Carrots and Sticks: How VCs Induce Entrepreneurial Teams to Sell Startups

Venture capitalists (VCs) usually exit their investments in a startup via a trade sale. But the entrepreneurial team – the startup’s founder, other executives, and common shareholders – may resist a trade sale. Such resistance is likely to be particularly intense when the sale price is low relative to VCs’ liquidation preferences. Using a hand-collected dataset of Silicon Valley firms, we investigate how VCs overcome such resistance. We find, in our sample, that VCs give bribes (carrots) to the entrepreneurial team in 45% of trade sales; in these sales, carrots total an average of 9% of deal value. The overt use of coercive tools (sticks) occurs, but only rarely. Our study sheds light on important but underexplored aspects of corporate governance in VC-backed startups and the venture capital ecosystem.

a

INTRODUCTION
Venture capitalists (VCs) play a significant role in the financing of high-risk, technology-based business ventures. Many of America's best-known public companies began as VC-backed firms: Amazon.com, Apple, FedEx, Intel, Microsoft, and Starbucks, to name a few. 1 Not surprisingly, venture capital is considered to be an important contributor to economic growth in the United States and elsewhere. 2 The steps in the venture capital cycle are by now familiar. 3 A venture capital firm creates and raises capital for a limited-life fund. 4 The VCs select portfolio companies for inclusion in the fund, investing in these companies through multiple rounds of financing. 5 Along the way, the VCs advise and monitor the portfolio companies, sometimes replacing the companies' founders. 6 Before the end of the fund's life, VCs "exit" from their investments in the portfolio companies and return capital to the fund's investors. 7 The fund's investors can then recycle the returned capital into another venture capital fund.
Venture exits usually take one of three forms: (1) an initial public offering (IPO) of the portfolio company's shares followed by the sale of the VC's shares into the public market; (2) a "trade sale" of the company to another firm; or (3) the dissolution and liquidation of the company. 8 Of these three types of exits, IPOs have received the most scrutiny. 9 This attention is not surprising. IPO exits tend to involve the largest and most visible VC-backed firms.' 0 And, perhaps just as important, the IPO process triggers public-disclosure requirements under the securities laws, making data on IPO exits easily accessible to researchers.
But trade sales are actually much more common than IPOs and, in the aggregate, are likely to be almost as financially important to VCs."1 Indeed, in certain industries-like medical devices-and during certain periods-like the last decade, when the IPO market was tepid-trade sales are likely to be more important to VCs than IPOs.12 Unlike IPOs, however, trade sales do not trigger the securities laws' intense public-disclosure requirements; they instead take place in the shadows.' 3 Thus, although trade sales are important to the venture capital cycle, researchers know relatively little about them.

8
Other forms of VC exit include VCs selling their interests in the startup to one or more private equity firms, and one VC selling its interest to another VC (a "secondary"). See Darian M. Ibrahim, The New Exit in Venture Capital, 65 VAND. L. REv. 1, 16-17, 19-20 (2012) (discussing how VCs are increasingly becoming sellers on the secondary market and how buyers on the secondary market include both VCs and private equity funds). These forms of exit merely replace some or all of the firm's existing VCs with new investors; they do not substantially alter the relationship between the startup's investors and its entrepreneurial team. Thus, these forms of exit do not create the types of conflicts explored in this paper-those that can arise in sales of the portfolio company to another operating company. Less commonly, a VC may exit by having the portfolio company redeem the VC's shares.  If the acquirer is publicly held and the acquisition is deemed a "material" agreement, then the acquirer must disclose in a Form 8-K "a brief description of the terms and conditions of the agreement. . . that are material to the company." SEC, Additional Form 8-K Disclosure Requirements and Acceleration of Filing Date, Exchange Act Release Nos. [Vol. 98:1319

CARROTS AND STICKS
The purpose of this Article is to shine light on how VCs arrange to sell startups in trade sales. In particular, we seek to investigate how VCs induce the "entrepreneurial team"-the founder, other executives, and common shareholders-to go along with a trade sale that they might have an incentive to resist.
We begin by describing the standard cash-flow and control rights that VCs receive when investing in startups. Turning first to cash-flow rights, we explain that VCs almost always invest through convertible preferred stock. In a trade-sale exit, VCs choose between retaining their preferred shares (and capturing most or all of the proceeds through their liquidation preferences) or converting the preferred shares into common shares. In an IPO, VCs (as a practical matter) must convert to common stock. 1 4 Turning next to control rights, we explain that VCs seek board seats and shareholder voting rights, in part to make it easier for them to exit and realize their cash-flow rights.
We then explain why founders, executives, and common shareholders are more likely to oppose a trade-sale exit than an IPO exit. In an IPO, the founder (if still the CEO) can continue running the firm and will face less direct oversight as shareholdings become more diffuse.' 5 Other executives of an IPO firm can typically keep their jobs, also with less shareholder oversight than before. And the original common shareholders of the IPO firm generally do well; VCs would not push for an IPO exit unless the common shares-to which the VCs must convert in an IPO-have considerable value.
By contrast, in a trade sale, the founder (if still the CEO) and other executives may lose their jobs or find themselves subject to more direct oversight as shareholdings become concentrated in the hands of a single shareholder (the acquirer). In other words, they become "mere" employees. And because VCs in trade sales often exit as preferred shareholders with liquidation preferences that must be paid in full before common shareholders receive any payout, common 33-8400, 34-49424 (Aug. 23, 2004), available at http://www.sec.gov/rules/final/33-8400. htm#seciic. Form 8-K disclosure only applies, however, to material transactions, and the acquirer does not need to file the actual merger agreement or provide other details about the target firm. See id.
14 In an IPO, VCs typically convert their preferred stock into common, either because it is contractually required or because preferred shares almost never survive an IPO given market resistance. Thus, VCs are likely to push for an IPO only when the common stock they would receive upon converting their shares is worth more than the preferred stock's liquidation preferences (plus, where relevant, the preferred stock's participation rights).

15
Black & Gilson, supra note 9, at 260-61 (noting that an IPO returns effective control of the firm to the founder-CEO).

CORAELL LAW REVEW
shareholders may receive little (if any) payout. 16 At the same time, the sale eliminates any "option value" (upside potential) of the common stock held by the founder, other executives, and employees. 17 For all these reasons, VCs pushing for a trade sale may face resistance from the entrepreneurial team, particularly when common shareholders receive very little.
Two points are worth emphasizing here. First, an entrepreneurial team resisting a trade sale may not necessarily believe that the startup has a reasonable chance of going public if it remains independent. Rather, team members may resist a trade sale today because they believe that, if the startup remains independent, there is a good chance that it can exit in the future via a more attractive trade sale-one that provides significantly better employment opportunities for the entrepreneurial team and more value for the common shareholders.18 Second, there may be situations where the entrepreneurial team favors a trade sale opposed by the VCs. For example, the entrepreneurial team might wish to accept an offer that is personally lucrative but which provides only a modest return to the VCs.1 9 Thus, we do not claim that the entrepreneurial team will always oppose a trade sale. Rather, our claim is that there are likely many scenarios where VCs will favor a trade sale that the entrepreneurial team opposes.
After describing the entrepreneurial team's potential incentive to resist a trade sale, we then discuss the various strategies the team might use to impede the sale. The executives and common shareholders can impede a trade sale through their influence over the board, which must approve the transaction. The common shareholders can try to block the sale by exercising their voting rights or by threatening litigation. Finally, the executives can refuse to cooperate in the sale process or, if the acquisition requires their continued participation in the enterprise, refuse to commit to such participation.
16 Of course, if the trade sale is at a high price relative to total liquidation preferences, common shareholders may well receive a large payout even if the preferred shareholders do not convert to common.

