ARTICLE IN PRESS
Journal of Financial Economics 72 (2004) 319–356
The effect of capital market characteristics
on the value of start-up ?rms$
Roman Indersta,c , Holger M. M .ullerb,c,*
a London School of Economics, London WC2A 2AE, UK
b Stern School of Business, New York University, New York, NY 10012, USA
c Centre for Economic Policy Research (CEPR), London ECIV 7RR, UK
Received 31 July 2002; accepted 5 June 2003
Abstract
We develop an equilibrium model of contracting, bargaining, and search in which the
relative scarcity of venture capital affects the bargaining power of entrepreneurs and venture
capitalists. This in turn affects the pricing, contracting, and value creation in start-ups. The
relative scarcity of venture capital is endogenous and depends on the pro?tability of venture
capital investments, entry costs, and transparency of the venture capital market. Supply and
demand conditions also affect the incentives of venture capitalists to screen projects ex ante.
We characterize both the short- and long-run dynamics of the venture capital industry, which
$We are grateful to an anonymous referee for valuable comments and suggestions. Thanks also to
Patrick Bolton, Mike Burkart, Denis Gromb, Thomas Hellmann, Pete Kyle, Stewart Myers, Enrico
Perotti, Per Str .omberg, Elu von Thadden, Masako Ueda, Jeff Wurgler, and seminar participants at
Dartmouth, Amsterdam, Stockholm School of Economics, Mannheim, Tilburg, Humboldt, the NBER
Corporate Finance Meeting in Chicago (2003), the Atlanta Fed Financial Markets Conference ‘‘Venture
Capital and Technology’’ in Sea Island (2002), the European Finance Association Meetings in Berlin
(2002), the Five Star Conference on Research in Finance at NYU (2001), the workshop ‘‘Strategic
Interactions in Relationship Finance: BankLending and Venture Capital’’ at LSE (2001), the workshop
‘‘Finance for the XXI Century’’ in Amsterdam (2001), and the European Summer Symposium in Financial
Markets (ESSFM) in Gerzensee (2001) for helpful comments and discussions. Earlier versions of this
paper circulated under the titles ‘‘Competition and Ef?ciency in the Market for Venture Capital’’ and
‘‘Venture Capital Contracts and Market Structure’’. Inderst acknowledges ?nancial support from the
Financial Markets Group (FMG).
*Corresponding author. Stern School of Business, New YorkUniversity, New York, NY 10012, USA.
E-mail addresses: r.inderst@lse.ac.uk(R. Inderst), hmueller@stern.nyu.edu (H.M. M .uller).
0304-405X/$ - see front matter r 2003 Elsevier B.V. All rights reserved.
doi:10.1016/j.j?neco.2003.06.001

---

ARTICLE IN PRESS
320
R. Inderst, H.M. M .uller / Journal of Financial Economics 72 (2004) 319–356
provides us with a stylized picture of the Internet boom and bust periods. Our model is
consistent with existing evidence and provides a number of new empirical predictions.
r 2003 Elsevier B.V. All rights reserved.
JEL classi?cation: D83; G32; L22
Keywords: Venture capital; Capital market competition; Bargaining power; Search
1. Introduction
The venture capital industry is highly cyclical, with periodic changes in supply and
demand conditions (Gompers and Lerner, 1999; Lerner, 2002). Between the fourth
quarters of 2000 and 2001 alone, for instance, total funds raised dropped by more
than 80%. In this paper, we askwhether, and how, such variations in capital supply
affect the competitive pricing, contracting, and value creation in start-ups.
Our theory is based on two building blocks. The ?rst is a model of contracting and
bargaining in start-ups. Building on Sahlman (1990), Kaplan and Str .omberg (2002),
and Hellmann and Puri (2001), we model the relationship between the entrepreneur
and venture capitalist as a double-sided incentive problem: A greater fraction of the
?rm owned by the venture capitalist improves the venture capitalist’s incentives but
weakens the entrepreneur’s incentives. Ef?ciency requires balancing the two
incentive problems, or equivalently, balancing ownership shares. Actual ownership
shares, however, are determined by bargaining, and thus by the relative strength of
the entrepreneur’s and venture capitalist’s outside options.
The second building blockis a search model linking outside options to the relative
scarcity of venture capital. An increase in capital supply, for instance, makes it easier
for an entrepreneur to obtain ?nancing, thereby increasing his outside option vis-"a-
vis a venture capitalist. The supply of capital, in turn, is endogenous and depends on
primitive market characteristics such as the pro?tability of investments, entry costs,
and capital market transparency.
The end result is an equilibrium model in which capital market characteristics
affect the relative supply and demand for capital, which in turn affects bargaining
powers and ownership shares, which in turn affects the pricing and value creation in
start-ups.
The core of our model is an analysis of the short- and long-run dynamics of the
venture capital industry. In the short run, the number of venture capitalists is ?xed.
In the long run, changes in market conditions lead to entry or exit of venture
capitalists, thereby making capital market competition endogenous. Our model
predicts that an increase in the return to venture capital investments—whether true
or merely perceived—leads to new entry and a long-run increase in capital market
competition, coupled with a rise in valuations and a decline in venture capitalists’
ownership shares. A decrease in investment returns, on the other hand, leads to exit,
a drop in valuations, and better deal terms for those venture capitalists remaining in
the market.

