The Bad Policy View

1
The Bad Policy View
2
Amending the First Generation Approach

• Main challenge to First Generation models to
  explain recent emerging markets crises the
  fiscal imbalances that are crucial for that
  approach seemed to be absent
• But perhaps one can argue that fiscal losses,
  although existing and sizeable, were not evident
  from conventional data

3
The Contingent Liabilities argument

• Dooley, McKinnon Governments in Asia provided
  implicit guarantees to private borrowers
• This encouraged excessive risk taking, capital
  flight, even , which resulted in cumulative
  losses that could be hidden for a while

4

• Crises occur when the accumulated losses exceed
  the amount of resources that the government is
  willing to use to bail out debtors. Then
  investors cash out and exchange their claims on
  (now worthless) ventures for the government
  resources.

5

• Financial liberalization, which preceded many of
  the crises episodes, encouraged this pattern
  because it was not accompanied by adequate
  improvements in supervision and regulation
• International financial assistance only
  exacerbated the problem

6
Assessment

• This view sounds plausible, and accounts for
  several aspects of recent crises the fact that
  fiscal losses were not evident, and the rapid
  growth in domestic credit in crises countries

7
Serious Issues

• By assuming that fiscal losses are not
  observable, the bad policy view is hard if not
  impossible to evaluate
• The data we do have does not suggest that less
  transparent countries were more prone to crises

8
(No Transcript)
9

• Also, why would the government choose the kinds
  of policies assumed?
• Why, in particular, would the domestic government
  choose to bail out foreign investors in a crisis?

10

• How is the bad policy view reconciled with the
  long period of prosperity in Asian tiger
  countries?
• How is that view consistent with the rapid
  recovery after crises?

11

• Finally, this view has yet to be articulated in a
  convincing way
• The recent work by Tirole goes some way in that
  direction

12
The Dual and Common Agency Approach
13
Main Idea

• Tirole A key aspect of international financial
  structure is that lender-borrower relations
  depend on the actions of the government, which
  private parties take as given
• Financial structure, in turn, affects the
  governments decision problem

14
Questions this view can address

• Level of Borrowing is there too much
  international capital mobility?
• Structure of borrowing debt vs equity, short
  term vs long term, domestic currency vs foreign
  currency denominated

15
A Model of Inefficient Borrowing Level

• Simplest setup three dates (t 0,1,2), a large
  number of domestic entrepreneurs, and a large
  number of foreign investors with opportunity cost
  of funds zero. All agents are risk neutral
• t 0 entrepreneur borrows I from foreign lenders

16

• t 1 government chooses an action a
• t 2 proceeds from investment are obtained and
  distributed between lenders and borrowers
• Let V(I , a) expected surplus created by
  investment, and W(I , a) expected amount paid to
  foreigners (chosen at t 0)

17

• The government maximizes the welfare of the
  domestic entrepreneur. However, it chooses a
  after I is set, which creates a time
  inconsistency problem, and a potential
  inefficiency

18
First Best (Commitment)

• Maximize
• U(I, a) V(I, a) - W(I, a)
• s.t. W(I, a) I
• Solution
• VI(IFB, aFB) 1 (Mg. benef. of I MC)
• Va(IFB, aFB) 0 (policy maximizes value)

19
No Precommitment

• In equilibrium, the policy choice a is
  anticipated by investors, hence
• W(I, a) I
• At t 1, I is given and a maximizes
• U(I, a) V(I, a) - W(I, a)
• ? Va(I,a) Wa(I,a)

20

• Tirole assume that a is an investor friendly
  policy, so Wa(I,a) gt 0. It follows that
• Va(I,a) gt 0
• that is, the policy a fails to maximize the value
  of investment

21
Is there too little or too much I?

• The condition Va(I,a) Wa(I,a) is a
  reaction function, and yields a a(I). In
  equilibrium, domestic welfare is
• V(I,a)-I
• The impact of a small (forced) increase in I
  would be
• Va . a(I)

22

• This is negative if WaI gt VaI
• In such a case, a small capital control raises
  welfare. This is because of a commitment
  effect a reduction in I leads to an increase
  in a, which is socially desirable
• T plausible view, but so is the opposite

23
Credit Rationing

• Now introduce credit rationing there is a
  maximum amount of income that can be pledged to
  outsiders, call it W(I,a)
• WI lt 1
• Credit rationing means that firms borrow as much
  as they can and would borrow more if they could
• W(I,a) I
• VI(I,a) gt 1

24

• Now the effect of an policy that forces an
  increase in I is
• VI(I,a) 1 Va . a(I)
• The first term is a direct effect due to the fact
  that, at the margin, an additional unit of
  investment has a positive net value to the
  country, due to credit rationing

25

• Ts conclusion even if capital controls are
  desirable because of commitment effects, credit
  rationing goes in the opposite direction.
• Overall, there is no strong presumption that this
  model yields too much foreign investment.

26
Externalities and Borrowing Structure

• A representative entrepreneur risk neutral,
  limited liability, has initial wealth A, invests
  I gt A (I is a choice variable)
• Domestic savers provide funding F (fixed)
• Foreign investors provide funds X at interest
  rate 0.

