Transcript Joseph Stiglitz-Powerpoint presentation

MONETARY POLICY IN A
MULTIPOLAR WORLD
J. E. Stiglitz
Washington, DC
October 8, 2013
Motivation



Monetary policy by one large country can have
affects on other countries
Other countries are likely to respond
Key questions:
 How
to think about the resulting global equilibrium?
 How could global coordination improve matters?
 In the absence of (perfect) coordination, how can
global financial architecture (the rules of the game)
improve matters?
Example: QE2

Probably had only minimal effects on US




But many other countries believe it had adverse effects on them, as liquidity
stimulated their economies



Effects on LT government rates small
Effects on private sector lending rates even smaller
Has not led to much more lending to only sector which is constrained, SME
Money is going where it’s not needed, and not going where it’s needed
Especially with credit channel in US still clogged
These other countries offset the actions


Imposing capital controls
Buying dollars to keep their exchange rates from appreciation



Fed “sold” dollars, other CB “bought” dollars
Does such a move have much effect on US, other countries, global economy?
Monetary policy in a world of globalization, and several “large” players
may be markedly different in the closed economy
Second Motivation


To move away from “perfect markets” models which
have dominated monetary policy debate
Towards models in which
 Credit
availability is important
 Distributive consequences are important
 The difference between T-bill rates and lending rates
(to businesses) is endogenous and important
 The market equilibrium, on its own, is not in general
Pareto efficient
Outline


Monetary policy in a limiting case—money issued
by two monetary authorities are perfect substitutes
Variable exchange rates
 “perfect”
capital markets
 credit constraints

Why monetary policy has any effects
Monetary policy in a limiting case
Single country with two CB, E and W
 Economies in two regions imperfectly correlated
 Two forms of money, ME and Mw , are perfect
substitutes
 Utility functions Vi (M, T)
Where M = ME + MW total money supply
T is transfer paid by one region to other

Nash equilibrium (no cooperation)
T=0
 VE ‘(ME + MW , 0) = 0, VW’(ME + MW , 0) = 0
Note effects of policy in East (in recession) on West (in
boom)

 Spill-overs
from trade (exports, imports), from financial
flows, from commodity prices
Pareto Optimum
Max VE(M, T) + VW(M,-T)
Generating f.o.c.
(3A) VEM + VWM = 0
(3B) VET = VWT

Nash equilibrium is not, in general, Pareto Optimal
Note: “Independence” of central banks makes it
difficult to achieve PO, because of the absence of
compensatory fiscal instruments
Additional instruments can reduce
externalities
Vi = V(M, T, αi )
Where
αi is a vector of other variables
Nash equilibrium
Viαi = 0
A coordinated solution would be preferable:
VEαi + VWαi = 0
Even if compensatory payments are not allowed,
cooperative agreements can be reached, i.e.
denoting the Nash equilibrium by {Mi*, αi*} , there
exists an alternative policy vector {Mi**, αi**} which
is Pareto Superior, where each party shifts more of
its policy agenda towards variables that have
smaller spillovers
Variable exchange rate
Additional channel through monetary policy exerts its
affects—with strong externalities on other country
(countries)
Slight generalization of earlier model:
Vi(Mi, Mj, αi, αj, βi, βj)
Where βi are variables that affect spillovers (in or
out), like capital control

Note: unlike trade liberalization, there is neither
theory or evidence that capital and financial
market liberalization is necessarily welfare
enhancing to both countries
 And
even if it were, there could be strong distributive
consequences that are not easily offset
 Hence, there are strong distributive consequences across
countries resulting from these liberalization measures
 Explaining
up markets
why many countries have to be “forced” to open
Nash Equilibrium
ViMi = Vi αi = Vi β i= 0
Not Pareto Optimal
Restricting externality reducing actions (like capital
controls) makes things worse
Credit availability
In the real world, what matters as much or more than
interest rate is the availability of credit.
High r can lead to an influx of capital, increasing
credit availability, L(r), L’ > 0
Assume potential output is Y*. Actual output is
Y(L(r))
Optimal r solves
Y(L(r*)) = Y*
Demand side shock
Yd = Yd(L(r), ε )
Raising r increases capital inflow, exacerbates
inflationary pressure
Should use additional instruments. Assume
V (Yd(L(r, α)) , α, r)
Optimal policy entails
VY YL Lr + Vr = 0
VY YL L α + V α = 0

Optimal policy
1)
2)
Takes account of the effect of r directly on welfare
(e.g. through distributive effects)
Can be improved upon by adjusting regulations
(reserve requirements, capital adequacy
regulations, lending standards)
Generalization:
Composition of output also matters





