? Or, as David Haber at Goldman Sachs put it, who will be the Sequoia of debt?
The answer may well turn out to be Sequoia. The big players are already thinking of issuing debt as they go more full stack. Most notably, General Catalyst has been considering a credit fund for a while now, but presumably others have at least thought about going along this route.4 Essentially, as companies stay private longer, the need for more complex capital facilities increases, and being private increases the importance of personal relationships and signalling value, which advantages the existing Tier 1 venture capital firms (though they will likely basically act as wrappers for credit lines from bigger underwriters). This is a conflict that startups will have to navigate. But ultimately, banks that offer this kind of credit, like Goldman Sachs and JP Morgan, already invest in startups so this will just favor them in later rounds. Plus, the usual suspects like SVB, Clearbanc, and others will step in.5 Just keep in mind that bankers usually don’t care about the company’s success the same way investors do.
Lastly, Danco writes that the bankers of the future will want to securitize some of these growth stage cashflows. Securitization sounds extremely dangerous. Startups have started to dip their toes into securitization, but mostly fintech companies that are trying to securitize their own assets. This doesn’t translate to other contexts. While SaaS revenues are sometimes compared to annuities, they aren’t. SaaS revenues are contingent on a user, which is not an asset because a user can disappear tomorrow, and the revenue stream is contingent on providing a service. Furthermore, securitization depends on noncorrelation of the underlying assets and the ability to spread risk across different tranches, neither of which are really doable for startups. And outside of SaaS, most companies don’t have a similar structure that is amenable to derivative financial products.
Nondilutive capital is good
It will take time for debt to percolate through the ecosystem. Silicon Valley is a place that tries to avoid business model and financial innovation in order to focus more on product and technology innovation. You see it everywhere: startups are all vanilla Delaware C-corps selling similar software products to the same set of traditional customers and evaluated by familiar metrics.6 Avoiding debt is another way to keep things simple, both by keeping an extra party out of the mix and by taking out another thing to manage. But some complexity is there because it comes with a benefit, and like many other forms of business complexity eschewed by Silicon Valley, perhaps it’s time to retire this particular aversion.
It’s a good thing that startups are talking about nondilutive fundraising methods, but there are other methods to acquire such capital. For science startups, the obvious one is grants from agencies like DoE, ARPA-E, DARPA, NIH, and NSF, among many others which, unlike debt, are truly nondilutive. There are also competitive purses, like the X Prize, which are less sustainable than grants but can be more lucrative. They aren’t better or worse, just sometimes smarter contextually.
It’s good that founders are getting smarter about preserving their ownership percentage. But the use of debt will require a lot of thought, produce a lot of heartbreak along the way, and have a significant impact on startup financing and operations down the road.
Note: thank you to John Loeber for invaluable comments.
Convertible notes are usually thought of as the province of early stage financing, but back in the day it was mostly used by later-stage companies. Amazon famously saved the company with a convertible bond issuance in 1999. And, yes, until they convert they are technically debt. This is one of their main protections for investors.
This will probably change, though it’s likely less because of the coupling of financial capital and productive capital at this phase of the Perez cycle and more because of overabundant capital looking for yield in ever-exotic places.
The reticence surrounding debt isn’t unique to Silicon Valley; it’s just higher here. But high debt can even be a problem in public markets, so it’s incorrect to suggest that an aversion to debt is unique to tech. Warren Buffett himself recently called out overreliance on debt.
Several VC firms have recently become Exchange Act investment advisors or are considering it. But this is not the first time VCs have forayed into alternative assets. Firms like Sequoia have considered raising alternative asset funds before.
By far the most interesting experiment in this space so far is Stripe Capital.
One example of what this was Silicon Valley’s reaction to the WeWork S-1. WeWork’s corporate structure was quite complex, which caused a lot of uproar. But WeWork is a real estate company, and real estate companies have very complicated ownership structures! Even the Up-C structure isn’t uncommon for tech, having been used by companies like GoDaddy and Tradeweb.