Selling out or going public? A real options signaling approach

Accepted Manuscript

Selling out or going public? A real options signaling approach Michi Nishihara PII: DOI: Reference:

S1544-6123(16)30310-5 10.1016/j.frl.2017.04.003 FRL 704

To appear in:

Finance Research Letters

Received date: Revised date: Accepted date:

15 November 2016 31 March 2017 11 April 2017

Please cite this article as: Michi Nishihara, Selling out or going public? A real options signaling approach, Finance Research Letters (2017), doi: 10.1016/j.frl.2017.04.003

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Highlights • We study the choice between selling out and going public under asymmetric information. • The high-growth firm signals to outsiders through the IPO timing. • The high-growth firm goes public earlier while the low-growth firm sells out later.

AC

CE

PT

ED

M

AN US

• IPO market and IPOs in a cold IPO market.

CR IP T

• Strong asymmetric information and low cash flow volatility foster trade sales in a hot

1

ACCEPTED MANUSCRIPT

Selling out or going public? A real options signaling

Michi NISHIHARA†

Abstract

CR IP T

approach∗

AN US

We examine a dynamic model in which a firm chooses between selling out and going public under asymmetric information. Suppose that the firm prefer to sell out under symmetric information. Under asymmetric information, we have a separating equilibrium in which the high-growth firm goes public earlier while the low-growth firm follows the first-best sales policy because the high-growth firm signals to market investors by doing an IPO. This result is consistent with the empirical evidence. The model also yields

M

implications about the effects of asymmetric information and cash flow volatility on the decisions.

ED

JEL Classifications Code: G31; G34; D82.

AC

CE

PT

Keywords: IPO; signaling; real options; asymmetric information; acquisition; earnout.

∗

This is the version on 31 March, 2017. The first version was written on 15 November, 2016. This work

was supported by the JSPS KAKENHI (Grant number JP26350424, JP26285071) and the JSPS Postdoctoral Fellowships for Research Abroad. The author would like to thank the editors, anonymous referees, Haejun Jeon, Paulo Pereira, and the participants of the 20th EBES Conference and the 19th Czech-Japan Seminar on Data Analysis and Decision Making under Uncertainty for helpful comments. † Corresponding Author. Graduate School of Economics, Osaka University, 1-7 Machikaneyama, Toyonaka, Osaka 560-0043, Japan, E-mail: [email protected], Phone: 81-6-6850-5242, Fax: 81-6-6850-5277

ACCEPTED MANUSCRIPT

1

Introduction

A number of papers (e.g., Lowry and Schwert (2002), Pastor and Veronesi (2005), Ljungqvist, Nanda, and Singh (2006), Chemmanur and He (2011)) have discussed the cyclical nature of initial public offerings (IPOs); namely hot and cold IPO markets. They have explained the mechanism of IPO cycles in terms of investor sentiment, market conditions, and asym-

CR IP T

metric information, among others. In recent years, several papers (e.g., Bustamante (2012), Banerjee, Gucbilmez, and Pawlina (2016)) have examined the IPO timing as a signaling tool under asymmetric information using the framework of real options signaling game. They showed that firms with high growth opportunities accelerate the IPO timing to signal quality of growth opportunities to outsiders. However, they do not consider an alternative choice to go public. In practice, many firms prefer to sell out to another firm rather than going public

AN US

as an independent firm (e.g., Cumming, Fleming, and Schwienbacher (2006), Giot and Schwienbacher (2007), Poulsen and Stegemoller (2008), and Gao, Ritter, and Zhu (2013)). This paper contributes to the literature by revealing how asymmetric information affects the choice between the alternatives as well as its timing.1

This paper builds on the dynamic IPO model of Bustamante (2012). An entrepreneur

M

of a private firm has an opportunity to make a growth investment along with an IPO. There are two types of firms with high and low growth opportunities. The entrepreneur knows the firm’s type, while outside investors, who price the IPO, do not observe the

ED

type. The firm only signals its type to outside investors thorough the IPO timing. In addition to this setup of Bustamante (2012), we assume that the firm has an option to sell out to a large firm rather than going public. Large firms, which bid for the private

PT

firm, do not observe the firm’s type. Following the empirical evidence of Gao, Ritter, and Zhu (2013), we suppose that a bidder can expand the business more efficiently because

CE

of economies of scope and scale so that the entrepreneur always chooses selling out under symmetric information. In the model, we show that one of three types of equilibria occurs depending the

