Gordon: a couple of very raw thoughts on your VC argument.

First, VC-backed startups very rarely have significant creditors. The difficulty in servicing (presumably high-interest and low-collateral) debt is exactly what drives creditors away and gives rise to VCs with their convertibles and control rights. So, what’s exactly at stake in your creditor-protection story? Who is being protected? A tiny group of trade creditors and landlords, who rarely accumulate significant unpaid debts over time? Feels like much ado about nothing, no?

Second, the reason why VCs’ interests are aligned with those of (hypothetical would-be) creditors is not because VCs’ hold convertibles (which only establish VCs’ liquidation priority vis-à-vis the common), but because in the absence of creditors, VCs hold first priority at liquidation. Now suppose a young company has a significant group of creditors with a higher liquidation priority than that of VCs. Now, VCs’ interests are no longer closely aligned with those of creditors – regardless of the type of the security that VCs hold. As the debt/equity ratio increases, VCs with all their convertibles will start acting increasingly more like common shareholders.

To put the two together: VCs will act as natural protectors for creditors only when a company has no creditors to speak of. But when a company has serious creditors, VCs’ interests are no longer aligned with creditors’ interests, so creditors can’t rely on VCs for protection.

Am I missing something? I am happy to be persuaded otherwise.

Kate: "VC-backed startups very rarely have significant creditors."

I would be interested to see a study on this. I see a lot of startups around here with bank credit, either secured by assets (computers, lab equipment, etc.) or backed by personal guarantees of the founders or their friends. These loans are often in the high-tens or hundreds of thousands of dollars, so not insignificant to a startup.
Lenders often get warrants as part of the debt deal, so they may have interests that are aligned with the equity, though probably not in an insolvency context. I spoke to a VC who was at the conference, and he did not think it was much ado about nothing.

Kate: "Now suppose a young company has a significant group of creditors with a higher liquidation priority than that of VCs. Now, VCs’ interests are no longer closely aligned with those of creditors...."

I agree that the liquidation preference (and cumulative dividends, where they exist) make VCs look a lot like creditors where no creditors exist. You suggest that the existence of creditors would change the VCs' incentives, and my response is: maybe. Two points:

(1) You are right if the liquidation preference and cumulative dividends have no value. In that circumstance, VCs look just like common stockholders.

(2) On the other hand, recognizing that the liquidation preference is a diminishing claim in a declining company, VCs have an incentive to spot declining companies early and extract their investment. How often are they able to do that? Sometimes. And in those circumstances, they are more like creditors than common stockholders.

But here is the point: this supposed duty kicks in when the firm is in the "vicinity of insolvency." Most creditors have fairly effective contractual protections once the firm defaults, so my analysis relates to situations "in the vicinity," but not over the border. I would argue that startups are perpetually "in the vicinity" and preferred stockholders offer some constraints on the casino mentality in that context. Once they cross the border into insolvency, their incentives are as you describe.

Gordon,

The only creditors you described in the comment above are secured creditors. They by definition do not need board protection – their interests are protected by the collateral. If their papers are drawn correctly, they don’t even care whether the firm’s assets are squandered in its last days of pre-bankruptcy existence, so long as nobody (fraudulently) squanders the collateral.

So, the only relevant group of creditors for your purposes is unsecured creditors. I know of no studies on the matter, but I have never seen an empirical or theoretical paper discussing the presence of a significant group of unsecured creditors in a startup. I can’t think of a good finance reason for the use of unsecured debt. Whenever a rare unsecured creditor exists, I bet he would have an equity kicker – a warrant or some sort – which would effectively turn his interest into a convertible preferred or participating preferred.

(By the way, even secured creditors are quite rare in my experience, so I am surprised about your observations to the contrary. I’ve seen secured credit used mostly as a signaling device where founders pledge some of their own assets to show that they truly believe in their business. This arrangement acts more like a secured credit to founders, rather than to the company, and surely does not trigger your “board protection” concerns).

Now, suppose a company for some odd reasons has a meaningful group of unsecured creditors. Every new unsecured creditor reduces the value of VCs’ liquidation preference, forcing VCs to act more like the common. So, every new unsecured creditor makes the issue of board protection more meaningful for discussion, but at the same time makes your conclusion (about alignment of VCs’ and creditors’ interests) less correct.

We can think of it as a continuum: first come secured creditors (irrelevant for our purposes); then, unsecured creditors of various priorities; then, VCs; then, the common; then, holders of options for the common. At the first sign of distress, option holders are wiped out. At which point everyone moves up one notch: the common starts acting more like option holders, and VCs start acting more like the common.