27

• So, at t 0, total investment is
• I A F X
• At t 2, the project may fail, yielding nothing.
  If success, the yield is RI, but only an amount
  rI is pledgeable to outsiders

28
(Why r may be less than R)

• Moral Hazard suppose that the probability of
  success drops from pt to qt , q lt p, if the
  entrepreneur does not exert effort. Effort costs
  BI.
• For the entrepreneur to exert effort, his payoff
  if success, say PI, must be large enough
• (pt) PI (qt)PI BI ? P B/(p-q) R - r

29

• The probability of success is p t, where t is
  chosen by the government at t 1, and costs g(t)
  to domestic residents.
• In terms of previous model, W (p t)rI

30
Optimal Commitment Policy

• Choose I, t to maximize
• (p t)R g(t) 1 I
• s.t. I A (p t)rI
• The latter is the credit rationing constraint and
  implies that investment is a function I(t) of
  (expected) government policy

31

• The solution requires
• g(tSB) gt R
• That is, the marginal cost of t exceeds its
  direct marginal benefit on expected value. This
  is because it also affects total investment, and
  there is credit rationing

32
Equilibrium without commitment

• Investment is given by credit rationing
  constraint I A (p t)rI, or
• I I(t) A/(1- (p t)r)
• But now t must maximize domestic welfare at t
  1, taking I as given
• (p t)(R-r) r (F/(XF)) g(t) I

33

• The solution yields
• g(t) R r (X/(XF))
• lt R lt g(tsb )
• ? t is too low, for two reasons dynamic
  inconsistency problem, and the fact that t
  benefits foreigners, whose welfare is not
  included in government objectives

34
Implications

• An increase in domestic savings (F) reduces
  policy inefficiency (and increases foreign
  investment)
• Consider a tax t on foreign investment at t 0.
  Domestic savings are fixed at F, so the tax
  raises their return to (1 t). The break even
  constraint is now
• (1 t) (I A) (p t)rI

35

• One can solve the model as before, and derive the
  effect of a marginal increase in t on domestic
  welfare
• (p t)R I g(dI/dt)
• (R-r)I dI/dt (dt/dt)
• The tax reduces I, which lowers welfare, but
  increases t, which is beneficial. The latter may
  dominate if foreign investment is large

36

• For the same reason, an increase in r has
  ambiguous effects on welfare. In this sense,
  better corporate governance has a negative
  impact.

37
Debt vs Equity

• Assume that, in addition to the random return R,
  investment I yields PI for sure at t 2
• It is optimal for the firm to sell PI at t 0,
  as safe debt. (The rest is sold as equity)

38

• Let aE and ad be the fraction of equity and debt
  owned by foreigners. Then the break even
  constraints are
• X aE (p t)R ad P I
• F (1-aE) (p t)R (1-ad)P I
• which imply an overall constraint
• I - A (p t)R P I
• ? Given t investors are indifferent between debt
  and equity as long as the above hold

39

• However, the governments choice is affected by
  the foreign equity-debt composition t maximizes
• (p t)(R-r) (1-aE)r g(t) I
• g(t) (R-r) (1-aE)r
• So, an increase in foreign equity holdings
  reduces government discipline and country welfare!

40
Liability Dollarization

• Now assume that there are two goods, a tradable
  one (dollars) and a nontradable one (pesos)
• Foreigners only value dollars
• Domestic residents have preferences
• c u(c) v(g)
• where c and c denote consumption, and g a
  public good expenditure, all at t 2

41

• For simplicity, assume credit rationing away, and
  no domestic savings
• At t 0, domestic agents invest I dollars to
  obtain y(I) pesos at t 2.
• They borrow I from foreigners, who are promised
  a repayment of d dollars and d pesos at t 2.
  If e real x-rate (pesos per dollar), the break
  even constraint is
• d d/e I

42

• How is the exchange rate determined? At t 2,
  domestic residents will maximize utility subject
  to the budget constraint
• c ec income in pesos
• Maximization requires
• e u(c) (show this)
• So a fall in dollar consumption will induce a
  real devaluation

43
Government Policy

• The government is assumed to own R dollars
  (reserves), spends g dollars at t 2, and
  rebates the remainder to domestic residents as a
  lump sum
• Hence income in pesos is
• y(I) e(R-g) d ed
• c ec as already mentioned

44

• But in equilibrium,
• c y(I)
• So
• c d d/e R - g
• (demand for dollars supply of dollars)

45

• Assume g is chosen at t 1, so it maximizes c
  u(c) v(g) s.t.
• e u(c)
• c y(I)
• c d d/e R - g
• Crucially, I, d, and d are taken as given here

46

• The FOC for the preceding includes
• v(g) u(c) dc/dg
• (Exercise Complete equilibrium description)
• Note that investors are indifferent as to
  currency composition, so there is a continuum of
  equilibria

47
Commitment Policy

• Suppose that the government can choose g at t
  0, before contracts are written
• The appropriate resource constraints are
• c y(I)
• c R - g - I (why?)
• Government chooses g, I to maximize
• c u(c) v(g) subject to the above

48

• The solution requires
• u(c) y(I) v(g) e
• This will coincide with the equilibrium outcome
  only if dc/dg 1
• This requires d 0 i.e. all debt to be in
  dollars! (show why)

49

• So, of the continuum of equilibria when there is
  no government precommitment, the best one for
  this country entails only dollar debt
• The key to this argument ex post, peso debt
  creates an incentive for the government to
  devalue, in order to expropriate foreigners.

50
Recap

• Before concluding that some forms of finance are
  dangerous and must be reformed, it may be crucial
  to understand why private agents choose to
  structure contracts in that way
• In Tiroles discussion, heavy emphasis on
  government behavior
• Multiplicity of Equilibria is not averted

51

• Tiroles approach is yet to be extended to
  explain crises, only the inefficiency of
  international finance