AC

parameter values: both types of firms sell out at the same time (hereafter, pooling equilibrium of trade sales), both types of firms go public at the same time (hereafter, pooling equilibrium of IPOs), or the high-growth firm goes public earlier and the low-growth firm sells out later on (hereafter, separating equilibrium). In particular, we show that the separating equilibrium is likely to occur under stronger asymmetric information and lower volatility. We add to previous results about IPOs in Bustamante (2012) and Banerjee, Gucbilmez, and Pawlina (2016), by providing implications about the choice between the alternatives

1

Compared to numerous studies on IPOs, a smaller number of papers (e.g., Zingales (1995), Pagano and

Roell (1998)) have examined the mechanism of the choice between the alternatives. The previous papers, unlike this paper, do not consider the effects of the IPO and sales timing as a signaling tool.

3

ACCEPTED MANUSCRIPT

as well as its timing. Indeed, the separating equilibrium result predicts that firms with low growth opportunities sell out after firms with high growth opportunities go public. This result is consistent with the empirical evidence. Poulsen and Stegemoller (2008) showed that firms with higher growth opportunities tend to go public rather than selling out. Gao, Ritter, and Zhu (2013) also showed similar findings. Giot and Schwienbacher (2007) examined the dynamics of exit options for venture capital funds and showed that

CR IP T

the IPO exit timing tends to be earlier than the trade sale exit timing. Furthermore, the comparative statics analysis yields the following testable predictions. Firms with strong asymmetric information and low cash flow volatility tend to sell out in a hot IPO market, while they tend to go public in a cold IPO market.

We also show that contingent payments in acquisitions can decrease inefficiency due to asymmetric information and induce the high-growth firm to sell out in the separating

AN US

equilibrium. This result can account for empirical findings that contingent contracts are more likely to be used in acquiring private and younger firms with stronger asymmetric information (e.g., Kohers and Ang (2000), Datar, Frankel, and Wolfson (2002), Ragozzino and Reuer (2009)).

The remainder of this paper is organized as follows. Section 2 introduces the setup. After Section 3.1 briefly explains the symmetric information case, Section 3.2 shows the

Setup

ED

2

M

equilibrium results under asymmetric information. Section 4 concludes the paper.

Consider an all-equity and private firm which is run by an entrepreneur who is the sin-

PT

gle shareholder. The firm receives a continuous stream of cash flows X(t) following a geometric Brownian motion

CE

dX(t) = µX(t)dt + σX(t)dB(t) (t > 0),

X(0) = x,

where B(t) denotes the standard Brownian motion defined in a probability space (Ω, F, P)

AC

and µ, σ(> 0) and x(> 0) are constants. For convergence, we assume that µ < r, where

a positive constant r is the risk-free interest rate. Assume that x is sufficiently low to exclude a firm’s IPO or selling out at time 0. The firm has a growth opportunity where the entrepreneur expands the cash flows to

θi X(t) by issuing a constant fraction of shares, α ∈ (0, 1), and paying the investment cost

I(> 0).2 The growth option potentially takes on two types: θg (i = g: good type) and θb (i = b: bad type), where θg > θb . From now on, we call firms with i = g and b good and bad firms, respectively. Assume θb > 1 because we consider a growth opportunity. 2

Following the basic setup of Bustamante (2012), we assume that the firm issues an exogenously given α

shares in the IPO to finance the growth investment project. Refer to the appendix of Bustamante (2012) to see how α can be determined by a financing constraint.

4

ACCEPTED MANUSCRIPT

The probability of drawing the good type is q ∈ (0, 1). The entrepreneur knows the firm’s

type. Then, the value of the entrepreneur who does an IPO at time T becomes Z ∞ (1 − α)θi X(T ) − I + Pi (T ), ET [ e−r(t−T ) (1 − α)θi X(t)dt] − I + Pi (T ) = r−µ T

(1)

where Pi (T ) denotes the proceeds from the IPO. In addition to this basic setup of Bustamante (2012), we assume that the firm has

CR IP T

an option to sell out instead of going public. We denote by Qi (T ) the proceeds from sales at time T .3 In addition to the payments Qi (T ), an acquirer pays the adjustment and investment costs associated with the acquisition and receives a continuous stream of cash flows ηθi X(t) from the firm’s business. For simplicity, we assume that η ≥ 1 and the acquirer’s adjustment and investment costs equal I.4 This means that an acquirer

can expand the business more efficiently than the entrepreneur due to economies of scope