---

ARTICLE IN PRESS
R. Inderst, H.M. M .uller / Journal of Financial Economics 72 (2004) 319–356
321
While stylized, our model can provide a useful picture of the Internet boom and
bust periods. As winners often tend to materialize quicker than losers (poor
performers may be able to hold out until their cash is ?nally burned up), the initial
success stories at the beginning of the Internet boom period might not have been
representative of the industry as a whole. The general public and investors might
have therefore overestimated the true returns to Internet investments. As more and
more ?rms began to fail, investors realized that their initial assessment was wrong.
They consequently adjusted their return estimates downward. In our model, the
boom and bust of the Internet correspond to an increase and decrease, respectively,
in the perceived return to venture capital investments.
We also examine the equilibrium effects of changes in entry costs and capital
market transparency. Such changes affect outside options either directly or via their
effect on capital market competition. Our model predicts that an increase in
transparency improves the value created in start-ups, while a decrease in entry costs
can destroy value if the aggregate capital supply is already relatively high. Finally,
capital market competition affects the incentives not only after but also prior to the
formation of a new venture. In an extension of our model, we show that venture
capitalists are more likely to screen projects in ‘‘down markets’’ when competition
among investors is weak, and less likely in ‘‘hot markets’’ when competition is
strong.
Our search model contrasts with traditional venture capital contracting models,
which consider an isolated setting with one entrepreneur and one venture capitalist
at a time. These models assume a ‘‘competitive capital market’’ in which
entrepreneurs extract all the surplus. In a world with many venture capitalists and
many entrepreneurs, however, it is not clear why entrepreneurs should extract all the
surplus. On the contrary, anecdotal evidence suggests that the ability to extract
surplus shifts backand forth between entrepreneurs and investors, depending on
who is currently in short supply.1 In this paper, we explicitly depart from the
standard assumption that entrepreneurs have all the bargaining power. As we show,
this standard assumption is not innocuous. By giving venture capitalists bargaining
power, one can get closer to the sharing rule that maximizes the joint surplus. In fact,
if venture capitalists and entrepreneurs each have the ‘‘right’’ amount of bargaining
power, ef?cient surplus sharing is possible.
A central tenet of our model is that changes in capital market competition and
bargaining power translate into changes in ownership shares. Supporting evidence is
provided by Gompers and Lerner (2000), who ?nd a positive relation between the
valuation of new ventures and capital in?ows, suggesting that ‘‘increases in the
supply of venture capital may result in greater competition to ?nance companies and
1 The following statement compares the height of the Internet bubble—when ‘‘too much money was
chasing too few deals’’ (Gompers and Lerner, 2000)—with the aftermath: ‘‘If you went into a ... start-up
three to six months ago, you almost certainly got a very bad deal. Companies could askfor anything they
wanted [in terms of valuation]. Now entrepreneurs are much more realistic’’ (Financial Times, ‘‘Open
Season for Europe’s Turkeys,’’ January 11, 2001). Along similar lines, Bartlett (2001b) argues that the
burst of the bubble brought about ‘‘changes in deal terms ... all of which are designed to enhance returns
and the quantum of control enjoyed by nervous investors.’’