17
For a discussion of the option value of common stock in VC-backed firms, see infra Part II.B.L 18 VCs, on the other hand, may want to sell the firm now and capture most of the sale price through their liquidation preferences. First, the VCs' liquidation preferences cause the VCs to bear most or all of the downside risk associated with keeping the startup independent. SeeJesse M. Fried & Mira Ganor, Agency Costs of Venture Capitalist Control in Startups, 81 N.Y.U. L. REv. 967, 978-79 (2006). Second, VCs may wish to sell the firm now to show prospective investors for the VCs' next fund that the VCs can successfully exit their investments. Third, VCs may wish to exit now because the fund in which the portfolio company is held is coming to the end of its life.
Next, we identify the various bribes (carrots) and coercive tools (sticks) that VCs can use to induce a reluctant entrepreneurial team to support, or at least not impede, a trade sale. For the founder and other executives, the VCs can offer bonuses for a successful sale (carrots). The VCs can also threaten termination or blacklisting if the executives do not cooperate (sticks).
For common shareholders, VCs can offer to share part of the VCs' cash-flow rights with the common shareholders through a "carve-out to common" (carrot). They can also use vote buying or other transactions that dilute common shareholders' voting rights to undermine common shareholders' ability to block a transaction via their voting rights (sticks).
To investigate the use of carrots and sticks in trade sales, we use a hand-collected database of 50 VC-backed Silicon Valley firms sold to acquirers in 2003 and 2004. The firms are primarily in the biotech, telecommunications, software, and internet sectors. The average sale price is $55 million, but there is considerable variance in outcomes. A number of sales are essentially liquidations, while other sales are for well over $100 million; one firm sold for over $500 million. For each firm, we collect data on the carrots and sticks used in connection with the sale. 20 We find in our sample a heavy reliance on carrots. To induce executives to cooperate in selling their firms, VCs frequently offer sale bonuses. In 16 of the 50 firms, VCs pay an average sale bonus of $1.63 million. In 11 of 50 firms, VCs give common shareholders as a class an average of $3.7 million extra. In 45% of the firms, VCs give at least one type of carrot, with these carrots on average amounting to 9% of the deal's value. Across all 50 firms in our sample, an average of 4.3% of deal value-2.4% on a dollar-weighted basis-is used to fund these two types of carrots: sale bonuses and carve-outs to common. 2 1 We also find some use of sticks, such as threats to blacklist founders who refuse to cooperate and attempts to undermine common shareholders' voting rights. But the overt use of these sticks is rela-20 This paper builds on earlier work using portions of the same data set. See Brian Broughman &Jesse Fried, Renegotiation of Cash Flow Rights in the Sale of VC-Backed Firns, 95J. FIN. ECON. 384, 385-86 (2010) (showing that common shareholders' ability to extract carve-outs from VCs depends on the allocation of control rights between common shareholders and preferred shareholders). In this Article, we use additional information gathered in our study to provide a more complete picture of VCs' efforts to induce entrepreneurial teams to sell firms by describing (1) the carrots given to both common shareholders and executives and (2) VCs' use of various sticks, such as blacklisting threats, vote buying, and vote dilution.
21 The use of carrots is greater in the 42 firms in our sample where the VC-investors do not convert to common stock. In such firms, there is a greater conflict between the entrepreneurial team and VCs. In this subsample of 42 firms, we find that VCs use an average of 5.0% of deal value (3.1% on a dollar-weighted basis) to fund carrots. tively infrequent. And although 60% of the founders in our firms are replaced before the sale, we find no evidence that VCs fired or threatened to fire founders to ease the sales of firms.
Our study makes three important contributions to the literature. First, it sheds light on an important but underexplored aspect of corporate governance in private, VC-backed firms. In particular, it highlights the potentially divergent interests of different players in VCbacked firms around trade-sale exits and shows that VCs frequently must overcome potential opposition to these sales.
Second, our study provides additional evidence that managers of a target firm can extract value by holding up a sale of the firm. Researchers have reported on the use of side payments to executives in the sale of publicly traded firms. 2 2 We show that such "bribes" are used not only in public firms with dispersed ownership but also in closely held private firms.
Third, our study provides some evidence on whether VCs are constrained from abusing their power in startups. The relatively infrequent use of sticks in our sample supports the view that reputational or other nonlegal considerations substantially constrain (but do not completely prevent) sharp-elbowed behavior by VCs in Silicon Valley.23 The remainder of the Article is structured as follows: Part I describes the typical structure of VC cash-flow and exit-facilitating control rights in startup firms. Part II explains why founders, other executives, and common shareholders may wish to impede a trade sale as well as the means they have to do so. Part III describes the carrots and sticks that VCs may use to overcome such resistance. Part IV presents our dataset. Part V describes the carrots and sticks actually used by VCs in our sample.

A. VCs' Cash-Flow Rights
In the United States, VC-backed startups almost always issue two classes of stock: common and convertible preferred. 24  and employees of the startup hold the common stock. 25 The convertible preferred stock is held by the VCs, who invest in startups almost exclusively through this type of security. 26 Most venture-backed startups issue a new series of preferred stock for each round of financing. 27 Like most preferred stock, VCs' preferred shares carry a liquidation preference and are convertible into common stock. 28 If VCs retain their preferred stock and preserve their liquidation preferences, those preferences must be satisfied before common shareholders receive any payment. In this scenario, the VCs' cash-flow rights resemble debt. 2 9 If the VCs convert their preferred shares to common, giving up their liquidation preferences, the VCs have the same cash-flow rights as common shareholders. 3 0 Thus, convertible preferred stock combines downside protection with upside potential.
In a trade sale, the VCs will choose whether to exit as preferred shareholders and reap the benefits of their liquidation preferences or 27 While some of the rights of the preferred shareholders may be class rights, each series of preferred stock is assigned its own exclusive rights and preferences, including liquidation preferences that may be senior or junior to those of other series. This, in turn, may give rise to conflicts within the preferred class over exits and other transactions. Indeed, in certain situations, the interests of the most junior preferred shareholders may be closer to those of the common shareholders than to those of the most senior preferred shareholders. In this Article, we implicitly assume that all of a firm's VCs share the same interests (vis-a-vis common shareholders, the founders, and other executives) around the potential sale of the firm.
28 See Kaplan & Stromberg, supra note 24, at 285 tbl.1 panel G (finding that almost all VC financings in the United States involve securities senior to common).
29 Unlike the liquidation preferences of most public-company preferred stock, VCs' liquidation preferences can easily exceed the original purchase price of the stock: the liquidation preference of VCs' preferred stock sometimes confers the right to be paid a multiple of the purchase price before common shareholders may receive any payment. See, e.g.,  VCs' convertible preferred stock may have "participation rights" that entitle holders not only to a liquidation preference but also to share (with common shareholders, on a pro rata basis) in any additional value available for distribution to shareholders, usually up to a specified amount (say, three times the original investment amount). Thus, the VCs will convert their preferred shares into common stock only if the amount they would receive as common shareholders exceeds the sum of their liquidation preference and the value of their participation rights. See Brian J. Broughman to exit as common shareholders. If the VCs exit via an IPO, underwriters will typically insist that the VCs convert their preferred shares to common shares and give up their liquidation preferences along with other rights attached to the preferred stock. To prevent holdouts, a startup's charter will, with respect to the holders of each series of preferred stock, generally either (a) require all of the holders of that series to convert to common stock or (b) allow a majority of the holders of that series to forcibly convert all holders of that series to common stock. 31 Thus, one way or the other, VCs in IPO firms will typically exit as common shareholders.

B. VCs' Exit-Facilitating Control Rights
To exit their investment, VCs must arrange an IPO, trade sale, or dissolution of the firm. Below, we describe the two types of control rights most relevant to facilitating an exit: (1) the right to fill seats on the board of directors and (2) shareholder voting rights. 3 2

Board Seats
The board of directors is responsible for managing the day-to-day business and affairs of the company. 3 3 Critically, the board is also the only corporate body that can initiate fundamental transactions, such as mergers, IPOs, and dissolutions. 3 4 These transactions, in turn, are almost always necessary for the VCs to exit; hence, we will call them "VC exit transactions." Thus, VC exit transactions-including trade sales-cannot go forward without board approval. But, as we will see, board approval alone is not sufficient. Shareholder approval is also necessary.