AN US

and scale. The parameter η can also capture synergies in acquisition (e.g., Poulsen and Stegemoller (2008)) by taking account of an increase in the acquirer’s cash flows in the related business.5 We also assume that all agents are risk-neutral and that investors and bidders are competitive and have all information except for the firm’s real type.6

M

3.1

Model Solutions

Symmetric information

ED

3

In this section, we suppose that investors and bidders observe the firm’s real type. Then, Z Pi (T ) = ET [

PT

we have

and

3

CE

Qi (T ) = ET [

∞

e−r(t−T ) αθi X(t)dt] =

T

Z

∞

T

e−r(t−T ) ηθi X(t)dt] − I =

αθi X(T ) r−µ

(2)

ηθi X(T ) −I r−µ

(3)

For simplicity, we assume that the entrepreneur sells all shares to the acquirer. Our results remain unchanged

AC

even if the entrepreneur keep holdings of a lower fraction of shares than in the IPO case, which means that the effect of asymmetric information is stronger in the sales case than in the IPO case. 4 Our results remain unchanged as long as an acquirer can expand the business more efficiently than the entrepreneur. 5 In practice, acquirers do not necessarily increase more cash flows than entrepreneurs. In some cases, loss of the entrepreneur’s expertise and experience lowers firm value. It depends on the trade-off, but this paper assumes the acquirer’s advantage over the entrepreneur following the recent evidence in Gao, Ritter, and Zhu (2013). 6 Our model applies a situation of selling out by auction rather than by negotiation. In the negotiated sales, unlike in our setup, through the time-consuming negotiation process with a particular firm, the seller and the acquirer fill gaps in information and prices.

5

ACCEPTED MANUSCRIPT

By (1)–(3) and η ≥ 1, selling out generates the higher value than doing an IPO. The optimal sales time is the solution to the optimal stopping problem:   Z T ηθi X(T ) e−rt X(t)dt + e−rT Vi∗ (x) = sup E[ −I ] r−µ T 0   x (ηθi − 1)X(T ) −rT = + sup E[e − I ], r−µ r−µ T

(4)

CR IP T

where the sales time T is optimized over all stopping times. As in the standard real options literature (e.g., Dixit and Pindyck (1994)), we can easily derive the following solution. Proposition 1 The firm sells out. The sales trigger is x∗i =

β (r − µ)I β − 1 ηθi − 1

(i = g, b)

(5)

Vi∗ (x) where β = 1/2 − µ/σ 2 +

x = + r−µ

AN US

and the entrepreneur’s value at the initial time (hereafter, called the firm value) is 

x x∗i

β

I β−1

(i = g, b),

(6)

p (µ/σ 2 − 1/2)2 + 2r/σ 2 (> 1).

∗∗ For later uses, we denote the IPO trigger and firm value by x∗∗ i and Vi (x), which equal

Asymmetric Information

ED

3.2

M

(5) and (6) with η = 1.

Under asymmetric information, investors and bidders do not know the firm’s real type.

PT

The entrepreneur cannot directly prove the firm’s type to them and can signal the type only through the choice between selling out and going public and its timing.7 Note that the bad firm may have an incentive to imitate the good type and receive Pg (T ) and Qg (T )

CE

defined by (2) and (3). Following Bustamante (2012), we derive the least-cost equilibrium for the good firm that pays all costs due to asymmetric information. We first derive

AC

the least-cost separating equilibrium, in which the good firm maximizes the firm value conditional on separating itself from the bad firm. As usual, we restrict our attention to threshold policies expressed as Tg = {t ≥ 0 | X(t) ≥ xg }.

When the good firm conducts an IPO, separates itself from the bad type, and receives

Pg (T ) defined by (2), the good firm value becomes Z E[

Tg

e−rt X(t)dt +

0

x = + r−µ 7



x xg

β 

Z

∞

Tg

e−rt (1 − α)θg X(t)dt + e−rTg

 (θg − 1)xg −I , r−µ



 αθg X(Tg ) −I ] r−µ (7)

Following the basic model of Bustamante (2012), we omit signaling by underpricing in the IPO. Bustamante

(2012) also showed that signaling by the IPO timing tends to be less costly than signaling by underpricing.