---

ARTICLE IN PRESS
322
R. Inderst, H.M. M .uller / Journal of Financial Economics 72 (2004) 319–356
rising valuations.’’ (Ceteris paribus, a higher valuation implies a smaller ownership
share for venture capitalists).
Another tenet of our model is the linkbetween ownership shares and incentives.
Empirical support is provided by Kaplan and Str .omberg (2002), who ?nd that equity
incentives increase the likelihood that venture capitalists provide value-adding
support activities. Similarly, industry observers have expressed concerns that
unfavorable deal terms (from the perspective of entrepreneurs) in the recent down
market might have had a negative effect on entrepreneurial incentives: ‘‘The terms of
current venture ?nancings can be such that founders may well lose interest. ... They
start plotting their next career move, perhaps with a competitor, from the date the
deal is closed. In short, the VCs, while putting in place extremely favorable terms
from their point of view, face the possibility of shooting themselves in the feet’’
(Bartlett, 2001a).
Michelacci and Suarez (2002) also have a search model of start-up ?nancing.
Unlike this paper, however, Michelacci and Suarez do not consider incentive
contracts or the inef?ciencies arising from an imbalance of ownership shares. Rather,
they focus on search inef?ciencies, using an insight from the search literature that
entry creates externalities for the matching chances of other market participants.
The rest of the paper is organized as follows. Section 2 presents the model. Section 3
characterizes the equilibrium when capital market competition is exogenous. Section 4
endogenizes the level of capital market competition. The exogenous variables are
primitive market characteristics such as investment returns, entry costs, and capital
market transparency. Section 5 considers robustness issues as well as welfare and
policy implications. Section 6 examines the incentives of venture capitalists to screen
projects ex ante. Section 7 summarizes the empirical implications and compares them
with the available evidence. Section 8 concludes. All proofs are in Appendix B.
2. The model
The model has two building blocks: (i) a model of contracting and bargaining in
start-ups, and (ii) an equilibrium model of search. We ?rst derive the contract frontier
characterizing the utilities of the entrepreneur and venture capitalist for all Pareto
optimal contracts. We then derive the bargaining solution, which determines a point
(i.e., contract) on the contract frontier. In the bargaining problem, we take the
entrepreneur’s and venture capitalist’s outside options as given. We ?nally embed the
bargaining problem in a search market to endogenize outside options as a function
of the relation between capital supply and demand, or degree of capital market
competition. The degree of capital market competition, in turn, is taken as given. It
will be endogenized in Section 4 when we introduce free entry of capital.
2.1. Financial contracting
A penniless entrepreneur has a project requiring an investment I > 0: Financing is
provided by a venture capitalist. The project payoff is XlX0 with probability 1 À p

