7 posts categorized "August 2009"

The "No Exit" Management Equity Incentive Plan

A central theme of this blog has been the way many Web 2.0 entrepreneurs and early stage investors are questioning the venture investment paradigm requiring "exits" in order to reward investors, founders and early-stage employees.

Among reasons to forgo a predetermined "exit strategy" are the following:

  • the venture may grow better organically, putting capital to work only as needed, and may blossom on a timeline that may not mesh with the liquidity agenda of a given investment fund;
  • certain investors, inclined to dis-engage from Wall Street or any investment ecosystem looking like a ponzi scheme, would just as soon have their capital returned from cash flow, and realize upside in the form of a steady annual return over the lifecycle of the venture (let's presume this could be 15-20 years); and
  • founders have agendas for their ventures as well, and these are at least as compatible with fiduciary duties to investors as pushing an exit at a time that may not be in the interests of all shareholders, preferred and common alike.

There is a new puzzle, however, for the New Internet ventures and their backers who are inventing the new investment rationales:  how best to reward early stage management who, culturally, expect (and deserve) equity-like upside if the venture succeeds? I'm referring here not so much to founders as to that class of seasoned management that tends to be recruited, still at an early stage, but after the startup has been validated (by customers, partners, revenue, or some combination of objective success other than profitability). This is a breed of management whose risk tolerance is higher than that typical of the professional executive class, who often display as much wanderlust as serial entrepreneurs.

One answer is straightforward:  if management should be incentivized to cause the company to generate distributable cash, rather than to achieve an exit, the company's board can set up a profit sharing plan. 

But here's a clever variation on the profit sharing concept that some early stage companies are adopting: a management incentive plan, structured like a change-in-control plan, that kicks in, not on a sale of the company, but on any declaration of dividends.

A typical "Management Change Of Control Incentive Plan" works like this:  immediately prior to any sale of the company, but contingent on the sale closing, a bonus pool is established, in an amount equal to a percentage of the proceeds from the sale. Sometimes, the percentage of the proceeds increases as the base amount of the sale price increases (to incentivize management to drive the shareholder proceeds higher). "Proceeds" might typically be defined as "the sum of all cash and the fair market value of any securities, other property, or any other form of consideration paid by an acquirer of the Company to all holders of the Company's capital stock for the equity, assets or business of the Company" in connection with a change in control.

A "New Internet Company Management Equity-Participation Incentive Plan" works instead as follows:  immediately prior to the declaration of any distribution, a bonus pool is established, in an amount equal to a percentage of the distribution. "Distribution" might be defined as "the sum of all cash and the fair market value of any securities, other property, or any other form of value distributed by the Company as a dividend on any class of the Company's capital stock." 

An advantage of the New Internet Company Management Equity-Participation Incentive Plan, as distinct from a simple profit-sharing plan, is that management's reward is directly tied to events that cause and justify distributions of cash to shareholders (those profits needed for reinvestment in the business are not distributed, either to shareholders or to management). Whether payments under the New Internet Company Management Equity-Participation Incentive Plan would also be made on a change in control would depend, for the most part, on whether management had stock (whether subject to options or in the form of restricted stock), and, if so, whether those interests were too far behind preferred stock liquidation preferences as to represent upside.

TheFunded's Ideal First Round Term Sheet: A Critique

Earlier this week, TheFunded Founder Institute released a model term sheet intended for venture-oriented first round financings. In his post announcing the term sheet, Adeo Ressi of TheFunded credited the blogosphere for inspiration; a recent post by Chris Dixon and others by Fred Wilson are among those supporting the idea of standard first round terms. (I'll link to TheFunded.com's post announcing the term sheet here, but their site is closed and I think their post will not be accessible unless you have paid a subscription fee.) The goals of the communal project (if I may so idealize it; there are substantive disagreements among those in the dialogue) are to (a) strike a fair balance between and among the interests and concerns of founders, companies and venture capitalists, and (b) standardize terms so that legal costs can be controlled (all agree that the legal fees typical for Series A rounds are way out of proportion to the amounts being raised).