6

ACCEPTED MANUSCRIPT

where Tg and xg denote the IPO time and trigger, respectively. The good firm maximizes (7) under the incentive compatibility condition (ICC):  β   (θg − 1)xg x x ss + −I Vg (x) = max x r−µ xg (≥x) r − µ  g β   ((1 − α)θb + αθg − 1)xg x x s.t. + − I ≤ Vb∗ (x). r−µ xg r−µ

(8)

CR IP T

In (8), the left-hand side of the ICC corresponds to the value of the bad firm that imitates the good-type IPO timing and receives Pg (T ) defined by (2), whereas the right-hand side of the ICC corresponds to the value of the bad firm that follows the policy in Proposition 1. Then, under the ICC, the bad firm has no incentive to imitate the good type.

When the good firm sells out, separates itself from the bad type, and receives Qg (T )

AN US

defined by (3), the good firm value becomes   β    Z Tg (ηθg − 1)xg ηθg X(Tg ) x x −I ]= + − I , (9) e−rt X(t)dt + e−rTg E[ r−µ r−µ xg r−µ 0 where Tg and xg denote the sales time and trigger, respectively. As in the IPO case, the

(10)

M

good firm value maximizes (9) under the ICC:  β   (ηθg − 1)xg x x s Vg (x) = max + −I x r−µ xg (≥x) r − µ  g β   (ηθg − 1)xg x x s.t. + − I ≤ Vb∗ (x), r−µ xg r−µ

ED

where the left-hand side of the ICC corresponds to the value of the bad firm that imitates the good-type sales timing and receives Qg (T ) defined by (3), whereas the right-hand side of the ICC corresponds to the value of the bad firm that follows the policy in Proposition

PT

1. By taking the higher value of (8) and (10), we have the following solution.8 Lemma 1 Separating equilibrium:

CE

Case (I): η ≥ θg /θb . The good firm sells out. The good firm’s sales trigger and value are xsg and Vgs (x) = Vb∗ (x), where xsg ∈ (0, x∗g ) is the solution to ηθg xsg /(r − µ) − I = Vb∗ (xsg ).

Case (II): η < θg /θb . The good firm goes public. The good firm’s IPO trigger and

AC

ss ∗∗ ∗ ∗∗ value are xss g and Vg (x) as follows. If ((1 − α)θb + αθg )xg /(r − µ) − I > Vb (xg ) holds, ∗∗ xss g ∈ (x, xg ) is the solution to

((1 − α)θb + αθg )xss g − I = Vb∗ (xss g ), r−µ

and Vgss (x) =

x + r−µ



x xss g

β 

(θg − 1)xss g −I r−µ



= Vb∗ (x) +



(11)

x xss g

β

(1 − α)(θg − θb ) r−µ

∗∗ ss ∗∗ Otherwise, xss g = xg and Vg (x) = Vg (x). In each case, the bad firm sells out, and its

sales trigger and value are x∗b (x) and Vb∗ (x). 8

We can easily show that the solution satisfies the ICC for the good firm, i.e., the good firm has no incentive

to imitate the bad type.

7

ACCEPTED MANUSCRIPT

Proof. In Case (I), problem (10) dominates (8). We can easily derive the trivial solution xsg to problem (10). In Case (II), problem (8) dominates (10). Note that x∗∗ g , which is defined after Proposition 1, is the solution to problem (8) without the ICC. If x∗∗ g does not satisfy the ICC, we obtain the solution xss g defined by (11) because the objective function in (8) monotonically increases with xg ∈ (x, x∗∗ g ]. From the triggers, we can easily calculate the firm values. 

CR IP T

The belief of investors and bidders is as follows: For sales at a trigger in (x, xsg ] in Case (I) and an IPO at a trigger in (x, xss g ] in Case (II), the firm is good at probability one; otherwise, the firm is bad at probability one. In Case (I), the result is trivial. Because of ηθb ≥ θg , the good firm prefers to sell out even though its payoff is the same as that of the bad firm. In Case (II), both the policy and the value depend on the firm’s type. The good firm signals to outsiders by doing an IPO although it is better off selling out

AN US

under symmetric information. In the binding case, as in Bustamante (2012), the good firm accelerates the IPO timing to avoid imitation, while in the non-binding case the choice of ∗∗ ∗ an IPO is enough to separate itself from the bad firm.9 Note that xss g ≤ xg < xb holds

in Case (II). This implies that the good firm’s IPO timing is earlier than the bad firm’s sales timing.