32
VCs also negotiate for numerous protective provisions that enable them to block a variety of corporate actions. See Fried & Ganor, supra note 18, at 987. By tightly constraining the firm, these provisions give significant leverage to VCs and make it harder for founders and other common shareholders to resist the VCs' demands to sell the firm. Staged financing-the ability to withhold needed cash-also gives VCs substantial influence over corporate decision-making and increases their ability to sell the firm. See id. The board, in turn, is elected-and can be replaced-by the shareholders. 3 5 In the typical public company, default statutory voting rules give the holder of each share one vote for each open board seat. 3 6 By contrast, in a VC-backed firm, the parties typically negotiate a voting agreement that overrides default statutory voting rules. 37 The agreement specifies the allocation of board seats among: (1) representatives of the common shareholders; (2) representatives of the VC investors; and (3) so-called "independent" directors, who are mutually appointed by the common shareholders and the VCs, and typically are industry experts or other outsiders with valuable experience and connections. 3 8 Because board approval is necessary for VC exits, VCs negotiate aggressively for board seats when investing in startups. 3 9 While data on board composition in private, VC-backed firms are scarce, two persistent patterns emerge from various small-scale studies of VC contracting documents and VC-backed firms. First, VC board representation tends to increase with new financing rounds, especially outside rounds that bring in new VCs. 40 Second, in a plurality of firms, neither common shareholders nor VC-appointed directors ever achieve outright control; the swing vote remains in the hands of an independent director. 4 1 Keeping the swing vote in the hands of an independent director makes it more difficult for either the common shareholders or the preferred shareholders to act opportunistically with respect to the other class of shareholders. 42 But as one of us has  (reporting that by the time of the IPO, the median number of VC-directors is three, the median number of management directors is two, and the median number of outside directors is two); Kaplan & Stramberg, supra note 24, at 287-88 (finding that in a survey of 118 startups, neither VCs nor common shareholders had control in 60% of the startups).

42
See generally Broughman, supra note 30, at 484 (showing that independent directors serving as tiebreakers can limit opportunistic conduct that may occur if entrepreneurs or investors were to control the board); Brian Broughman, Independent Directors and Shared argued in other work, these studies (and the outwardly neutral appearance of independent directors) may understate the extent of de facto VC control over startup boards. 4 3

Shareholder Rights
As noted above, for VCs seeking an exit transaction, board approval is not sufficient. Shareholder approval is also necessary.
As a matter of corporate law, some shareholder approval is always required. The nature of that approval will depend on the corporate law of the state in which the firm is domiciled as well as the firm's charter and other organizational documents. In some states, such as Delaware, the default arrangement is that a majority of preferred and common shares (voting together) must approve exit transactions. 4 4 But alternative arrangements, including provisions giving particular series of preferred stock veto rights, are common. 4 5 In other states, including California, a majority of each class of shares (including common stock) must approve exit transactions. 4 6 As a matter of business practice, the buyer in a trade sale may also insist on shareholder approvals not required by corporate law or the firm's organizational documents. For example, a buyer might be concerned about appraisal rights. Under corporate law, dissenting shareholders in a merger (those voting against the transaction) can exercise appraisal rights and have a court determine the "fair value" of their shares; the buyer is then forced to pay fair value to these dissenters. 4 7 If the court's determination of fair value is much higher than the amount paid per share in the transaction, the cost of acquiring the target could rise substantially. To minimize potential appraisal claims, an acquirer may insist that a supermajority of shareholders ap-Board Control in Venture Finance, 9 REv. L. & ECON. 41 (2013) (providing a formal model of decision making under shared board control with an independent director holding the tiebreaking seat). 43 See Fried & Ganor, supra note 18, at 988-89 (noting that (1) at least one director designated as a representative of the common shareholders will be the CEO, and if the CEO is an executive hired by the VCs instead of the founder, he or she may be loyal to the VCs rather than to the common shareholders; and (2) "independent" directors may not be truly independent of the VCs-even when the directors are approved by both the common shareholders and VCs-if they are drawn from the VCs' professional network and expect to interact with the VCs in the future). prove the transaction. 4 8 The bottom line is that corporate law may provide only a floor for shareholder-approval requirements.
Because common shareholders' approval is sometimes needed as a matter of corporate law (in states where approval by each class of stock is required) or business practice, VCs can (and sometimes do) negotiate for "drag-along" rights that allow them to force common shareholders to vote for any transaction favored by the VCs. 49 If the common shareholders have agreed to vote their shares as directed by the VCs, they cannot block an exit transaction via a class vote. 50 Instead, they must seek other means to stop the transaction.

POTENTIAL ENTREPRENEURIAL-TEAM RESISTANCE TO TRADE SALES
This Part explores why the entrepreneurial team may resist a trade sale that the VCs propose. Subpart A explains why founders and other executives may oppose, and may be able to impede, a trade sale. Subpart B explains why common shareholders may oppose, and may be able to impede, a trade sale.

A. Resistance by the Founder and Executives
The founder and other executives may have both reasons and the means to oppose a trade sale.

Incentive
Consider a founder who is still the CEO. The founder-CEO may have ambitious (perhaps grandiose) financial and nonfinancial aspirations for the startup. He or she may believe that these aspirations can be realized only if the startup remains independent and he or she 48 Historically, another reason for seeking a high percentage of target shareholder votes was to obtain favorable accounting treatment for the transaction. See, e.g., Orban v. Field, No. 12820, 1997 Del. Ch. LEXIS 48, at *17 (Del. Ch. Apr. 1, 1997) (the acquirer of a VC-backed startup insisted that 90% of the common shareholders approve the acquisition in order to obtain pooling-of-interests accounting treatment).
49 See Robert P. Bartlett, III, Venture Capital, Agency Costs, and the False Dichotomy of the Corporation, 54 UCLA L. REV. 37, 62 n.98 (2006) (describing drag-alongs). Depending on its terms, the drag-along may not be effective unless the board has also voted to approve the sale.

50
To the extent that common shareholders have agreed to vote their shares as directed by the VCs, and the shares are voted in favor of a transaction, the common shareholders may lose their right to appraisal, which is generally available only to shareholders who vote against the transaction. See, e.g., DEL. CODE ANN. tit. remains in control. A trade sale may well snuff out these aspirations. To the extent the founder-CEO wants to keep the startup independent so that it can realize its full "potential," he or she may oppose a trade sale.
Whether or not the founder is still part of the management team, the firm's executives might have two very practical reasons to oppose a trade sale. First, the executives may well be replaced or subject to tighter supervision as the firm comes under the control of a single shareholder (the acquirer). The acquirer could thus force the executives to give up the pay, perquisites, and prestige associated with their positions-or to work harder for them. 5 1 Second, executives may own substantial amounts of common stock in the firm, especially if they are part of the founding team. As we discuss below, a trade sale may yield little for common shareholders while eliminating the option value of their stock. 5 2 To the extent executives own common stock that they will be forced to give up for less than its option value, they have another reason to try to block a trade sale.

Ability
Executives of VC-backed firms seeking to block a trade sale have three potential means to do so. First, and most importantly, they can refuse to cooperate in the sale of the firm. They can refuse to report acquisition interest to the VCs, they can drag their feet in dealing with a potential acquirer, and they can refuse to provide information or emphasize negative aspects of the business to a potential acquirer. Moreover, if a potential acquirer wants members of the executive team to stay on after the acquisition (for example, to continue developing a startup's technology), executives can block the deal by refusing to work for the acquirer. In short, executives can use their positional power in the firm to undermine and sabotage efforts to sell it.
Second, to the extent executives have seats on the board, they can threaten to vote against a sale and try to persuade independent directors to oppose the transaction. Attorneys advising a board may prefer a vote to be unanimous because it reduces litigation risk. The

51
Executives of VC-backed startups that face potential trade sales are in the same position as executives of publicly traded companies that face either takeover attempts or merger proposals. See generally Hartzell, Ofek & Yermack, supra note 22, at 38 (discussing how executives at target companies experience high turnover rates at the time of acquisition and during the years immediately proceeding acquisition). Of course, a trade sale might make the executive team better off in some cases. In those cases, executives would not oppose the sale.

52
See infra Part II.B.1. desire for a unanimous vote would increase the ability of even a single director to impede a sale.
Third, if common-shareholder approval is necessary for the deal and executives own a large block of common shares, they can threaten to vote their shares against the deal and lobby other common shareholders to oppose the deal. This brings us to common-shareholder opposition to trade sales.

B. Opposition by Common Shareholders
Common shareholders, like executives, may also have an incentive (and some ability) to impede a trade sale.

Incentive
VCs exiting via trade sale can always choose to retain their preferred shares rather than convert them into common shares. In this scenario, the VCs have liquidation preferences that must be paid in full before common shareholders receive any payout. Thus, common shareholders may not receive much, if anything, in a trade sale. However, even if the VCs' liquidation preferences exceed the sale price, the common stock might have considerable option value at the time of sale: if the firm is not sold now, it might later hit a home run and be sold later for a price that far exceeds the preferred shareholders' liquidation preferences, giving a large payout to common shareholders. 53 The less common shareholders receive and the greater the loss of option value, the more common shareholders will wish to block a trade sale.
In blocking a sale, the common shareholders' goal may not be to keep the startup independent so that it can later go public. Rather, these shareholders may resist a trade sale today because they believe that, if the startup remains independent, there is a good chance that the future will bring a more attractive trade sale-one that provides significantly more value for common shareholders.