---

ARTICLE IN PRESS
R. Inderst, H.M. M .uller / Journal of Financial Economics 72 (2004) 319–356
323
and Xh > I with probability p: The success probability p ¼ pðe; aÞ depends on the
entrepreneur’s and venture capitalist’s non-contractible efforts eA½0; 1? and aA½0; 1?:
Effort costs are denoted by cðeÞ and gðaÞ; respectively, and are strictly convex. All
agents are riskneutral, which implies our results continue to hold if Xl and Xh are
expected values instead of ?nal cash ?ows.
In the base model (Sections 2–4) we assume that Xl ¼ 0: The division of the
project payoff is then fully characterized by a sharing rule sA½0; 1? representing the
venture capitalist’s ownership, or equity share. The case where Xl > 0 is considered in
Section 5. In that section, we also consider the possibility that the venture capitalist
pays the entrepreneur a wage.
Given some sharing rule s; the entrepreneur’s and venture capitalist’s utilities are
uðsÞ  pð1 À sÞXh À cðeÞ and vðsÞ  psXh À gðaÞ; respectively. By varying s from zero
to one, we can trace out the utility possibility frontier depicting the set of all possible
u À v combinations. This frontier need not be decreasing everywhere. For instance, if
one side is more productive than the other, both utilities might be increasing over
some range (see Fig. 1). The same is true if efforts are complements. We call the
undominated, i.e., decreasing, segment of the utility possibility frontier the equity
frontier. The equity frontier depicts the set of all Pareto-optimal u À v combinations.
It is derived from the set of Pareto-optimal equity shares and denoted by u ¼ cðvÞ:
The domain of the equity frontier is ½v; %v?; where vXvð0Þ and %vpvð1Þ are the venture
%
%
capitalist’s utilities under the lowest and highest Pareto-optimal equity share,
respectively.2
The shape of the equity frontier depends on the production technology pðe; aÞ: We
focus on production technologies that are well behaved in the following sense:
(i) the function cðvÞ is decreasing and strictly concave over ½v; %v?;
%
(ii) the sum uðsÞ þ vðsÞ has a unique maximum in the interior of ½v; %v?; and
%
(iii) the venture capitalist’s utility vðsÞ is increasing in s over ½v; %v?:
%
Properties (i)–(ii) follow naturally from the fact that the incentive problem is two-
sided and effort costs are strictly convex. Maximizing the sum of utilities then
requires balancing the two incentive problems. In particular, giving one side a very
high and the other side a very low equity share does not maximize total utility, as the
side with the high equity share will then provide effort at a level where his or her
marginal effort cost is extremely high. Consequently, the total utility u þ v ¼
cðvÞ þ v has a unique maximum in the interior of ½v; %v?: De?ne #s  arg max½uðsÞ þ
%
vðsÞ?; #v  vð#sÞ; and #u  uðvð#sÞÞ: We refer to the surplus-maximizing sharing rule #s as
the ef?cient sharing rule (or ef?cient equity share).3 Evidently, it holds that c0ð #vÞ ¼
2 See Appendix A for examples. If the venture capitalist and entrepreneur have different productivities
or if efforts are complements, the lowest Pareto-optimal equity share will be positive and the highest
Pareto-optimal equity share will be less than one, implying that v > vð0Þ and %vovð1Þ: (See Fig. 1 for an
%
illustration.) In this case, it is not Pareto-optimal to give the venture capitalist or the entrepreneur all the
equity.
3 Formally, #s is ef?cient within the class of budget-balanced mechanisms. Allowing budget-breaking
mechanisms "a la Holmstr .om (1982) might yield superior outcomes. Hence #s is constrained (or second-best)
ef?cient.

---

ARTICLE IN PRESS
324
R. Inderst, H.M. M .uller / Journal of Financial Economics 72 (2004) 319–356
À1; i.e., the equity frontier has a slope of minus one at the point where total utility is
maximized. The third property states that the venture capitalist’s utility is increasing
in her equity share. This automatically implies that along the equity frontier we have
u0ðsÞo0; ruling out situations where a given value of cðvÞ is associated with more
than one equity share.
All three assumptions are innocuous and satis?ed by many production
technologies. In Appendix A we give two examples of technologies used in the
venture capital literature that satisfy (i)–(iii): the linear technology pðe; aÞ ¼ ga þ
ð1 À gÞe used in, e.g., Casamatta (2003), and the Cobb-Douglas technology pðe; aÞ ¼
age1Àg used in, e.g., Repullo and Suarez (2000). Under the linear technology the two
efforts are substitutes, while under the Cobb-Douglas technology they are
complements. To make the problem nontrivial, we assume that #v > I; i.e., the
surplus-maximizing allocation is suf?ciently pro?table to allow the venture capitalist
to breakeven.
The entrepreneur’s and venture capitalist’s deal utilities are U ðsÞ  uðsÞ and V ðsÞ 
vðsÞ À I ; respectively. The deal utilities are the overall utilities from the contract ðs; I Þ:
The utilities derived from equity shares, uðsÞ and vðsÞ; are thus merely one component
of the agents’ deal utilities. The investment outlay I is another component. In
Section 5 we will introduce two more components: the safe project payoff Xl and
wage payments.
The set of all Pareto-optimal U À V combinations is called the contract frontier.
The contract frontier is derived from the set of Pareto-optimal contracts ðs; I Þ and
denoted by U ¼ CðV Þ: In the base model where Xl ¼ 0; the contract frontier is
obtained by shifting the equity frontier cðvÞ to the left by I ; implying that CðV Þ 
cðV þ I Þ; see Fig. 1. In Section 5 when we reintroduce Xl > 0 and wage payments,
the construction of the contract frontier is more complicated. The domain of the
contract frontier is ½V ; %
V?; where V  maxfv À I ; 0g and %
V  %v À I: (As the venture
%
%
%
capitalist and entrepreneur never bargain to a point where the venture capitalist’s
deal utility is negative, the constraint that V X0 is without loss of generality.)
%
Fig. 1 depicts the utility possibility frontier characterizing all possible u À v
combinations, the equity frontier characterizing all Pareto-optimal u À v combina-
tions, and the contract frontier characterizing all Pareto-optimal U À V combina-
tions for the Cobb-Douglas technology.
2.2. Bargaining
It is reasonable to assume that when bargaining over a contract, the entrepreneur
and venture capitalist select a contract that is Pareto ef?cient. Bargaining thus
corresponds to choosing a utility pair ðV ; UÞ on the contract frontier. If the
bargaining breaks down, the entrepreneur and venture capitalist realize their out-
side options U o and V o; respectively. For the time being, we shall take these
outside options as given. They will be endogenized in Section 4 as a function
of relative supply and demand in the capital market. Our bargaining concept is
the generalized Nash bargaining solution. Accordingly, the bargaining outcome
consists of deal utilities U ¼ CðV ÞXU o and V XV o maximizing the Nash product