M

Next, we derive the least-cost pooling equilibrium for the good firm. Consider the ¯ sales case, in which a bidder pays η θX(T )/(r − µ) − I, where θ¯ = qθg + (1 − q)θb . The

ED

sales time is the solution to the optimal stopping problem:   ¯ Z T η θX(T ) p −rt −rT − I ]. V (x) = sup E[ e X(t)dt + e r−µ T 0

PT

In a standard manner, we can easily derive the following solution. Lemma 2 Pooling equilibrium of trade sales:

CE

The firm sells out. The sales trigger is xp =

AC

and the firm value is

regardless of its type.

V p (x) =

β (r − µ)I β − 1 η θ¯ − 1

 x β I x + , r−µ xp β−1

The belief of outsiders is as follows: For selling out at the trigger xp , the firm is good at probability q and bad at probability 1 − q; otherwise (off the equilibrium path), the firm is

bad type at probability one.10 It is easy to see x∗g < xp < x∗b and Vb∗ (x) < V p (x) < Vg∗ (x). 9

As in Leland and Pyle (1977), retention of a fraction 1 − α of shares in the IPO is a signal of good quality

to outsiders. 10 Note that the belief of outsiders must be consistent with the firms’ policies. As in Bustamante (2012), we consider the belief of the bad type off the equilibrium path so that the belief can satisfy the consistency. Indeed, under this belief, neither good nor bad firms can increase the payoff by deviation due to the low sales price off the path.

8

ACCEPTED MANUSCRIPT

¯ We now consider the IPO case, in which investors pay αθX(T )/(r − µ). The IPO time

is the solution to the optimal stopping problem:  ¯  Z T Z ∞ αθX(T ) pp −rt −rt −rT Vg (x) = sup E[ e X(t)dt + e (1 − α)θg X(t)dt + e −I ] r−µ T 0 T

Lemma 3 Pooling equilibrium of IPOs: The firm goes public. The IPO trigger is xpp =

(r − µ)I β , β − 1 (1 − α)θg + αθ¯ − 1

regardless of its type. The good firm value is  x β I x + r−µ xpp β−1

and the bad firm value is11 Vbpp (x) =

AN US

Vgpp (x) =

CR IP T

In a standard manner, we can easily derive the following solution.

  x β  ((1 − α)θ + αθ¯ − 1)xpp x b + − I . r−µ xpp r−µ

M

The belief of outsiders is as follows: For an IPO at the trigger xpp , the firm is good at probability q and bad at probability 1 − q; otherwise (off the equilibrium path), the firm

ED

pp < x∗∗ and V ∗∗ (x) < V pp (x) < V ∗∗ (x). is bad type at probability one. We have x∗∗ g g g
By comparing Vgs (x), Vgss (x), V p (x), and Vgpp (x) in Lemmas 1–3, we obtain the least-

PT

cost equilibrium for the good type.12

Proposition 2 The least-cost equilibrium:

CE

Case (I): η ≥ θg /θb . The pooling equilibrium of trade sales is chosen. Case (II-i): (1 − α)θg /θ¯+ α ≤ η < θg /θb . If V p (x) ≥ V ss (x) holds, the pooling equilibrium g

AC

of trade sales is chosen. Otherwise, the separating equilibrium is chosen. Case (II-ii): 1 ≤ η < (1 − α)θg /θ¯ + α. If Vgpp (x) ≥ Vgss (x) holds, the pooling equilibrium of IPOs is chosen. Otherwise, the separating equilibrium is chosen.

Case (II-ii) with η = 1 corresponds to the basic result of Bustamante (2012). The separating equilibrium in Cases (II-i) and (II-ii) predicts that firms with high growth opportunities go public earlier and firms with low growth opportunities sell out later on. This result is consistent with the empirical evidence about IPOs. For instance, Chemmanur and He

11

Vbpp (x) could be lower than Vb∗ (x) for some parameter values. However, when the pooling equilibrium of

IPOs is chosen (see Proposition 2), Vbpp (x) is higher than Vb∗ (x), which means that the bad firm has no incentive to deviate from the pooling equilibrium of IPOs. 12 As shown in Bustamante (2012), the least-cost equilibrium for the good firm is chosen by the intuitive criterion and the Pareto-dominant criterion of alternative equilibrium refinements in the Bayes-Nash equilibrium in which both types of firms jointly have incentives to reveal their type or cluster.