53
Consider, for example, a startup with $50 million in total liquidation preferences. Assume there is a 50% likelihood that, within one year, the firm will be worth $90 million and a 50% likelihood that it will be worth $0. A hypothetical risk-neutral buyer content to earn a 0% return would pay $45 million for all of the equity of the startup. Preferred shareholders would get $45 million; common shareholders would get $0. But if the startup were to remain independent, the common stock would have an expected value of $20 million because there is a 50% likelihood of a $90 million sale yielding $40 million for common shareholders. This $20 million is the option value of the common stock that is lost in the sale of the firm today for $45 million.

Ability
Common shareholders can block a trade sale in three ways: by (a) using their shareholder voting rights; (b) litigating or threatening litigation; or (c) using their influence over the board.

a. Shareholder Voting Rights
As we explained in Part I.B.2, common-shareholder approval may be necessary to conduct a trade sale. In particular, such approval will be necessary if (1) corporate law or the acquirer requires commonshareholder approval, and (2) the VCs fail to negotiate effective dragalong rights enabling them to force the common shareholders to vote for a VC-favored sale. If neither corporate law nor the acquirer demands common-shareholder approval (or the VCs have effective dragalong rights), then common shareholders cannot threaten to hold up the deal through their voting rights.

b. Fiduciary-Duty and Other Litigation
To try to block a deal, common shareholders can also threaten to sue the board or the VCs directly for, among other things, breach of fiduciary duties under corporate law. Common shareholders' litigation leverage will depend in part on the corporate laws of the state in which the firm is domiciled. The easier it is to demonstrate a breach of fiduciary duty under that state's corporate law, the more leverage common shareholders will have.
But there are two limitations to common shareholders' ability to credibly threaten litigation. First, aggrieved common shareholders will often lack the financial ability to sue VCs. 5 4 Second, there may well be large reputational costs to litigating against VCs. 55 All in all, the actual likelihood of litigation will probably be low.

c. Board Influence
A third mechanism by which common shareholders can impede a trade sale is by asking their representatives on the board (if there are any) and any independent directors (if there are any) to vote against 54 See Fried & Ganor, supra note 18, at 1000-01. Lawyers generally will not take such cases on a contingency basis because the amounts involved in any given case are likely to be relatively small. The plaintiffs would thus need to finance the potentially extensive litigation themselves. Defendants, whose litigation will be financed by the target's directors and officers insurance policies, can be expected to engage in a scorched-earth defense to financially exhaust the common shareholder plaintiffs. Members of the entrepreneurial team, even collectively, are unlikely to have the resources necessary to finance such litigation.

55
See id. at 1001. The founders and employees holding common stock may wish to raise money from VCs in the future for other ventures, or work at other VC-backed startups. Acquiring a reputation as a "troublemaker" who sues VCs is likely to make it more difficult to raise funds from VCs or to get positions at VC-backed firms in the future. the transaction. Common shareholders will have considerable influence on how the board votes if a majority of the board's seats are held by a combination of common-shareholder representatives and independent directors. Indeed, given directors' preference for unanimous board votes, even a stubborn minority of directors who care about the interests of common shareholders might be able to block a sale that is unfavorable to common shareholders.

VCs' POTENTIL CARROTS AND STICKS
In Part II, we saw that founders, other executives, and common shareholders of VC-backed startups may have the incentive and ability to resist a trade sale. In this Part, we describe the carrots and sticks that VCs can use to induce founders and other executives (subpart A) and common shareholders (subpart B) to go along with trade sales. 56 Subpart C explains that, from the VCs' perspective, both carrots and sticks entail costs.

A. Founders and Executives
To induce founders and other executives to facilitate rather than impede a trade sale, VCs may either offer carrots or deploy various sticks.

Carrots
For founders and other executives, the VCs can use direct carrots (management bonuses) or indirect carrots (carve-outs to common shareholders).

a. Management Bonus
The direct carrot is a "management bonus" triggered by the sale and generally tied to the sale price. The bonus may be structured in a variety of ways. To begin, executives can be promised ex ante, even before an acquirer is identified, a portion of the sale proceeds. Alternatively, executives can be given a contemporaneously negotiated payment. This ex post payment might be labeled as a "bonus" or, 56 Because VCs provide financing in stages and have the ability to prevent the firm from getting debt or equity financing elsewhere, they can always wield the "implicit" stick of starving the company of cash until management and common shareholders agree to sell the company. However, use of such a stick might also reduce the value of the VCs' investment in the firm. In addition, if members of the entrepreneurial team expect to get little in the sale, the threat of destroying the firm may not induce them to support the sale. For these two reasons, there are limits to the power of this stick.

CORNELL LAW REVEW
alternatively, characterized as consideration for a noncompete or consulting agreement. 57 The cost to VCs of using a management-bonus carrot will depend on the sale price relative to the aggregate liquidation preferences at the time of sale. If the sale price exceeds the aggregate liquidation preferences, the original common shareholders will pay for much (or all) of the management-bonus carrot. If the sale price is less than the aggregate liquidation preferences, the VCs will shoulder all of the cost of the carrot.

b. Carve-Out to Common
The indirect carrot is a carve-out to the common shareholders as a class. A carve-out entitles common shareholders to a portion of the proceeds from the trade sale even if the firm is sold for less than the VCs' liquidation preferences. To the extent that the founders and executives own shares of common stock, they will share pro rata in the carve-out. For example, if founders and executives own 50% of the common stock, a carve-out to common of $10 million will yield founders and executives $5 million.

Sticks
VCs facing recalcitrant founders and other executives have two potential sticks at their disposal.

a. Termination
First, VCs could threaten to fire the founder and other executives if they refuse to cooperate in a trade sale. Termination would cut off future compensation and perhaps cause the forfeiture of unvested equity. Such a step would be quite painful for the executives.
But such a tactic would also be quite costly for the VCs. The startup would lose its management team, which may be crucial to a potential buyer; the VCs would be forced to spend time and energy finding a replacement; and the new team would then need to be brought up to speed and adequately incentivized to assist in selling the firm. As a result, we would expect VCs to wield this stick rarely, if at all.

b. Blacklisting
Second, VCs could threaten to blacklist founders and other executives who refuse to cooperate in a trade sale. VCs are repeat players in both funding entrepreneurs and hiring professional executives for 57 A retention agreement with the acquiring firm may also function as a carrot if its actual purpose is to overcome executives' resistance to the sale rather than to increase the value of the acquirer by binding valuable human capital to the acquiring firm. [Vol. 98:1319 CARROTS AND STICKS startups. Neither entrepreneurs nor professional managers will want to displease the VCs, all else being equal.

B. Common Shareholders
To induce common shareholders to support a trade sale, the VCs have one carrot at their disposal and a number of sticks.

Carrot: Carve-Out to Common
The carrot VCs can offer common shareholders is a carve-out from their liquidation preferences (or other cash-flow rights) so that common shareholders receive more than they are contractually entitled to receive. Such a carve-out to common can both induce common shareholders to vote for the sale and cause them to refrain from suing.

Sticks
VCs can wield a number of sticks to compel the class of common shareholders to vote for the trade sale. However, these sticks cannot prevent a lawsuit and may in fact increase the likelihood of one.