---

ARTICLE IN PRESS
R. Inderst, H.M. M .uller / Journal of Financial Economics 72 (2004) 319–356
325
Fig. 1. Equity frontier and contract frontier for the Cobb-Douglas technology. The dashed curve
represents the utility possibility frontier (UPF) depicting the entrepreneur’s and venture capitalist’s utilities
from equity shares, u and v; for all possible equity shares sA½0; 1?: The decreasing segment of the UPF is
the equity frontier. It contains all utilities u and v associated with a Pareto-optimal equity share. Shifting
the equity frontier to the left by the investment amount I yields the bold contract frontier CðV Þ; which
depicts the entrepreneur’s and venture capitalist’s utilities from Pareto-optimal contracts, U ¼ CðV Þ and
V ; respectively.
½V À V o?Z½CðV Þ À Uo?1ÀZ; where ZAð0; 1Þ: For expositional convenience, we de?ne
b  Z=ð1 À ZÞ:
As the contract frontier is strictly concave, the bargaining problem has a unique
solution. In Appendix B we show that this solution must lie in the interior of ½V ; %
V?:
%
Denote the bargaining solution by ðV d; U dÞ; where U d  CðV dÞ: The superscript
indicates that V d and U d are the equilibrium deal utilities. Maximizing the Nash
product with respect to V ; we obtain
Lemma 1. The equilibrium deal utilities V d and U d are uniquely determined by
V d À V o
b ¼ ÀC0ðV dÞ
;
ð1Þ
CðV dÞ À U o
where U d ¼ CðV dÞ:
The venture capitalist’s deal utility V d is increasing in her outside option V o and
decreasing in the entrepreneur’s outside option U o: The reverse is true for the
entrepreneur. As is well known, the axiomatic Nash bargaining solution can be
derived as the limit of a non-cooperative bargaining game where the two parties
bargain with an open time horizon under the riskof breakdown (Binmore et al.,
1986). It is worth noting that our results do not depend on the speci?cs of the Nash
bargaining solution. All we need is that an agent’s deal utility is positively related to
his or her own outside option and negatively related to the counterparty’s outside
option. Any bargaining solution with this feature yields similar results.