No question, legal fees for first round financings are too high and need to come under control. Standardizing a term sheet won't hurt that cause, because, to the extent there is more overt industry consensus about what terms are fair to ask for and expect, and what "asks" are truly extraordinary or exceptional, key terms should be concluded quickly, and fewer ambiguities will have to be confronted and worked out in the process of drafting the definitive deal docs. But, simply in terms of controlling legal costs to document and close a first round deal, here's what would help more:  standardizing the stock purchase agreement.

In fact, the flavors and variations of term sheets are fairly well known. So, for that matter, are Investors' Rights Agreements and the arcane registration rights provisions and multiple indemnification provisions they contain (there is surprisingly little variation in the boilerplate language for these provisions, once you have key terms settled on). Even the better proportion of charter provisions (in Washington, we refer to a domestic corporation's charter as its "Articles," and for a corporation formed in Delaware the charter is called a "Certificate") have pedigreed boilerplate language that most securities lawyers are loathe to change for fear of unintended consequences (there are exceptions -- we could use tighter standardization of the wording of a proliferating variety of preferred class protective provisions that are a key part of the Articles or the Certificate, as the case may be).

The stock purchase agreement often ends up being the most highly customized document in the suite -- which is odd, in a way, when you consider that the document is destined to soon be backward looking. In particular, the reps and warranties from the company to its investors that a stock purchase agreement contains consume a huge proportion of the time spent preparing deal docs. Ideally, drafting the stock purchase agreement would take no more time than amending and restating the Articles or the Certificate, a document which, at least in a company going places, will be a living, governing constitution, referred to and relied upon for the life of the venture, long after the stock purchase agreement ceases to be relevant.

The NVCA has a model stock purchase agreement (available as a Word doc!), and I wonder if another effort in the spirit of TheFunded's model term sheet might be to work from this model agreement to create standard variants of the reps and warranties, with a goal of setting more overt norms about what reps are typically fair to ask for and to receive. Now, I am not talking about putting form over substance in terms of disclosure. Disclosure from a business perspective should be a matter of the credibility and integrity and candor of the players; from a technical legal perspective, disclosure should be a function of the PPM, or the business plan with some thoughtful risk factors, together with a good disclosure schedule and appropriate investor due diligence. But if the industry more readily agrees about, say, a suite of standard IP reps, maybe we will spend fewer days customizing, for example, another firm's "no open source" IP reps when the company being funded is an SaaS businesses?

At the end of the day, I suspect that in order to succeed the project is going to require a full suite of standard documents, a full suite of variant provisions, and probably also some commentary to speak to the intentions and purposes of one variant over another. All of this should be accessible and manageable online, and hopefully not balkanized among too many different law firms (I am a big fan of the Wilson and Orrick term sheet generators, but those are first generation apps and need to be improved and opened up further). I don't know if the others in the blogosphere are contemplating this, but, as I see it, the lawyer's value will be delivered as much on the front end of the deal -- the term sheet -- as in the quality control of the output of the deal documents. (In my experience, founders, companies and investors are all better served, and legal fees are lower, when lawyers come in at the term sheet stage, weigh in on how it is shaped, and take co-ownership of the instructions for the deal.)

Lonely Hucksters, Communal Digerati

I'm not at Gnomedex, taking place in Seattle today at the sleek & slender Bell Harbor Conference Center. I'm up the hill at the bigger, older, clunkier WA State Convention Center, at LSI's "Fourth Annual Advanced Conference of Current IP Licensing." So I'm missing a slate of talks by folks who (I imagine) are inspiring techies and geeks to push the social media frontier ever farther, and hanging out instead with those of my trade who are trying to make sense of how old laws are adapting to keep pace.

Jeffrey Greenbaum of New York gave us an excellent presentation this morning about how the FTC is trying to keep up with abuses in the tracking of online consumer behavior. (Earlier this year, the FTC issued suggested self-policing guidelines for online behavioral advertising.)  Greenbaum also made clear that privacy policies, tucked away by link at the bottom of a page, are not going to be conspicuous enough to adequately disclose to consumers what you may be doing with the data you collect tracking where they go and what they do. 