a. Cross-Voting
To eliminate common shareholders' ability to block a trade sale, VCs can either partially convert their shares into common stock or exercise warrants to buy common stock so as to acquire a sufficient amount of common stock to achieve the necessary approval threshold. Consider the transactions in Orban v. Field, 5 8 a Delaware case involving a common-shareholder lawsuit against a VC-controlled board. Office Mart, a Delaware corporation, arranged to be acquired by Staples in a merger that provided no payout to its common shareholders. 5 9 The charter allowed preferred shareholders to vote alongside common shareholders on an as-converted basis (as if their stock had been converted into common shares), and the preferred shareholders had enough votes to ensure shareholder approval of the transaction as required by corporate law. 6 0 However, for accounting reasons, Staples insisted that at least 90% of Office Mart's common shares be voted in favor of the transaction. 6 1 Common shareholders, led by Office Mart's founder and former CEO George Orban, refused to back the deal, demanding $4 million in exchange for their votes. 6 2 58 No. 12820, 1997 Del. Ch. LEXIS 48 (Del. Ch. Apr. 1, 1997). 59 See id. at *2.
Office Mart's board refused to provide a carve-out to common. 63 Instead, the board arranged a series of transactions that, in essence, assisted the VCs in converting a portion of their preferred stock into common, thereby diluting the original common position down to less than 10% of the class. 64 The VCs, now holding over 90% of the common stock, voted their common stock in favor of the merger, which allowed the transaction to go forward and wiped out the original common shareholders. The maneuver cost the VCs nothing; although the partial conversion into common reduced the VCs' aggregate liquidation preferences, it still left them with enough liquidation preferences to absorb all of the merger consideration. Orban sued, arguing that Office Mart's board had violated its fiduciary duties to common shareholders. The Delaware Chancery ruled in favor of the board, concluding that the "common stockholders had no legal right to a portion of the merger consideration under Delaware law or the corporate charter" 65 and noting that "[t] here is no claim" that the challenged transaction "was not in the best interests of the corporation." 6 6

b. Vote Buying
VCs can also overcome common-shareholder opposition by buying the votes of shareholders, such as executives, who already own a large block of common stock. For instance, assume that three executives of a startup collectively hold 55% of the common stock and that a trade sale cannot occur without approval by 50% of the common shareholders. Rather than providing a carve-out to all common shareholders, the VCs could simply award each of these three individuals a lucrative cash bonus or other payment in connection with a trade sale in exchange (implicitly) for voting their common shares in favor of the transaction. Vote buying via payments to executives can thus kill two birds-executive opposition to the trade sale and common-shareholder opposition to the trade sale-with one stone.
Alternatively, a VC-controlled board can issue (additional) common shares to certain executives or employees who expect to gain from the board's proposed transaction and will thus vote their common shares in favor of the transaction. 6 7 Although vote buying in 63 Id. at *21. 64 Id. at *22-24. 65 Id. at *32. Of course, the issue here was not, as the court implies, how to divide the proceeds of an already-effected merger. In fact, the merger had not yet taken place. Rather, the issue was whether a preferred-controlled board could use corporate resources to dilute the voting power of common shareholders objecting to a proposed merger that benefitted the preferred shareholders and left the common shareholders with nothing. Id. at *2-3.
66 Id. at *8 n.23. 67 For example, the plaintiffs bringing suit in Kalashian v. Advent VI L.P., No. CV-739278, 1996 WL 33399950 (Cal. App. Dep't Super. Ct. Oct. 4, 1996) alleged-among some or all of these cases may violate the board's duty to refrain from improperly tampering with the shareholder vote, 6 8 it is very difficult to prove that the purpose of a particular payment or equity issuance to executives or employees is to buy votes.

C. Carrots vs. Sticks
We have seen that VCs have at their disposal both carrots and sticks to induce entrepreneurial teams to cooperate in trade sales. Carrots tend to involve out-of-pocket costs. For example, carve-outs to common come solely at the VCs' expense. 69 But many sticks, such as threats to terminate or blacklist a founder or other executives, do not involve any cash outlay. 70 However, sticks can give rise to substantial indirect costs. First, although litigation is unlikely in any given case, the use of sticks may well increase the risk of litigation, which, should it occur, can be very costly in terms of VCs' time and reputation. 7 ' Second, VCs may be deterred from "misbehaving" because of reputational consideraother things-that the VCs had given common stock to new management solely to undermine the voting power of the original common shareholders for the purpose of preventing these shareholders from blocking an economically dilutive financing. After trial, the VCs

69
Management bonuses would come solely at the expense of the VCs if the firm were sold for an amount less than total liquidation preferences. However, if the firm were sold for an amount greater than total liquidation preferences, management bonuses would be subsidized by the common shareholders, either partially (if VCs have participating preferred stock) or completely (if VCs have nonparticipating preferred stock).
70 Vote buying and cross-voting may or may not require dipping into the VCs' pockets. First, consider vote buying. Vote buying can occur completely at the expense of the common shareholders if value the original common shareholders would otherwise (absent the vote buying) have received is used to pay for the votes. For example, suppose that absent vote buying, common shareholders as a class would receive $5 million in a sale and a majority of the common shareholders would oppose such a sale. The VCs then pay executives a net amount of $4 million to exercise underwater options to acquire (on a postexercise basis) 50% of the outstanding common shares and vote their shares in favor of the transaction. Common shareholders as a class now receive $1 million, with 50% of that $1 million going to the executives. Next, consider cross-voting. Cross-voting can often occur with no out-of-pocket cost to VCs. Suppose, for example, that VCs' liquidation preferences are $50 million and the sale price is $40 million. The VCs can convert 20% of their shares into common stock and vote these common shares in favor of the sale without reducing the value of their liquidation preferences and their payout in the sale.

71
See tions. 72 As websites like TheFunded.com make the world smaller and more transparent, and serial entrepreneurs play a more important role in startups, the reputational costs to VCs who bully entrepreneurial teams with sticks are likely to become higher. Thus, both carrots and sticks can be costly to VCs.

A. Sample Population
We study the use of carrots and sticks in trade sales of VC-backed startups using a hand-collected data set of VC-backed Silicon Valley firms. This section describes the data-collection process and provides descriptive statistics for the firms in our sample.

Data Gathering
The data used for this project were collected in 2005-2007 as part of a study supported by the Kaufmann Foundation and the University of California, Berkeley. We obtained from VentureReporter.net a list of VC-financed companies located in California that were sold in trade sales in either 2003 or 2004. We filtered out all firms except those located in and around San Francisco, San Jose, and Oakland (broadly defined as "Silicon Valley"), 7 3 leaving a population of 193 firms.
For each firm, we next sought to locate and obtain data from one or more persons knowledgeable about the firm's life-including the circumstances surrounding its sale. We identified the current business addresses for the founders and executives-all of whom we call "entrepreneurs" for convenience-of 141 of the 193 companies. We mailed letters asking entrepreneurs from each firm to provide us with data, promising to keep confidential the identity of the entrepreneur and the startup firm.
Entrepreneurs from 57 of the 141 firms agreed to provide us with data-a response rate of 40.4%. The information obtained, supplemented by publicly filed corporate charters, covered each firm's entire life span. Among the data gathered were the state of 72 See Black & Gilson, supra note 9, at 262-63 (arguing that geographical proximity between VCs and investment portfolio companies gives rise to reputational constraints, deterring opportunistic acts by VCs); Sahlman, supra note 4, at 513 (arguing that VCs refrain from abusing their power because they wish to attract the best entrepreneurs, who could always obtain funding from other VCs or alternative sources of capital).
73 We used LinkSV to filter out firms that did not meet these criteria. See LINKSV, http://www.linksv.com (last visited Feb. 4, 2013). LinkSV profiles all companies located in Silicon Valley (in or around San Jose, San Francisco, and Oakland) that receive venture capital funding. Companies that did not appear on LinkSV were removed from our sample. incorporation; cash-flow rights and control rights negotiated in each VC financing round; the identities and backgrounds of the CEO and directors; and the circumstances and terms of the sale, including amounts paid to management and various classes of shareholders.
From the original set of 57 firms, we removed seven for lack of adequate data, leaving us with 50 firms. In most of these sales (42 out of 50), the VCs exited as preferred shareholders. In the remaining eight firms, the VCs converted into common stock in connection with the sale, giving up their liquidation preferences.

Selection Issues
Because our sample is limited to Silicon Valley firms sold in 2003 or 2004, one must be somewhat cautious about extrapolating from our firms to VC-backed firms generally. Silicon Valley is a close-knit community with its own norms and ways of doing business. 74 Our sample firms were sold several years after the tech bubble collapsed, a period in which VCs lost considerable amounts of money. 7 5 These losses may have heightened the conflicts between VCs and entrepreneurial teams around exit events. The use of carrots and sticks within our sample firms could thus be a function of the post-bubble time period and factors unique to Silicon Valley, limiting the generalizability of our results. . 76 In addition, our sample consists only of companies whose entrepreneurs voluntarily responded to our request for information. There could be systematic differences between firms whose entrepreneurs responded to our inquiries and firms whose entrepreneurs did not. While we sought to minimize such biases by soliciting data from every entrepreneur we could locate and by offering confidentiality, our sample might not even be completely representative of Silicon Valley firms sold in 2003 and 2004. One reason to think that our sample might be unrepresentative, for example, is the higher incidence of litigation in sample firms compared to other firms. Among the 50 sample firms, there were three lawsuits filed by founders against VC investors, a litigation rate of approximately 6%. (Two of these lawsuits related to the sale of the firm; one did not. For confidentiality reasons, we cannot disclose any additional information about these lawsuits.) Among the 143 out-of-sample firms, there was only one founder who filed a lawsuit against VC investors, a litigation rate of less than 1%. This suggests that the frequency of sticks outside our sample might be lower than the frequency of sticks in our sample.

