Unformatted text preview: ceiving
nothing in this event. Otherwise it can reﬁnance the loan for one more period, injecting an additional unit of money into the operation. In this event, with probability p it generates Rp
for the bank and with probability (1 − p) it fails, generating a liquidation value Lp < L. The
entrepreneur receives private beneﬁts of Bp in both cases. If the second period re-ﬁnanced project fails, the bank collapses if not bailed-out as we assume its net position is then negative
(more on this below). To keep matters simple, we assign all the bargaining power to the bank
so that the bank is able to extract all the monetary returns from ﬁnancing the projects and
entrepreneurs are limited to private beneﬁts.
In order to formalize the idea of bank moral hazard, that is, the idea that banks may be
passive in their lending practices with the expectation that the government will bail them out
if their projects turn out poorly, we graft a game between the government and the bank on
top of this game between the bank and the entrepreneur. Suppose the government can decide
to monitor the bank to ascertain whether the bank liquidates bad projects or remains passive.
4 The government does this by choosing ex-ante, in the ﬁrst period, on a detection probability γ
that involves a cost c(γ ), c > 0, c > 0. Detection takes place after the reﬁnancing stage, and
if detected the management of the bank is ﬁred and the bank is recapitalized (bailed-out) to
the extent of its negative net position. In other words, by the time bank passivity has been
detected by the government, the damage has already been done; it is too late to undo the banks
The timing of the game is as follows. In period 1 the government decides on γ and incurs
the monitoring cost c(γ ). The bank lends to ﬁnance projects, and entrepreneurs with poor
projects decide on their eﬀort level. In period 2, returns are observed and the bank decides whether to liquidate or not, by choosing a level of activity a ∈ [0, 1]. Immediately afterward
the government monitors and ﬁres passive bank managers, in which case they get nothing, and
recapitalizes the bank if it has a negative net position. The timing and structure of the game
is depicted in Figure 1.
[Insert Figure 1 here]
There are thus two layers of soft-budget constraint problems in this environment, one at the
level of the government-bank interaction and the other at the level of the bank-ﬁrm interaction.
As we will see, a soft-budget constraint at the ﬁrst level can be transmitted to the next level.
We analyze the model by applying standard backward induction techniques.
We start by asking what will be the eﬀort choice of the entrepreneur with the bad project.
In period 2, if the bank re-ﬁnances the project, the entrepreneur receives Bp . This happens
with probability (1 − a). Thus the entrepreneur shirks if (1 − a)Bp ≥ Bg . In other words, ﬁrms have hard budget constraints only if the bank is suﬃciently active; t...
View Full Document
- Fall '13