Some sectors (1) more associated with instability (real
estate)
Some (2) have greater technological spillovers/learning
benefits
Some more dependent on capital inflows (real estate),
affected by restrictions on cross border flows (β)
Some more affected by restrictions on domestic legislation
Increase in r increases flow of funds to real estate, reduces
flow of funds from domestic banks to other sector

Capital inflows limit effectiveness of monetary policy through
traditional channels
Welfare maximization
V(Yd1(βL1 (r, α), r) , Yd2 (L2 (r, α),r) α, β,r)
F.O.C.
VY1 Yd1L L1 + Vβ = 0
VY1 Yd1L β L1α + VY2 Yd2L L2α + V α = 0
VY1 [Yd1L β L1r + Yd1r ] + VY2 [Yd2L L2r + Yd2r ]+
V r = 0.
Implications
If an increase in r leads to an expansion of
unproductive sector and contraction of productive
sector, we set r lower than we otherwise would
(tolerate more inflation)
 If we can restrict capital inflows, we should
 If inflationary pressures related to sum of outputs,
Yd1 + Yd2 = Y*,
Any inflation target an be achieved by large number
of policies {α , β, r}. Choose the one which
optimizes sectoral composition

Optimal to use multiple instruments


Increasing reserve requirements and lowering
interest rates (or raising them less than they
otherwise would be) may be preferable to just
raising interest rates
But imposing controls (taxes) on the inflow of capital
may be still preferable
General point:
Pervasiveness of macro-externalities

With credit rationing, collateral, incentive compatibility,
self-selection constraints (and/or imperfect risk
markets), economy is essentially never constrained
Pareto efficient (Greenwald-Stiglitz)


Those with access to international capital markets borrow
excessively
Public policy should be directed at “correcting” these
market failures
No presumption that markets, on their own, are efficient
 No presumption that market based interventions (“r”) are
the best set of interventions

Some general observations about
monetary policy

Generalized MM theorem predicts that monetary
policy shouldn’t have much effect
 Especially
with CMA accounts allowing using, in effect, T
bills for transactions purposes

But monetary policy has some effects: Why?
 Market
imperfections (capital constraints)
 Institutional features
 Distributive consequences
 Market “irrationality” (not seeing through public veil)

Modern monetary theory lives in a half-way house
of incompletely articulated assumptions of
imprecisely defined market imperfections and
distributive effects, leading to speculative
observations about possible channels through which
monetary policy might yield effects, with
ambiguous quantitative significance .
Examples

Banks are central—mediate flow of credit to SME’s
 Focus
should be understanding bank behavior
 “Liquidity trap”—circumstances in which monetary
policy does not lead to more lending
 Distinctively
different from Keynesian liquidity trap
 Note that in Great Depression, real interest rate was
greater than 10%, now it is -2%.
 No evidence that zero lower bound is critical constraint—
would lowering real interest rate to -3% make any real
difference?
Temporary interventions




With infinitely lived rational individuals, without
capital constraints why should they have any effect?
With life cycle model, elderly with stock will
consume more
But elderly depending on interest payments will
consume less
Distributive consequences are crucial
Market imperfections help explain
ineffectiveness of QE

Domestic lending channel blocked (especially to
SME’s)
 Community
and regional banks still weak—
disproportionately responsible for SME lending
 SME lending collateral based; collateral real estate;
real estate prices still markedly down
 Refinancing limited
 Banking
sector concentrated—incentive not to increase
lending (take lower government rate as profit)
 Large fraction of homeowners who are not underwater have
already refinanced
Insolvency vs. Illiquidity


A central distinction in Central Bank doctrine
But is it totally persuasive?
 In
a world with perfect information, a firm that was
solvent would presumably have access to funds
 Of course, every bank will believe the market has
“misjudged” its future prospects
 But why should a Central Banker believe that his
judgments are better than that of market, when no one
is willing to supply funds to the bank?
Conclusions


In a world of truly free capital mobility, the effects of
monetary policy are different (typically weaker) than in
a closed economy
Restrictions on capital flows may enhance the ability of
the government to maintain the economy near full
employment
Macro-benefits more than offset micro-distortions
 But in more properly formulated model, no presumption that
markets on their own are efficient


Cooperation may lead to Pareto improvements

More likely to be true—more likely to be able to obtain
cooperation—if there are more instruments

Restrictions on cross-border capital flows and other instruments can
reduce cross-border externalities, improve efficiency, and mitigate
distributive consequences


Strong presumption that markets are not Pareto efficient
Strong presumption that optimality requires using a panoply of
instruments


Single-minded focus on interest rate has been foolish
Governments/monetary authorities should be concerned about the
structure of the economy and the distribution of income


If so, they should use a multitude of instruments, not just interest rates
And engage in unorthodox policies

To respond to a potential influx of distorting capital from abroad