---

ARTICLE IN PRESS
326
R. Inderst, H.M. M .uller / Journal of Financial Economics 72 (2004) 319–356
2.3. Search
To endogenize outside options, we embed the bargaining problem in a market
environment. We consider a stationary search market populated by entrepreneurs
and venture capitalists.4 The measure of entrepreneurs and venture capitalists in the
market is Me and Mv; respectively. A key variable is the ratio of venture capitalists to
entrepreneurs, or degree of capital market competition, Mv=Me  y: A high value of y
indicates that the capital market is very competitive. Each venture capitalist has
capital k; which implies she can ?nance at most one project. All our results extend to
the case where each venture capitalist can ?nance a ?nite number of projects. As each
venture capitalist has a ?xed amount of capital, the ratio Mv=Me  y also indicates
the relative magnitude of capital supply and demand. Time is continuous, and the
discount rate is r > 0:5
The measure of deals, or matches, per unit of time is given by the matching
function xðMe; MvÞ: From the perspective of a venture capitalist, the (Poisson)
arrival rate of a deal is xðMe; MvÞ=Mv; while the entrepreneur’s deal arrival rate
is xðMe; MvÞ=Me: We assume that the matching function exhibits constant returns
to scale. This has the convenient implication that arrival rates depend solely on
the degree of capital market competition y: (See Section 5.4 for details.) Speci?-
cally, the venture capitalist’s deal arrival rate is xðMe; MvÞ=Mv  qvðyÞ; which
is decreasing in y with limy-0 qvðyÞ ¼ N and limy- q
N
vðyÞ ¼ 0:
Hence a
venture capitalist is more likely to meet an entrepreneur in a given time interval
if the ratio of venture capitalists to entrepreneurs is low. As the measure of deals
per unit of time is MvqvðyÞ; the entrepreneur’s deal arrival rate is yqvðyÞ  qeðyÞ;
which is an increasing function of y: Hence an entrepreneur is more likely to meet a
venture capitalist in a given time interval if the ratio of venture capitalists to
entrepreneurs is high.
If the search is successful, the venture capitalist and entrepreneur bargain over a
contract. The outside options in the bargaining, U o and V o; are the utilities from
going back into the market and searching anew. Given that the market is stationary,
the utility from going backinto the mark
et equals the utility from entering the
market in the ?rst place. Hence U o and V o represent both the outside options in the
bargaining as well as the overall utilities from searching.
We now determine the outside options. Consider ?rst the entrepreneur’s outside
option U o: The Poisson arrival rate of a deal for the entrepreneur is qeðyÞ: The
probability that a deal occurs in a small time interval D is thus qeðyÞD: With
probability 1 À qeðyÞD no deal occurs, and the entrepreneur continues the search.
The expected discounted utility from searching is therefore
U o ¼ qeðyÞD expðÀrDÞUd þ ð1 À qeðyÞDÞexpðÀrDÞUo:
4 In the search literature, our frameworkis commonly known as the Diamond-Mortensen-Pissarides
model (Pissarides, 1990).
5 Frictions are thus expressed as costs of delay. The model can be easily extended to include search costs.
For convenience, we assume that both sides use the same discount rate.

---

ARTICLE IN PRESS
R. Inderst, H.M. M .uller / Journal of Financial Economics 72 (2004) 319–356
327
Solving for U o and letting D-0; we obtain
q
U o ¼
eðyÞ
U d:
ð2Þ
qeðyÞ þ r
Eq. (2) illustrates the relation between the entrepreneur’s overall utility U o and his
deal utility U d: The overall utility is the discounted expected utility from searching
or, alternatively, the utility from a deal minus the expected cost of delay.
Accordingly, it holds that U ooUd: Moreover, by (2) the difference between Ud
and U o is smaller the smaller is the discount rate r and the greater is the speed of
matching qeðyÞ: Rearranging (2) yields the asset value equation
rU o ¼ qeðyÞðUd À UoÞ:
ð3Þ
Similarly, the venture capitalist’s outside option V o is given by the asset value
equation
rV o ¼ qvðyÞðV d À V oÞ;
ð4Þ
which implies that the venture capitalist can invest funds at the interest rate r while
searching for an investment opportunity.
To close the model, we need to specify what the in?ows and out?ows are.
Stationarity requires that the in?ow of venture capitalists and entrepreneurs matches
their respective out?ow. Let mv and me denote the measure of venture capitalists and
entrepreneurs arriving in the market over one unit of time. The in?ows mv and me are
fully, and uniquely, determined by the stationarity conditions mv ¼ qvðyÞMv and
me ¼ qeðyÞMe; respectively. Accordingly, the model is fully pinned down by its
stocks: The stocks Mv and Me determine (i) the level of capital market competition
y  Mv=Me; (ii) the size of the market, (iii) the out?ows qvðyÞMv and qeðyÞMe; (iv)
the in?ows mv and me; and (v) the equilibrium values of V d; Ud; V o; Uo; and s (see
Proposition 1 below).
2.4. Equilibrium conditions
The following de?nition summarizes the equilibrium conditions.
Capital market equilibrium. An equilibrium is characterized by the following
conditions:
(i) the deal utilities ðV d; U dÞ maximize the Nash product ½V À V o?Z½CðV Þ À U o?1ÀZ;
(ii) the outside options ðV o; U oÞ satisfy the asset value equations (3)–(4); and
(iii) the ?ows and stocks of entrepreneurs and venture capitalists, ðme; mvÞ and
ðMe; MvÞ; satisfy the stationarity conditions qvðyÞMv ¼ mv and qeðyÞMe ¼ me:
The equilibrium is the solution to a system of four equations: the ?rst-order
condition characterizing the bargaining solution (1), the asset value equations (3)
and (4), and the identity U d  CðV dÞ: In the bargaining solution, the deal utilities
V d and U d are a function of the outside options V o and U o: Conversely, in the
asset value equations, outside options are a function of the deal utilities. Inserting