Greenbaum touched on an FTC action involving an individual (non-corporate) blogger who favorably reviewed a product without disclosing that the manufacturer had given him a free unit. At that point, I started to wonder:  for this kind of blunder, do we need regulation from above, or should we let a common law evolve from the very real, very powerful, if not yet culturally-normative digital ethos of personal authenticity? Policing retailers and telemarketers and financial products and all the other usual suspects from the analogous analog world, sure -- that crowd presumably needs a 21st Century version of the FTC to stay on the beat. But personal brands (twitter handles, personal blogs, facebook vanity urls) will wear the transparent baggage of their own activities, no?

Steve Tapia of Microsoft followed and observed that the idea of primacy of the individual, rather than the community, dates back only to the Romantics, that even in Western culture individualism has no deeper roots. Provocatively, Tapia said that facebook is a phenomenon where "everyone rushes home to be by themselves so that they can pretend to be with other people." (That may not be an exact quote, but it's pretty close.) All this by way of explaining the engine driving "user generated content":  the newer wave is following an ancient impulse, whereas the older generation prefers what is modern, corporate and lonely.

"User generated content" presents an unsolved puzzle for legal doctrines (copyright, defamation). In Tapia's opinion, we know little more today about the law governing user generated content than we did a decade-plus ago. User generated content creates tension -- between those with an innate sense of privacy about where they are and with whom they associate, versus those who think the time, place and substance of their daily activities are communal information -- and perhaps this tension will have to play out further before we can know the rules. Tapia told an anecdote about a convention or conference where a participant suddenly disclosed that he had been webcasting the proceedings, and everyone else felt offended. I felt slightly self-conscious about my tweeting from the event after that, though I kept right on tweeting just the same!

Here's What's Happened to the 1x Liquidation Preference

In an earlier post this year about trends in first round financings, I asked the question, "Whatever Happened to the 1x Liquidation Preference?" Based on data in a report from Cooley Godward Kornish for Q2 2009, it would appear that the 1x liquidation preference remains far and away the norm for Series A deals, and has been all year.

According to the Cooley report, 100% of all Series A deals in the survey in the first three quarters of 2008 had a 1x liquidation preference. In Q4 2008, that "fell" to 93% of Series A deals, and 7% of Series A deals in that quarter had a liquidation preference of somwhere between 1x and 2x. Was that the beginning of a trend? No, not according to the Cooley report. For Q1 2009, 100% of all Series A deals in the survey again had a 1x liquidation preference. And for Q2 2009, while there was again some movement away from 1x -- 10% of Series A deals had a greater than 1x liquidation preference (5% at between 1x and 2x, and 5% between 2x and 3x) -- 90% of surveyed Series A deals in Q2 2009 were at 1x.

In a footnote on the first page, the report states how many completed deals were surveyed for each quater reported. It does quantify the total number of deals in each quarter by round, but the next-to-last page of the report does provide a bar chart approximating how the deals breakdown by series.  If I am reading it correctly, then the number of Series A deals being reported on in the Cooley report are approximately 30, 25 and 20 for Q4 2008, Q1 2009 and Q2 2009, respectively.

Note: all of the above pertains to Series A rounds. The Cooley report does show a trending for liquidation preferences of greater than 1x in later stage financings. (In two other posts, I reported on Sequoia's 4x and 2.738x liquidation preferences in late stage Zappos financings, and analyzed the impacts of these preferences on distribution of shareholder proceeds in the sale to Amazon; but those Zappos financings pre-date the trending charted in the Cooley report.)

There is also data in the report showing that the percentage of deals with participating preferred has increased, from 55% in Q2 2008 to 60% in Q2 2009. The percentages are not further broken out by round. However, caps on participating preferred are broken out by round, although it is not clear to me how the caps are trending quarter by quarter.

Battle of the Big Law Term Sheet Generators

The Orrick law firm has now made a venture capital term sheet creator available on the web, so I thought I'd take it for a test drive and compare it to the WSGR term sheet generator which has been available for some months now. Good news:  entrepreneurs and investors now have two viable web apps to consult when planning and modeling a preferred stock financing.

