P. Z. LIN
b) If bank provided capital for SME and the enterprise re-
turned the loan repayment in time, then the situation would be
win-win cooperation. The revenue collection is (+1,+1);
c) If bank provided capital for SME, the enterprise refused to
return the loan repayment and bank borne loss. Then the enter-
prise obtained interests exclusively and bank lost its principal.
The revenue collection is (+2,−1);
d) If bank provided capital for SME, the enterprise refused to
return the loan repayment and bank dun to enterprise. At this
point, the inefficient judicial system would make the company
still gain interests of 1.5, and the bank’s interests would be
protected to some extent as 0.5. The revenue collection is (+1.5,
+0.5).
Examine the pattern of this game between the bank and the
enterprise, we will find that the original equilibrium is broken.
The new Nash equilibrium should be (0,0). That is the banks
refused to offer money from the very beginning.
This phenomenon is not uncommon in real life. As a result of
Chinese judicial system’s inefficiency when it duns for ar-
rearage, bank can get back little money but pays high transac-
tion cost. In other words, the desired effect cannot be achieved
when bank pursuits debt recovery through the judicial system.
The expected return of dunning to enterprise, for the bank, is
less than debt recovery cost in this case. Then the bank tends to
not pursue the matter, and make its threat become incredible.
Therefore, when the company knows the bank's rational choice,
its best option is “deadbeat”. Thus, in order to avoid losses (−1),
banks would choose not to trust SMEs and retain principal
when they know the company would refuse to repay the debts.
The best choice of banks is not offer capital to SMEs. It is the
so called “credit crunch” when banks expected to lose their prin-
cipal. In this circumstance, Nash equilibrium of the dynamic
game between the bank and the SME becomes (0,0). At this
time, in order to guarantee its principal safe, the bank would
like to invest in large enterprises, who own less credit risk,
instead of SME (Lu, 2005). This is why SME loans accounted
for a relatively small proportion in total corporate loans.
Equilibrium Analysis
Through the above analysis of game, we know that bank
could not recover its lost totally in time when the enterprise
chose to break the contract. The bank inclined to select “credit
crunch”. Thus, SMEs cannot gain enough funds to obtain their
own development. At the meantime, standard financial system
has not been set up generally in SMEs, which makes banks face
extremely high costs of searching for business information of
SMEs. The small scale of most SMEs financing creates the
small income of bank. The bank’s cost-profit ratio (profits/costs,
costs include the cost of searching for information and the am-
ortized cost of non-performing loans of SMEs) from SMEs is
much lower than the one from large enterprises. Therefore,
bank credit funds flow to the large enterprises eventually.
Thence, standard financial system should be set up in SMEs to
overcome the information asymmetry in the process of SME
financing on the one side, and effective judicial system should
be established in society to increase the costs of breaking con-
tract on the other hand. Increasing bank’s cost-profit ratio from
SMEs would make part of bank credit funds flow from the
large enterprises to the SMEs. This process would not stop until
the two cost-profit ratios are equal. That is capital liquidity
maintains a balanced state.
A Dynamic Game of Incomplete Information
The Game’s Basic Assumptions
1) The players: banks, small and medium-sized enterprises.
SMEs are divided into two types: enterprises with good eco-
nomic efficiency and ones with poor economic efficiency. The
good ones can repay all the bank loans on time, whereas the
poor ones would make banks suffer losses.
2) The revenue of bank from offering capital is set to Rg, and
. The revenue of bank from offering capital to enter-
prises with good economic efficiency is set to R1, ; the
interests of enterprises who choose repudiation is set to R2,
; the revenue of bank from offering capital to enter-
prises with poor economic efficiency is set to Rb,
Rg 0
R2 0
R1 0
Rb 0
.
The Process of Game Development
1) Whether the SMEs apply for loans from banks or not in
this game.
Game tree can be used to represent this incomplete informa-
tion game. The second layer of this decision tree contains two
decision nodes, which means that information of business
situation owned by bank is incomplete. In other words, the
bank does not know which path reflects the true operating
situation of the SMEs. Then banks cannot make decision sepa-
rately according to the two different cases. There are three pos-
sibilities for the banks on the promise of the enterprises apply
for a loan. If banks choose to accept the application, they may
get a profit (enterprises with good economic efficiency) and
may suffer a loss (enterprises with bad economic efficiency). If
the banks choose not to accept the application, they would miss
the opportunity of gain. As a result, the banks need to assess the
probabilities of enterprises’ efficiency before they make the
decision.
R2 0
, means that enterprises with poor economic effi-
ciency can gain benefits through defrauding banks of funds.
Therefore, it is easy to know that enterprises will apply for a
loan whether they are in good or poor management state.
Whether the banks accept the SMEs’ application for loans or
not in this game.
On the promise of enterprises apply for a loan, we use
PGA stands for the conditional probability of enterprises
with good economic efficiency and
PBA
stands for the
conditional probability of the ones with bad economic effi-
ciency.
PG and
PB stand for the probabilities of enter-
prises in good and in bad management state respectively (both
probabilities only can be estimated by empirical data or general
knowledge from the average case).
PAG
and
PAB
stand for the conditional probabilities of enterprises in good and
in bad management state respectively when they apply for loans.
Based on the above analysis, we know:
PAGPAB 1
According to Bayes’ rule:
PG|APGPA|G PGPA|G PBPA|B
From
PBA1 PGA , it is easy to find
PBA .
PGA
PGPAGPGPAG PBPAB
PG PG PBPG
Copyright © 2012 SciRes. 65