B. Sample Description
Our sample firms are "high-tech" businesses, primarily in the biotech, software, telecommunications, and internet sectors (Panel A of Table 1). The concentration of IT-related businesses is representative of VC-financed firms generally. 7 7 At the time of sale, the firms had received an average of $42 million in VC funding and had been operating for an average of approximately five years. The mean sale price is $55 million. Panel B of Table 1 provides information on the amount invested, financing rounds, years of operation, and sale price.

C. VCs' Ex Ante Cash and Control Rights
We now turn to describe, for each of the firms in our sample, (1) the VCs' cash-flow rights; (2) the entrepreneurial team's board seats and shareholder voting rights; and (3) the identity of the CEO at the time of sale.

VCs' Cash-Fow Rights
Across all 50 firms, VCs' aggregate liquidation preferences at the time of sale are, on average, $47 million. In the first round of financing, the liquidation preference usually equals the amount invested (a "lx preference"), while the liquidation preference in subsequent rounds is more likely to be a higher multiple (i.e., 2x or 3x) of the amount invested (Panel C).78 At sale, aggregate liquidation preferences are on average somewhat greater than the amount invested (Panel B).
When VCs retain their preferred stock rather than converting to common stock, the allocation of the sale proceeds depends on the relationship between liquidation preferences and the sale price. If liquidation preferences exceed the sale price and contractual priority is fully respected, common shareholders get nothing. Liquidation preferences exceed the sale price in 31 of the 42 firms in which VCs exit as preferred shareholders (Panel D). Absent a carrot, common shareholders as a class have little incentive to approve a merger when their equity is underwater. In eight firms, it was in the VCs' interest to convert to common stock rather than maintain their liquidation preferences, meaning that the sale proceeds were allocated pro rata among 78 Liquidation preferences from early rounds of financing are sometimes waived or otherwise reduced in a subsequent round of financing (a "recap" financing). A recap financing may occur as part of a voluntary recapitalization of the firm, perhaps to eliminate "debt" overhang from relatively large liquidation preferences, or alternatively, a pay-to-play contractual provision may force a VC to convert to common stock (and thereby give up its preferences) if it fails to participate in a subsequent financing round. In our sample of 50 firms, there were 13 recap financing rounds (Panel C). all common shareholders (the original common shareholders and the converting VCs).

Power of the Entrepreneurial Team
This section describes the extent of the entrepreneurial teams' blocking power-their board seats and corporate-law rights-in our sample firms. The data are summarized in Table 2.

a. Board Seats
Recall that board approval is required for a trade sale. Thus, the entrepreneurial team may be able to impede a sale through its influence on the board. We divide directors into three categories: (1) VC appointed, (2) common-shareholder appointed, and (3) independent (outside). If a particular outside director had been selected exclusively by either the VCs or the common shareholders, we designate this person as a VC or common-shareholder director, regardless of how contracting documents label the board seat. 7 9 Panel A reports the allocation of board seats. At the time of sale, 56.5% of all directors were appointed by the VCs, and 22.8% were appointed by the common shareholders. Panel B shows that the VCs controlled the board in 29 of the 50 firms (58%). In our sample, common shareholders rarely controlled the board at the time of the sale (3 of the 50 firms). However, in 21 of the 50 firms, the combination of independent directors and common shareholders could have blocked a sale. Panel A reports the mean and median board representation for (1) common shareholders, (2) VCs, and (3) independent (outside) directors. Panel B shows board control at the time of sale. If the VCs (or common shareholders) control a majority of the board seats, we classify the VCs (or common shareholders) as having "Control." If the board has an even number of seats and the VCs (or common shareholders) appoint exactly half the directors, we treat this as "Blocking." "Shared Control" means that the VCs and the common shareholder each appoint fewer than 50% of the directors, with outside directors providing the tie-breaking vote. Panel C shows the state of incorporation at the time of sale.

79
Our "de facto" classification of directors differs from the "formal" classification used by Kaplan and Stramberg in their 2003 study, which treats any board seat intended for a director who is not a VC or a representative of the common shareholders as held by an outside director. Kaplan & Str6mberg, supra note 24, at 287. As one of us has explained elsewhere, nominally independent directors may not be truly independent tiebreakers between the VCs and the common shareholders. See Fried & Ganor, supra note 18, at 988-89.

Mean
Median SD

b. Shareholder Rights
All our companies were incorporated in either California or Delaware at the time of their sales, consistent with findings that most VCbacked firms incorporate either in their home state or in Delaware. 8 0 Panel C shows that 35 out of 50 firms were incorporated in Delaware at the time of the sale. As we explain below, California law may give the entrepreneurial team somewhat more power vis-i-vis VCs through both voting rights and the threat of fiduciary litigation.
(1) Voting rights. California and Delaware provide different voting rights for shareholders. In Delaware, mergers typically need to be approved by only the holders of a majority of all of the firm's outstanding stock, both preferred and common. 8 1 We find that at the time of sale, VCs almost always have sufficient voting power to dictate the outcome of a shareholder-wide vote. 82 California, on the other hand, requires a separate vote for each class of shareholders. 8 3 Thus, when the VCs remain preferred shareholders, common shareholders of firms domiciled in California can more easily impede a sale they oppose.
But it is not that simple. California purports to subject "quasi-California" corporations (corporations doing business in California but incorporated elsewhere) to the requirement of separate class [Vol. 98:1319 votes. 4 While California's legal ability to impose this requirement on firms incorporated elsewhere is contested, many (but not all) Delaware-incorporated companies located in California are advised by lawyers to hold separate class votes. In our sample, all but one Delawareincorporated firm held a separate class vote for the common shareholders.
Thus, as a practical matter, incorporation in California rather than in Delaware may not give common shareholders much more power through voting rights. Nevertheless, common shareholders' ability to impede a transaction is still likely to be somewhat greater in a California-incorporated firm, where a separate class vote is indisputably mandatory. 8 5 (2) Fiduciary litigation. During the years in which our sample firms were sold, Delaware law appeared to permit a VC-controlled board to make decisions that favor preferred shareholders at the expense of the common shareholders unless the common shareholders could show that the decision was not in the "best interests of the corporation." 8 6 In contrast, California law generally affords stronger protection to minority shareholders, including common shareholders in firms with VC-controlled boards. 8 7 Thus, directors of Californiadomiciled firms may have believed they faced greater risk of liability for harming common shareholders. 8 8 84 See id. § 2115(b).
85 As noted earlier, if the common shareholders have subjected themselves to drag-along rights, they cannot use their voting rights to block a sale. Three entrepreneurs in our sample (two California-domiciled firms, one Delaware-domiciled firm) indicated that drag-along rights limited their ability to vote against a trade sale. We have been told by venture capital lawyers that drag-along rights are now used much more frequently, especially in deals involving East Coast venture capital firms.
86 See Orban v. Field, No. 12820, 1997 Del. Ch. LEXIS 48, at *29-32 (Del. Ch. Apr. 1,   1997). However, a recent decision by Vice Chancellor Laster clarifies that a board controlled by preferred shareholders must do more than act in the "best interests of the corporation"; in particular, directors must "pursue the best interests of the corporation and its common stockholders..  464, 471 (Cal. 1969) (holding that majority shareholders have a fiduciary responsibility to minority shareholders to use their ability to control the corporation in a fair, just, and equitable manner); DiLillo v. Ustman Techs., Inc., No. B148198, 2001Cal. App. Unpub. LEXIS 1527(Nov. 19, 2001) (determining that a VC was a controlling shareholder of a firm and that plaintiff common shareholders could sue the VC directly for breach of a fiduciary duty in connection with the sale of the firm's assets).
88 However, a quirky feature of California corporate law may make it difficult for common shareholders to prevail against VCs on corporate-law claims when the VCs exit via a merger. Under California law, a firm's insiders have been permitted to eliminate personal liability for violations of corporate law by conducting a merger with an unrelated entity; once the firm has merged, shareholders' only remedy for corporate law violations, even for alleged fiduciary violations that occurred prior to merger, is appraisal. See CAL. CORP. CODE § 1312 (West 2012); Steinberg v. Amplica, Inc., 729 P.2d 683, 693-94 (Cal. 1986).