---

ARTICLE IN PRESS
328
R. Inderst, H.M. M .uller / Journal of Financial Economics 72 (2004) 319–356
(3)–(4) into (1), we obtain
r þ q
V d
b
vðyÞ ¼ ÀC0ðVdÞ
;
ð5Þ
r þ qeðyÞ
CðV dÞ
which implicitly de?nes the equilibrium value of V d as a function of y: Inserting the
solution in U d ¼ CðV dÞ yields the equilibrium value of U d: Finally, inserting V d and
U d in the asset value equations (3)–(4) yields the equilibrium values of V o and Uo:
3. Value creation in start-ups
3.1. Equity shares, individual utilities, and value creation
The following proposition characterizes how individual utilities and the value
created in start-ups depend on the level of capital market competition.
Proposition 1. For each level of capital market competition y there exists a unique
equilibrium. The venture capitalist’s equity share s; deal utility V d; and overall utility
V o are all decreasing in y: The reverse is true for the entrepreneur. The total value
created in the start-up V d þ U d is ?rst increasing and then decreasing in y:
Deal utilities and overall utilities move in the same direction. Consider, for
instance, the entrepreneur. An increase in capital market competition makes it easier
for the entrepreneur to obtain ?nancing, which reduces his cost of delay. The
entrepreneur’s outside option Uo therefore increases (and the venture capitalist’s
outside option V o decreases), which implies that the bargaining outcome shifts in
favor of the entrepreneur. Consequently, the entrepreneur’s deal utility U d increases
and the venture capitalist’s deal utility decreases. The increase in U d; in turn, feeds
backinto the search market dynamics. As the utility from doing a deal has gone up,
searching for a deal becomes more valuable. The overall utility U o therefore
increases again, and so on. This process continues until a steady-state equilibrium is
reached. Consequently, an increase in y corresponds to a move along the contract
frontier from the right to the left.
The rest follows from the construction of the contract frontier. As we move along
the frontier counterclockwise, the venture capitalist’s equity share s decreases. This
weakens the venture capitalist’s incentives and improves the entrepreneur’s
incentives. The total effect depends on the current level of s and its relation to the
ef?cient sharing rule #s: If s > #s; a reduction in s increases the total value created in the
start-up V d þ U d: If s ¼ #s; the value created in the start-up attains its maximum.
Finally, if so#s; a reduction in s decreases the total value created in the start-up.
3.2. Pre- and post-money valuation
A measure of the ?rm’s value commonly used in the industry is the post-money
valuation. The post-money valuation is an implied value calculated on the basis of

---

Document Outline

• The effect of capital market characteristics on the value of start-up firms
  • Introduction
  • The model
    • Financial contracting
    • Bargaining
    • Search
    • Equilibrium conditions
  • Value creation in start-ups
    • Equity shares, individual utilities, and value creation
    • Pre- and post-money valuation
    • Market value and success probability
  • Industry dynamics
    • Endogenous entry of venture capitalists
    • Changes in investment profitability
      • Expansion of the equity (and contract) frontier
      • Short-run analysis
      • Long-run analysis
      • Value and valuation
      • Internet boom and bust
      • Speed of matching
    • Changes in entry costs
    • Changes in capital market transparency
      • Short-run analysis
      • Long-run analysis
  • Discussion and robustness
    • Safe project payoff
      • Construction of the contract frontier
      • Equilibrium analysis
    • Wage (or transfer) payments
    • Endogenous entry of venture capitalists and entrepreneurs
    • Constant-returns-to-scale matching technology
    • Welfare and policy implications
  • Project screening
  • Empirical implications
    • Venture capitalists? equity shares
    • Pre- and post-money valuation of start-ups
    • Total value created, market value, and success probability of start-ups
    • Search time
    • Project screening
  • Concluding remarks
  • (Well-behaved) production technologies
  • Proofs
  • References

---