the product market in equilibrium.16 The optimal anti-competitive contract
commits the investor to fund only the entrants that the second investor would
also fund. As in the present paper, this commitment can only be achieved by
taking a riskier claim in the incumbent firm.
These points can easily be understood with the aid of the following simple
example.17 First notice that if Firm 2 could turn to a perfectly competing
source of funds, it would always enter the market; therefore, the first investor
would never bother to take (costly) equity in Firm 1, as this would not prevent
competition in the product market. Therefore suppose that the second investor
competes imperfectly with the first investor, because he has a higher cost of
funds: r2 >r1 = 1.OnceFirm1 has been funded by Investor 1, the second
investor will fund Firm 2 if and only if V2 - (r2 - r1)I2 ≥ 0. Then two possible
cases may arise. If 0 <V2 < (r2 - r1)I2 there is no competition to fund Firm 2,
and the first investor can prevent competition in the product market by simply
committing himself not to fund Firm 2. In this first case the Coase problem is
an issue and anti-competitive equity financing arises as in Proposition 3 above.
If instead V2 ≥ (r2 - r1)I2, then Firm 2 can always turn to the second investor
if denied funding by the first investor. Anticipating this, the first investor will
itself fund Firm 2 and take a debt claim in Firm 1.
The above example, though incomplete, illustrates two important points.
First, when the credit market is imperfectly competitive, anti-competitive goals
have the same qualitative impact on financial contracts as in the monopoly
case: investors take high-powered claims when they aim to deter entry (i.e.
when V2 is low enough for this to be feasible). Second, more competition in
the credit market spurs more intense competition in the product market (as
r2 becomes smaller, it is more likely that Firm 2 is funded in equilibrium).
This is because when competition in the credit market becomes more intense,
the efficient investor has less leeway in restraining entry in the downstream
industry.18
16For an analysis of bypass in a foreclosure setting, see Rey and Tirole (2000).
1 7 The following example is a simplification of our extended model. In the latter we assume
a continuum of potential entrants with different values V2 . Only those entrants with higher
values can afford to approach a less efficient investor, and thus are not subject to financial
entry deterrence.
1 8 The question of how increased credit market competition should affect the design of
financial contracts is more complex. Two conflicting effects are at work. On the one hand,
when the credit market becomes more competitive (r2 becomes smaller relative to r1 ), the
Coase problem arises ‘less frequently’, making the investor’s claim more low-powered or
debt-like. On the other hand, in the presence of competition, the reservation wages of the
firms become endogenous. A more efficient competing investor will also offer a better deal
to Firm 1, which has then a higher reservation utility W1 . This second effect means that the
first entrepreneur gets more rent and results in an increased payment for the entrepreneur
in the low state. As pointed out in section 4, a larger entrepreneurial stake then feeds back
18