Score one quick early round to the WSGR (Wilson) app, simply because it is accessible via Chrome and Firefox. Orrick's works only on Internet Explorer, which meant I couldn't access it at all from my family Mac at home.

But once I dusted off an old Dell laptop, found its powercord and remembered a three year old password, I was able to get onto the Orrick app. It's pretty good. Like the Wilson original, the Orrick term sheet creator takes you step-by-step through a series of fields, with questions, prompts and multiple choices. Inputs include information as straightforward as the company name and the names of the founders and investors, to things more esoteric such as how many S-3 demand registrations you want to offer your investors.

Orrick's app assumes you know more about what you are doing than the Wilson generator does. This is not to say, however, that the Wilson generator is more simplistic. Wilson's provides "tutorials" and links to WSGR market data (with just a few curious omissions), and represents a much better teaching tool. Where you make a choice that is outside industry norms, Wilson's generator is apt to tell you so, and nudge you back to (what it thinks) is the middle of the road. (Example of a text-box pop-up on the Wilson app when you make a questionable turn:  "It is very uncommon for preferred shares to have voting rights that differ from the common shares into which they convert.") Wilson's app comes closer to letting you in on the unwritten rules behind the typical term sheet-building process. Orrick's app is friendlier in the sense that you can get in and out of the app and generate a working draft term sheet faster.

A peeve about the Orrick app:  at the outset, it gives you a drop down menu of all 50 states, plus the District of Columbia, from which to choose the state in which your entity is organized. This might lead you to believe that your choice of state will lead to substantive differences, perhaps in references to corporate codes. But that is not the case. Although I picked Washington for the state of incorporation of my hypothetical company raising funds, the term sheet that was generated referred to my company's charter in brackets, as though it wasn't sure whether Washington corporations refer to charters as "articles" or "certificates" of incorporation. The Wilson generator limits its drop-down menu of states of organization to Delaware, California, or "other," which I've previously found a bit smug; now that design choice seems simply honest.

Wilson's app has more, but Orrick's has some things Wilson's doesn't. Orrick's app gives you a field of questions around whether you want to set up a charitable "entrepreneurs' foundation" at the closing of the round. There is no similar feature in the Wilson generator. Orrick's app also purports to have a track to generate a term sheet for a "founder's round," but on a brief tour this looked to me like an intake form (including fields for names, addresses and social security numbers) that a paralegal might use to set up a file and draft initial documents of incorporation. Not saying this is bad, just that it has more do do with corporate organization than with financing. (It complements the "Start-Up Forms Library" of legal templates that Orrick has made publicly available, concurrently with its term sheet creator app.)

One more thing, and this could be of distinctive benefit to pre-Series A stage startups in particular:  Orrick provides a separate track to generate a convertible note financing term sheet. I have not taken that track for a ride just yet.

Amazon Flows to Zappos Shareholders

Here's a reconstructed Zappos' cap table, and a model of how proceeds from a sale to Amazon might be distributed to shareholders:
MX-4501N_20090803_155905The table is imperfect. As with the analysis about Sequoia's take in my prior post, these numbers rely on (a) the disclosure in Amazon's SEC filing last week, and (b) terms of Zappos' charter on file with the California Secretary of State. Since my prior post, I've also had the chance to look briefly at the notes in the Zappos financial statements included in Amazon's SEC filing; these have numbers that in some categories are at odds with those in the summary cap information beginning at page 68 of the SEC filing and also don't track with the authorized share numbers in the Zappos charter. But the financial statements certainly helped pin down the number of shares outstanding from prior preferred stock rounds, and thus made it possible to present information in a cap table format. In any case, the aggregate numbers for the the holdings of the Sequoia entities are pretty consistent.

My initial point was that the late stage investor, Sequoia, benefitted from its liquidation preference. This certainly appears to be so, although the net "transfer" (looking at things from the point of view of equity distribution on an as-converted-to-common basis, with no preferences) from the other shareholders to Sequoia isn't quite as dramatic when you break it out class by class.