Perhaps it’s worth it to model this scenario formally to see under which exact sets of conditions (debt-equity ratio, preferred-common ratio, and company performance) enough unsecured creditors would exist to matter, but VCs would still act as creditors’ protectors. My guess is, that would be a very small spot on the continuum.

A quick note on one point you made above: "In the morning session, a venture capitalist said he hoped that fiduciary duties to creditors in the vicinity of insolvency did not apply to startups, but I can't see any such limitations in the cases."

I agree that I have not seen a case that diminishes a duty itself on this ground. However, I do seem to remember that Chief Justice (then Vice Chancellor) Steel's opinion in the Francotyp case supports the proposition that the test for insolvency for startup companies may be different thatn the test for insolvency for established firms. It has been quite some time since I have rtead the case, so pardon me if memory is playing tricks on me.

I have written two papers recently providing some evidence about the kinds of creditors that VC-backed startups have. The first (85 Cornell L. Rev. 134, 153-65 (1999)), reports details from interviews with banks that specialize in loans to VC-backed software startups. That paper suggests that at least in the software industry secured bank loans are quite common, although the collateral itself provides little or no protection to the lenders, who must depend on leverage and informal relations with venture capitalists. A more recent paper (82 Wash. U.L.Q. 1375, 1429 tbl. 13 (2004)) reports descriptive data from the bankruptcy schedules of all failed VC-backed startups in recent years in the software, biotech, and telecom industries. For those firms, secured and unsecured debt is pervasive. At least 47% have bank debt. The mean unsecured debt is $32 million. {For comparison, the mean secured debt in a sample of all Chapter 11 filings is about $1000.} It is difficult to generalize about the types of debt, but it generally falls into the typical types of unsecured debt for a business of this size (trade credit of various kinds, tax claims, wage claims, and lease claims).

All of these comments are very helpful. Thank you for contributing. I am in a bit of a time crunch right now, but I would be interested to hear more thoughts on VC and creditor interests. Ron, based on your work, I wonder if you have thoughts on whether their interests are aligned or divergent. Or perhaps their interests shift from aligned to divergent as the firm approaches insolvency?

Ronald's numbers are fascinating, but they are really addressing a different question: do early-stage companies that file for Chapter 11 have debt? I am not surprised that the answer is “yes”, since if you don’t have debt, why file for Chapter 11?

But the vast majority of early-stage companies don't file for Chapter 11. Ronald's table 13, from which the striking number of $32M in mean unsecured debt comes from, has only 62 companies in it -- and, as I understand, that's the entire universe of relevant Chapter 11 filers!

Furthermore, the huge standard deviation ($92M) means that most startups that file for Chapter 11 have no unsecured debt at all (or have very small debt) and very few filers have huge debt (perhaps for bizarre reasons that have nothing to do with real unsecured borrowing – for example, because they count the “debt + warrants” position as “unsecured debt” for bankruptcy purposes). For example: if you have 62 companies, 59 of which have zero unsecured debt, and 3 have $330M in debt each, you’ll get the $31M mean and $97M standard deviation in unsecured debt.

These numbers don't seem to support the view that unsecured debt is large and common in startups. They support the view that some of those who file for Chapter 11 have large debts, and that the distribution of debts among filers is highly skewed.

Finally, I’d like to know how many “unsecured creditors” have warrants and other equity kickers, and therefore are creditors mostly in the name (for Gordon’s purposes).

Again, these numbers are very interesting; I am just not sure they are relevant to Gordon’s argument.

For your purposes, Gordon, I think their interests are aligned. The Cornell piece explains the reputational constraints that are operating in this market and those seem to provide adequate protection for the banks here. It is harder to generalize about similar constraints that might protect the other creditors, predominantly lease creditors and trade creditors. Another point is that the venture capitalists as a practical matter make a shutdown decision when they decline to continue funding. At that point, the interests of most creditors of all kinds seem to coalesce around an expeditious transfer of technology.

Kate is correct to point out that the data are highly skewed. But it is not true that most companies have little or no debt. The tables report both means and medians. The median unsecured debt, for example is about $7M (about 70 times the median unsecured debt in Chapter 11 filings overall). Based on interviews with vcs and the main lenders (SVB and Comerica), I doubt that this debt picture differs substantially from the picture for solvent startups. As for warrants, the bank lenders will have them, but the unsecured creditors most likely will not.