Founders' Position in the Firm
The entrepreneurial team is likely to have a greater incentive to block a trade sale if the founder is still the CEO. First, a founder-CEO is likely to own more common stock than a replacement CEO; everything else equal, he or she will thus face a greater loss of option value if the firm is sold. Second, a founder-CEO may derive greater nonfinancial benefits from the business and thus may have additional incentive to keep the firm independent.
The entrepreneurial team may also have greater ability to impede a sale if the founder is still the CEO. A founder-CEO is likely to have substantially more common stock than a hired CEO, giving him more shareholder-blocking power. Compared to a hired CEO, the founder-CEO is also likely to have longer and deeper relationships with other directors and common shareholders. As a result, the founder-CEO is likely to have influence over directors and other common shareholders, increasing the likelihood that directors will resist the trade sale and further increasing the likelihood that common shareholders will vote against the transaction.
To reduce resistance to a trade sale, a firm's VC investors may use their control over the board of directors to replace the CEO before the planned exit. In our 50-firm sample, 29 of the founders (58%) were replaced as CEOs before the firm was sold. However, as we will discuss shortly, none of these replacements appeared to be made to overcome resistance to a trade sale.

V ACTUAL CARROTS AND STICKS IN THE SALE OF VC-BACKED STARTUPS
Part III described the potential carrots and sticks that VCs could deploy to induce entrepreneurial teams to sell the firm. Here, we describe the actual carrots (subpart A) and sticks (subpart B) used in our sample. We also describe the total out-of-pocket cost to the VCs in our sample of using carrots in connection with trade sales.
Delaware law, which is generally considered to be more insider friendly, does not shield insiders in this manner during ordinary mergers. See DEL. CODE ANN. tit. 8, § 262 (2011); Rabkin v. Philip A. Hunt Chem. Corp., 498 A.2d 1099, 1105(Del. 1985) (holding that courts should examine allegations of "specific acts of fraud, misrepresentation, or other items of misconduct" in connection with a merger even if appraisal is available to plaintiffs as a remedy). Thus, if a California court today were to follow Steinberg, common-shareholder plaintiffs would be worse off, in some respects, than if the firm had been subject to Delaware corporate law.

Common Carve-Outs
In our sample, 11 sales (22% of 50 sales) feature carve-outs to common shareholders-extra value above the amount to which they were contractually entitled. To measure these carve-outs, we compare the actual payout received by VCs to the VCs' contractual entitlement. When VCs convert their preferred shares to common shares, their contractual entitlement equals their pro rata share of the sale price. When VCs exit as preferred shareholders, their contractual entitlement is the lesser of their liquidation preferences and the sale price. 89 Any excess payment received by the original common shareholders is treated as a carve-out to common. Table 3 describes the carve-outs to common shareholders in our sample. Limited to the 11 sales in which a carve-out occurs, the average carve-out to common shareholders is $3.7 million, or approximately 10% of the total sale price. Though most firms do not provide a carve-out, carve-outs can be substantial when they occur. In our sample, one sale provided a $10 million carve-out to common shareholders, and in two deals, the carve-out was at least 25% of the total sale price.
Carve-outs to common shareholders occur only in the 42 firms where the VCs exit holding preferred stock. Limited to the subsample of 42 firms where VC investors did not convert their preferred shares to common shares, we find that common shareholders receive, on average, $969,000 more than their contractual entitlement, or approximately 2.5% of the total sale price. Among all 50 companies, common shareholders receive, on average, $810,000 more than their contractual entitlement, or approximately 2.2% of the total sale price.
Carve-outs to common shareholders are illustrated in Figure 1. Each bar represents a firm's total sale price (in millions of dollars). The sale price is divided into four components: (1) the amount actually paid to VCs (in black); (2) the contractual entitlement of the original common shareholders at the given sale price (in white); (3) the carve-out given to the original common shareholders (in dark grey); and (4) nonretention bonuses paid to senior management in connection with the sale (in light grey).

89
The sale price is defined as the amounts paid to VCs and common shareholders. In those cases where (a) the sale price exceeds the liquidation preferences and (b) the VCs exit holding participating preferred stock, we define VCs' cash-flow rights as the sum of the liquidation preferences and the participation rights.

FIGURE 1.
This figure shows, for each firm, the distribution of exit proceeds among VCs, common shareholders, and senior management. Each bar represents a firm's total sale price (in millions of dollars). The sale price is divided into four components: (1) the amount actually paid to VCs (in black); (2) the contractual entitlement of the original common shareholders (in white); (3) the carve-out given to the original common shareholders (in dark grey), and (4) non-retention bonuses paid to senior management in connection with the sale (in light grey). For ease of presentation, all firms with a sale price of more than $100 million are normalized to a sale price of $100 million.  In our sample, common carve-outs are used, at least in part, to overcome common shareholders' opposition to trade sales. Elsewhere, we showed econometrically that these carve-outs are more likely to occur when: (1) VCs lack de facto board control and thus need to convince other directors-such as independent directors and common-appointed directors-to support the sale; (2) the firm is domiciled in California rather than Delaware, and therefore common shareholders' rights are likely to be slightly stronger; or (3) common shareholders would otherwise be wiped out and receive none of the sale proceeds. 9 0 Qualitative interview data confirm that carve-outs to common shareholders are designed to induce common shareholders and their allies on the board to support the sale. In one California-domiciled firm, for example, the VCs carving out a portion of their liquidation preferences for common shareholders required each common shareholder to sign a liability waiver before receiving a portion of the carveout. According to the entrepreneur, the carve-out was offered only because the VCs were concerned about a possible common-shareholder suit challenging the terms of the sale. In another case, where the VCs lacked board control, the entrepreneur told us that the VCs were forced to give a carve-out payment to common shareholders to obtain the support of other directors.

Distribution of Sale
Of course, some of the carve-outs to common might have had other motives. VCs might have wanted to be "fair" to common shareholders or to build or maintain a reputation as "friendly" VCs. However, the fact that these carve-outs are more likely to occur when common shareholders have more power or more incentive to block a sale suggests that they are being used at least in part to induce common shareholders' cooperation with sales.

(Nonretention) Management Bonuses
Bonuses that do not require management to stay with the acquirer might be used as carrots to induce executives to support a sale. 9 1 Of the 50 firms in our sample, 16 offered various nonretention management bonuses in connection with the acquisition, with an average bonus in these 16 firms of $1.63 million (or 6.6% of the sale price). Bonuses are sometimes given to a broad class of employees (5 firms), but in most cases they are limited to key executives (11 firms).
Among all 50 companies, senior executives receive an average nonretention bonus of approximately $520,000. Across all companies, these payments represent approximately 2.1% of the total purchase price (or 0.95% on a dollar-weighted basis).*92 Compared to bonuses awarded in connection with the sale of publicly held firms, the payments we document are modest in dollar terms, but large as a fraction of deal size. 9 3 Nonretention management bonuses are depicted in Figure 1.
Of course, VCs might have other reasons to give executives a management bonus besides inducing them to sell the company. They may want to reward the executives for ajob well done or to curry favor with executives whom they hope to hire in the future. Thus, we cannot be certain that all of the nonretention bonuses observed were salerelated carrots.

91
Retention arrangements-arrangements that require executives to stay with the acquirer-might be used either as (1) arm's-length deals by the acquirer to keep the target's talent or (2) carrots by the VCs to induce executives to support the deal. In our sample, 19 founders reported that retention agreements or related compensation arrangements (such as options for the acquirer stock) were negotiated in connection with the sale. Because we were not given any reason to believe that these arrangements were designed to serve as carrots, and because their value is difficult to measure, we report data only on nonretention bonuses.
92 Most of these payments came at the expense of the VCs. Across the 50 firms in our sample, a total of $26 million was paid out in the form of nonretention management bonuses. Of this amount, $22.5 million came at the expense of preferred shareholders, and only $3.5 million came at the expense of common shareholders.

93
See Hartzell, Ofek & Yermack, supra note 22, at 43-46 (finding that on average, CEOs of publicly held target firms receive a nonretention merger bonus worth approximately $1.2 million, which is only about 0.1% of the average acquisition price-$1.2 billion-of the deals in the study).

