3 Result
We solve this game by backward induction. Given a license scheme (w, f ), the licensees
compete in quantities in the third stage. The profit maximization problem of each
licensee i (= 1, ..., n) is as follows:
max qi(P(Q) - w - c) - f. (1)
Qi
The first-order condition of each licensees i (= 1, ..., n) is given by
P(Q) - c - w + P,qi = 0. (2)
Note that since we assume that P'(Q) + QP"(Q) < 0, the second-order condition is
satisfied.
In the second stage, licensees enter the market as long as they can obtain positive
profits. Thus, we have the following zero-profit condition for each licensee i (= 1, ..., n):
qi(P(Q) - w - c) - f = 0. (3)
In what follows, we focus on the symmetric equilibrium where all the licensees choose
the same strategies, i.e., qi = q for all i (= 1,...,n). Now we have the equilibrium
conditions in the second and third stages as follows:
P(Q) - c - w + P'q = 0, (4)
q(P(Q) - c - w) - f = 0. (5)
Given w and f, let q(w, f ) and n(w, f ) be the solutions to (4) and (5). Therefore, from
equations (4) and (5), the implicit function theorem implies the following7:
P ' q + q2P '' P ' + P 'n + qP ''n q2P' (P - w - c) + nqP' |
dn -dw - |
= |
T .q. |
, | |
and |
P 'q + q2P '' P 'n + P ' + qP ''n |
dn |
= |
o’ 1 |
. |
This yields: | |||||
dn 1 (P - c - w - qP') - nq2P'' |
(6) (7) (8) . (9) | ||||
dw = ^ (P - c - w)(P' + qP'') - q(P') |
2 , | ||||
dw (-(P - c - w)q(P' + P''q)) + q2(P')2 , dn P ' + nqP '' + nP ' | |||||
df = (-(P - c - w)q(P' + P''q)) + q2(P')2 , | |||||
df (-(P - c - w)q(P' + P''q)) + q2(P |
')2 |
7Since the determinant is (P - c - w)(P' + wP") - q(P')2 < 0(≠ 0), the implicit function theorem
can be applied.