Sequoia's Preference in the Amazon / Zappos Deal

Investment entities managed by Sequoia Capital had liquidation preferences on their preferred stock investments in Zappos, and those preferences are going to make a financial difference, even in a deal contemplated to be worth hundreds of millions of dollars. Because of the liquidation preferences, the Amazon deal (assuming it closes) should work for Sequoia as though Amazon were to pay Zappos $1.1 billion in value.

Amazon's offer to purchase Zappos was widely reported last week. Of particular interest to many was the question of what Sequoia Capital would yield from the sale. The deal proposed is not for cash but for Amazon stock, so dollar values are going to fluctuate with the market for Amazon stock. But commentators I've read seem to think that Sequoia's payout may be derived by taking an eventual or assumed value of the Amazon stock offered as consideration, and multiplying that value by Sequoia's ownership interest in Zappos on an as-converted-to-common-stock basis. Zappos' Seventh Amended and Restated Articles of Incorporation (publicly available, with persistence) reveals otherwise.

Based on my review of this Zappos' charter document, together with information about Zappos' capitalization contained on pages 68-70 in a Form S-4 Registration Statement filed by Amazon, Sequoia will do better than its as-converted equity ownership percentage in Zappos would suggest. 

Given liquidation preferences of 4x and 2.738x, respectively, on investments it made in the aggregate amount of between $43,897,783 and $45,775,338, Sequoia is entitled to either $156,661,127 or $156,662,449, prior and in preference to all other Zappos preferred and common stock shareholders. One caveat and nuance:  the Amazon S-4 reveals that an investment vehicle affiliated with Michael Marks owns just under 3% of the Zappos preferred stock, and that these shares are of one, or both, of the two classes (or series) of preferred otherwise owned exclusively by Sequoia entities. The "play" in the numbers I reference in this paragraph are a result of my not knowing exactly how the Marks shares break down between the two Sequoia classes.

It's remarkable how little the aggregate Sequoia preferential payment moves, even with the ambiguity of the precise allocation of the two rounds between Sequoia and Marks. Though Zappos has six distinct preferred stock classes (or series), the first four will convert, as the equity value of the stock outstanding in each of those series is far greater than the 1x liquidation preference that the Zappos charter reveals for each.

How might Sequoia have fared, had it not had such an aggressive liquidation preference?  Assuming all Zappos preferred shares converted to common and were treated equally, and assuming a deal value of $847MM (that is a figure I saw reported by Tomio Geron in a WSJ blog; I don't care to track how Amazon's stock moves day to day, not for purposes of this exercise, but we need some number for the sake of illustration):  then Sequoia entities might have received $119,095,231 in value. In this example, then, Sequoia's actual liquidation preferences net it an additional $37.5 million.

Some other numbers, just for fun:  Sequoia would appear to be seeing a 3.42x return on its invested capital. This tracks with a multiple in the range of from 3x to 3.5x that John Cook cited last week as being reported in PEHub (I can't access the PEHub article without filing in templates I don't want to fill in), but, being a lawyer, I would stress that the actual liquidation preferences negotiated by Sequoia were, as noted above, precisely 4x and 2.738x, respectively, for the two Sequoia series. (The second multiple is a bit odd because it is designed to match the same target of $24.64 per share, in Zappos currency, that was set for the liquidation preference in the prior round; the two preferred rounds were priced at $6.16 and $9.00 per share, respectively.) The effective return of 3.42x is a function of the blend of the actual liquidation preferences and the overall amounts invested.

I am taking for granted the accuracy of the share numbers given in the Amazon S-4, and I am making other assumptions about Zappos corporate history (e.g., I am assuming the "original issue price" stated in the Zappos charter for each preferred series reflects economic reality and has not been gamed). I'd also like to thank an intrepid professional at my firm for finding the right filings with the California Secretary of State. Corporate charters are publicly available, yes, but sadly that does not yet always mean they are easily accessed.

Related Posts with Thumbnails