Cost of Carrots
In our sample, 23 of the 50 firms offer a management bonus or common carve-out. Limited to the group of 23 firms offering a carrot, Figure 2 illustrates the value of these benefits relative to the sale price. In most cases, total carrots represent between 5% and 12% of the sale price, with a mean amount of approximately $2.9 million (or 9.4% of deal value) ( Table 3). In two firms, however, total carrots exceed 20% of deal value.
Across all 50 firms, the total cost of carrots (those given to executives plus those given to common shareholders) is about $1.33 million for the average sample firm ( Table 3). Across all companies, the average carrot is approximately 4.3% of the purchase price (or 2.4% on a dollar-weighted basis).94 FIGURE 2.
This figure shows, for each of the 23 firms in our sample that provided a common carve-out or management bonus, the payouts as a percentage of the total sale price. Carve-outs to common are shown in dark grey, and management bonuses are shown in light grey. The use of carrots is greater in the 42 firms in our sample where the VC investors did not convert to common stock, and thus where the conflict between the entrepreneurial team and VCs was likely to be the greatest. In this subsample of 42 firms we find that the total 94 On average, preferred shareholders bore $1.26 million of the cost and common shareholders bore the remaining $70,000.

1352
[Vol. 98:1319 cost of carrots is about $1.51 million for the average firm and that approximately 5.0% of deal value (3.1% on a dollar-weighted basis) was used to fund carrots.

B. Sticks
As explained earlier, the two potential sticks available to VCs in dealing with recalcitrant executives are firing them and threatening to blacklist them, and the two sticks that can be used against common shareholders are cross-voting and vote buying. This section describes the use of such sticks in our sample.

Termination of CEO
Recall that 29 of the 50 founders of our firms had been replaced before the sale. But these replacements typically occurred well before the sale, usually in connection with a later round of financing. 95 After three years of operating, approximately 30% of the firms in our sample had replaced the founder-CEO. Of those firms still in operation after six years, approximately 63% of the firms had replaced the founder-CEO. 9 6 Data on the timing of CEO replacement in our sample are summarized in Table 4.  Table 4 reports the identity of the CEO in our sample firms (founder or replacement), as of the end of each year of the firm's operation. Panel B illustrates the frequency of CEO replacement over time. The horizontal axis represents the life span of each firm from formation to sale, normalized to one unit, and the vertical axis represents the percentage of sample firms that have replaced the CEO. 95 In 3 of the 29 firms where the founder was replaced as CEO, the founder-CEO was replaced in connection with the first round of VC financing (meaning the founder agreed to resign as CEO in exchange for the financing). In the other 26 firms, the founder-CEO was replaced at a later date, typically after several rounds of financing. 96 Our evidence regarding the timing of CEO replacement is roughly consistent with Michael T. Hannan, M. Diane Burton, and James N. Baron 525-26 (1996) (finding that the likelihood that a nonfounder will be appointed CEO is about 10% in the first 20 months of a company's life, 40% after 40 months, and 80% after 80 months). fired the founder-CEO to strip him or her of unvested equity, a possibility raised by Michael Klausner and Kate Litvak. 9 8 Indeed, most of the replaced founders reported significant involvement with the firm after the CEO change. The founder-CEO remained on the board following the CEO replacement in almost 60% (17 out of 29) of our sample firms. Even when the founder does not keep a board seat, he or she generally has some position with the firm after replacement (often serving as the Chief Technology Officer or in some other executive capacity).
To be sure, even if we do not observe forcible termination, the possibility of such termination is always present. The possibility of termination limits the value of management bonuses that executives can extract from the VCs.

Blacklisting
In two firms (Firm 35 and Firm 50), VCs explicitly threatened to prevent a founder from ever raising money again in Silicon Valley if he or she did not vote for the sale as a director. In each of these firms, the founder's support was not necessary as a matter of corporate law to conduct the sale-the VCs had control of the board-but the VCs wanted a unanimous board vote to reduce liability risk. And they got it in both situations.
Again, the fact that blacklisting was explicitly threatened in only two firms does not mean that founders or executives in other firms were not affected by the possibility of blacklisting. We have been told by VCs and venture capital lawyers that entrepreneurs are keenly aware, even if they have not been explicitly told, that if they behave "unreasonably" toward the VCs backing their firm, it will be very difficult for them to raise funds from other VCs in the future. Our results thus understate the importance of blacklisting constraints on the entrepreneurial team.

Cross-Voting
In one firm (Firm 29), VC investors engaged in cross-voting, the scheme used in Orban v. Fields. Firm 29, which was California-domiciled, received $50 million over three rounds of VC financing and was sold in 2004 for $20 million. At the time of sale, the VCs already controlled the board of directors, the founder had already been replaced as CEO, and the VCs collectively had liquidation preferences exceeding $100 million (the third-round financing included a 3x preference). The founder refused to vote his common stock in favor of the 98 See Klausner & Litvak, supra note 4, at 68 ("[T]he VC ... could fire the entrepreneur simply to take away his unvested stock . . . .").
sale. 99 Because the firm was incorporated in California, approval by a majority of the common shareholders was necessary to sell the firm, giving the founder some degree of blocking power. To eliminate this holdup, the VCs converted a fraction of their preferred stock into common shares. This maneuver did not cost the VCs anything because, as in Orban, the VCs were left postconversion with liquidation preferences far greater than the $20 million sale price. 0 0 Firm 29 was the only company in our sample where this form of cross-voting was actually used. However, other founders, especially those of firms whose common stock was far underwater, reported that they were aware that VCs could engage in this tactic. Thus, demands for carve-outs to common shareholders are made in the shadow of possible cross-voting. 0 1

Vote Buying
Vote buying reportedly occurred in one firm (Finn 49) in our sample. The firm was California-domiciled, and there were no dragalong rights. The founder reported that the VC-controlled board paid the CEO and other employees to exercise underwater options to acquire and vote common stock in favor of a trade sale for about $80 million. In addition to this stick, the VCs used a carrot: they provided a $4 million carve-out to common shareholders. The founder sued the VCs, alleging vote buying and improprieties associated with a presale financing round. The case was settled for an undisclosed amount.
Of course, implicit and essentially undetectable vote buying might be occurring through management bonuses because of the considerable overlap between shareholders and executives. Thus, we cannot rule out the possibility that vote buying was occurring in other firms in our sample.

CONCLUSION
Trade sales by VCs are actually much more common than IPOs and, in the aggregate, are likely to be just as financially important to VCs. However, unlike IPOs, trade sales do not trigger the intense public-disclosure requirements of the securities laws. As a result, we know relatively little about them.

09
The VCs in this firm did not appear to hold drag-along rights that could compel a vote of the common shareholders.
100 See Orban v. Field, No. 12820, 1997 Del. Ch. LEXIS 48, at *1 (Del. Ch. Apr. 1,  1997). 101 The cost to VCs of cross-voting is higher if the conversion of preferred shares to common shares reduces their total liquidation preferences below the sale price. In such cases, cross-voting would effectively result in a carve-out for the original common shareholders.
In this Article, we seek to shed light on how VCs arrange to sell startups in trade sales. In particular, we investigate how VCs induce the founder, other executives, and common shareholders to go along with a trade sale that they might otherwise have an incentive to resist.
We identify the types of bribes (carrots) and coercive tools (sticks) that VCs might use to induce a reluctant entrepreneurial team to support, or at least not impede, a trade sale. We then investigate the use of carrots and sticks using a hand-collected database of 50 VCbacked Silicon Valley firms sold to acquirers in 2003 and 2004. We find, in our sample, a relatively heavy reliance on carrots. Carrots are used in 45% of the firms, with carrots averaging 9% of deal value in these firms. We also find some use of sticks, such as threats to blacklist founders who refuse to cooperate and attempts to undermine common shareholders' voting rights. But the overt use of these sticks is relatively infrequent.
Our study makes three contributions. First, it sheds light on an important but underexplored aspect of corporate governance in private, VC-backed firms. In particular, it highlights the potentially conflicting interests of the different players in VC-backed firms around trade-sale exits and shows that VCs frequently must overcome potential opposition to these sales by founders, executives, and common shareholders.
Second, our study provides further evidence that managers of target firms can extract value by holding up a sale of the firm. Research has documented the frequent use of side payments to induce executives to participate in the sale of publicly traded firms. We show that such carrots are used not only in public firms with dispersed ownership but also in closely held private firms.
Third, our study provides some evidence on whether venture capitalists are constrained from abusing their power in startups. Although venture capitalists have considerable power in startups, the relatively infrequent overt use of sticks in our sample provides support for the view that reputational or other nonlegal considerations constrain misbehavior by Silicon Valley venture capitalists. We hope that our work is useful to practitioners and academics seeking to better understand the corporate governance of VC-backed firms.