9

ACCEPTED MANUSCRIPT

(2011), Bustamante (2012), and Banerjee, Gucbilmez, and Pawlina (2016) showed that firms with higher growth opportunities go public earlier. More notably, our result can also explain empirical findings about the choice between the alternatives as well as its timing. Poulsen and Stegemoller (2008) showed that firms with lower growth opportunities tend to avoid IPOs and choose to sell out, while Gao, Ritter, and Zhu (2013) showed that less profitable firms tend to sell out. Giot and Schwien-

CR IP T

bacher (2007), who examined the dynamics of exit options for venture capitalists, showed that the IPO exit timing tends to be earlier than the trade sale exit timing. ¯ which causes the pooling In Cases (II-i) and (II-ii), a higher q leads to a higher θ, equilibrium rather than the separating equilibrium. This is consistent with the result of Bustamante (2012). Further, we explore the effects of asymmetric information and volatility on the equilibrium in numerical examples. We set the base parameter values as

AN US

r = 0.07, µ = 0.03, σ = 0.25, α = 0.5, q = 0.5, x = 0.01, θg = 2.3 and θb = 1.7. In Figure 1 we change levels of θg − θb with θ¯ = 2 fixed, while in Figure 2 we change levels of σ. In the figures, the regions S, P S, and P I stand for the separating equilibrium, the pooling equilibrium of trade sales, and the pooling equilibrium of IPOs, respectively. We can see from the figures that the separating equilibrium occurs for high levels of θg − θb and low levels of σ. For low levels of η, a lower θg − θb and a higher σ lead to the

M

pooling equilibrium of IPOs. For high levels of η, a lower θg −θb and a higher σ lead to the

pooling equilibrium of trade sales. We relate hot and cold IPO markets to low and high

ED

η, respectively, because a lower η fosters an IPO. Then, our results yield the following predictions which have not been tested. Firms with strong asymmetric information and low cash flow volatility tend to sell out in a hot IPO market, while they tend to go public

PT

in a cold IPO market.

Our model is also related to a contingent contract, in which an acquirer first pays a fraction of the total consideration and later pays the remaining value contingent on

CE

the performance (e.g., see earnouts in Lukas, Reuer, and Welling (2012)). Consider a contingent contract in which a bidder pays a fraction of the firm value, denoted by α ˜ , at

AC

the acquisition timing and pays the remaining fraction 1 − α ˜ contingent on the realized cash flows. For instance, suppose that α ˜ is low enough to satisfy η((1 − α ˜ )θb + α ˜ θg ) ≤

(1 − α)θb + αθg . In such a case, the ICC in (10) is looser than the ICC in (8); hence,

the good firm sells out in the separating equilibrium. As α ˜ → 0, the good firm value

approaches the first-best value Vg∗ (x). In other words, the good firm can benefit from such a contingent contract because it reduces loss due to asymmetric information. This result is consistent with the empirical findings in Kohers and Ang (2000), Datar, Frankel, and Wolfson (2002), and Ragozzino and Reuer (2009). They show that contingent payments are more likely to be used in the acquisition of private firms, especially new ventures (i.e., in higher asymmetric information cases) than public and established firms (i.e., in lower asymmetric information cases).

10

ACCEPTED MANUSCRIPT

4

Conclusion

We examined a real options signaling game in which a firm chooses between selling out and going public. We shed light on the choice between the alternatives and its timing by deriving three types of equilibria; one separating and two pooling equilibria. Most notably, we showed that in the separating equilibrium the high-growth firm goes public

CR IP T

earlier and the low-growth firm sells out later on. This is because the high-growth firm avoids imitation and signals to outsiders by choosing an IPO and accelerating the IPO timing. The separating equilibrium result can account for empirical findings of the choice between the alternatives as well as its timing.

Furthermore, the comparative statics result showed that the effects of asymmetric information and cash flow volatility on the choice depend on whether an IPO market is hot or cold. Indeed, the result yielded testable predictions that strong asymmetric information

AN US

and low cash flow volatility foster trade sales in a hot market, while they foster IPOs in a cold market. The model can also account for empirical findings that contingent payments are more likely to be used in M&A under stronger asymmetric information. A limitation of the model is that we do not focus on the allocation of control rights between entrepreneurs and investors such as venture capitalists. An entrepreneur, unlike

M

venture capitalists, can prefer an IPO due to private benefits associated with an IPO even when acquisition offers a higher payoff. Actually, Cumming (2008) and Cumming and Johan (2008) showed that stronger venture capital control rights increase the likelihood

ED

of trade sales rather than IPOs. Further, whether an IPO is preferable to a trade sale may depend on the quality of the exchange. For instances, Cumming, Fleming, and Schwienbacher (2006) showed that legal structures affect IPO exits of venture capitalists,

PT

while Johan (2010) showed that listing standards matter to IPO decisions. It could be

CE

interesting topics for future research to incorporate these issues.

References

AC

Banerjee, S., U. Gucbilmez, and G. Pawlina, 2016, Leaders and followers in hot IPO markets, Journal of Corporate Finance 37, 309–334.

Bustamante, M., 2012, The dynamics of going public, Review of Finance 16, 577–618. Chemmanur, T., and J. He, 2011, IPO waves, product market competition, and the going public decision: Theory and evidence, Journal of Financial Economics 101, 382–412. Cumming, D., 2008, Contracts and exits in venture capital finance, Review of Financial Studies 21, 1947–1982. Cumming, D., G. Fleming, and A. Schwienbacher, 2006, Legality and venture capital exits, Journal of Corporate Finance 12, 214–245.

11

ACCEPTED MANUSCRIPT

Cumming, D., and S. Johan, 2008, Preplanned exit strategies in venture capital, European Economic Review 52, 1209–1241. Datar, S., R. Frankel, and M. Wolfson, 2002, Earnouts: The effects of adverse selection and agency costs on acquisition techniques, Journal of Law, Economics, and Organization 17, 201–238.

Press: Princeton).

CR IP T

Dixit, A., and R. Pindyck, 1994, Investment Under Uncertainty (Princeton University

Gao, X., J. Ritter, and Z. Zhu, 2013, Where have all the IPOs gone?, Journal of Financial and Quantitative Analysis 48, 1663–1692.

Giot, P., and A. Schwienbacher, 2007, IPOs, trade sales and liquidations: Modelling 702.

AN US

venture capital exits using survival analysis, Journal of Banking and Finance 31, 679–

Johan, S., 2010, Listing standards as a signal of IPO preparedness and quality, International Review of Law and Economics 30, 128–144.

Kohers, N., and J. Ang, 2000, Earnouts in mergers: Agreeing to disagree and agreeing to stay, Journal of Business 73, 445–476.

M

Leland, H., and D. Pyle, 1977, Informational asymmetries, financial structure, and financial intermediation, Journal of Finance 32, 371–387.

ED

Ljungqvist, A., V. Nanda, and R. Singh, 2006, Hot markets, investor sentiment, and IPO pricing, Journal of Business 79, 1667–1702. Lowry, M., and G. Schwert, 2002, IPO market cycles: Bubbles or sequential learning?,

PT

Journal of Finance 57, 1171–1200.

Lukas, E., J. Reuer, and A. Welling, 2012, Earnouts in mergers and acquisitions: A

CE

game-theoretic option pricing approach, European Journal of Operational Research 223, 256–263.

AC

Pagano, M., and A. Roell, 1998, The choice of stock ownership structure: Agency costs, monitoring, and the decision to go public, Quarterly Journal of Economics 113, 187–

225.

Pastor, L., and P. Veronesi, 2005, Rational IPO waves, Journal of Finance 60, 1713–1757. Poulsen, A., and M. Stegemoller, 2008, Moving from private to public ownership: Selling out to public firms versus initial public offerings, Financial Management 37, 81–101. Ragozzino, R., and J. Reuer, 2009, Contingent earnouts in acquisitions of privately held targets, Journal of Management 35, 857–879.

Zingales, L., 1995, Insider ownership and the decision to go public, Review of Economic Studies 62, 425–448.

12

ACCEPTED MANUSCRIPT

1.1 1 0.9 0.8

θg−θb

0.7 0.6 0.5 0.4

PS

0.3

PI

0.1

1

1.02

1.04

1.06

1.08

AN US

0.2

CR IP T

S

1.1 η

1.12

1.14

1.16

1.18

1.2

0.4

PT

PI

ED

M

Figure 1: The effects of asymmetric information on the equilibrium.

0.35

CE

PS

S

AC

σ

0.3

0.25

0.2

0.15

0.1

1

1.02

1.04

1.06

1.08

1.1 η

1.12

1.14

1.16

1.18

1.2

Figure 2: The effects of volatility on the equilibrium.

13