10. When monetary policy becomes are only two assets: money and bonds.

Money pays a
ineffective: liquidity traps. zero (institutionally determined, hence exogenous) nom-
inal interest rate and bonds an endogenous nominal in-
A liquidity trap is a situation in which monetary policy terest rate. For the time being let us consider only the
becomes ineffective because the policymaker’s attempt (discrete-time) money market equilibrium condition
to influence nominal interest rates in the economy by Mt Yt
altering the nominal money supply is frustrated by pri- = , (1)
Pt 1 + it
vate agents’ willingness to accept any amount of money
which, using Fischer equation, can be rewritten as
at the current interest rate.
The traditional theory of the liquidity trap assumed Mt Yt Pt
= . (2)
that the LM curve becomes perfectly elastic at some P t (1 + rt ) P̄ t+1
level of the nominal interest. The modern reincarnation Note that I am assuming that individuals have perfect
of this theory spells out more carefully the conditions foresight so that Pt+1e
= P̄t+1 and that, at time t, Pt+1
that may generate a liquidity trap. which depends on what will happen from t + 1 onwards
A great deal on the current debate on liquidity traps is exogenous (i.e. I am not interested in modelling what
has been motivated by the prolonged Japanese depres- happens in the future).
sion and the apparent inability of the Bank of Japan to
do anything about it. Monetary policy (i.e. open market operations) alters
Let us first analyse the transmission mechanism of the relative supply of money and bonds. Suppose a
monetary policy. In our simple IS-LM-AS set up there money expansion policy: the central bank decreases the
supply of bonds and increases the supply of money by only goods prices and leaves real (but not nominal) asset
exactly the same amount. To induce private agents to returns unchanged. In the second case, the change in the
reallocate their portfolio from bonds to money the nomi- relative supply of assets affects real asset returns and,
nal interest rate (the opportunity cost of holding money) through this channel, real activity.
has to fall. The crucial point is: monetary policy affects the
Transmission mechanism: equilibrium if and only if it is able to alter
1. Flexible prices (Classical dichotomy holds). Yt = Y ∗ asset returns (be them nominal or real).
and rt = r∗. As the real interest rate is exogenous A liquidity trap is a situation in which monetary policy
a fall in the nominal interest rate requires a fall in cannot alter asset returns.
expected inflation P̄t+1/Pt. For given future price
If the statutory nominal return on money balances is
expectations, this is achieved by an increase in cur-
zero the economy is in a liquidity trap when the nominal
rent prices Pt.
interest rate on bonds is zero. The interest rate on bonds
2. Nominal rigidities. If current prices are sticky (as- cannot fall below zero (otherwise agents would be better
sume Pt is fixed throughout the period) and for given off not lending). If the nominal interest rate is zero,
future price expectations, expected inflation is fixed. agents are indifferent between holding money and bonds
The required fall in the nominal interest rate can only and will absorb any increase in the money supply at
be achieved through a change in the real interest rate unchanged asset returns.
rt.
In Japan short term interest rates have been very close
In the first case the transmission mechanism concerns to zero for a long time.
2
1 How can a liquidity trap happen? The above figure describes an economy in a liquidity
trap (note that the nominal interest rate features on the
Consider the following (discrete-time) model: vertical axis). At given current price level Pt (i.e. at
given expected inflation) the demand for goods is “low”,
LRAS Ȳ = F (L̄) (3)
µ ¶ in the sense that at zero nominal interest rate it falls
Pt short of full-employment output. In other words, saving
IS Yt = C(Yt − T̄t) + I (1 + it) + Ḡt (4)
P̄t+1 is too high with respect to investment. At the given level
Mt Yt of inflationary expectations P̄t+1/Pt, a negative nominal
LM = (5)
Pt 1 + it interest rate i∗ is required to clear the goods market.
Equation (3) describes the full employment level of out- A change in the level of the nominal money supply Mt
put. cannot move the economy to the right of point A, where
Y < Ȳ . Monetary policy is ineffective, since increasing
the money supply cannot push the nominal interest rate
i LRAS
below zero. To the right of point A the money market
IS LM equilibrium condition (4) is replaced by
it = 0. (6)
The LM coincides with the horizontal axis to the right
of point A.
A
-
Y Y
i*

3
 Since we know that it is the real interest rate that clears bound on the nominal interest rate it = 0, the real inter-
the goods market, let us consider what happens to the est rate. So even if the nominal interest rate hits its lower
real interest rate. bound at zero and monetary policy is ineffective, flexible
From Fischer equation prices ensure that the real interest rate takes whatever
value (negative if necessary) it takes for the goods mar-
P̄t+1
1 + it = (1 + rt) . (7) ket to be in equilibrium at the full-employment level of
Pt output. For example at point A current prices Pt would
So if the market-clearing i∗ < 0, it has to be that either decrease so that expected inflation is high enough1 for
r∗ < 0 or P̄t+1/Pt < 1 or both. For the economy to be the real interest rate to fall to the value that is consistent
in a liquidity trap either the goods market-clearing real with full employment. The fall in Pt and the increase in
interest rate is negative or agents expect a deflation or expected inflation shift the IS curve to the right until it
both. intersects the LRAS (point B in the figure below).
Does the Classical dichotomy fails when the economy is But what if the current price level Pt is sticky (assume
in a liquidity trap, i.e. when the market clearing nominal it is fixed)? Expected inflation is fully pinned down and
interest rate is constrained from the zero bound and the cannot increase to induce the required fall in the real
money market clearing condition (5) is replaced by (6). interest rate. At it = 0, the real interest rate cannot
No! Equations (3) and (4) still determine real variable. change and the IS curve determines the equilibrium level
For given P̄t+1, output Ȳ and it = 0, the IS curve still of output. The economy is stuck at point A at less than
determines the real interest rate. In fact it determines 1
Think about it, the current price level has to fall for inflation (the change in the
Pt. This pins down expected inflation and, for a given price level) between today and tomorrow to increase to the required level.
4
full employment until prices fall to reduce the real inter- asset market crash and fed by the current unwillingness
est rate. In the meantime the economy is in recession. of Japanese banks to lend have shifted the IS curve to the
LRAS
point where the equilibrium real interest rate in Japan
i

IS’
is negative. Furthermore, since prices have been falling
IS
LM’
for some time, expectations of more deflation may also
have set in, further preventing the real interest rate from
becoming negative.
A B
-
Y Y
3 How to get out of a liquidity trap?

Japan is in the most prolonged recession ever seen since The economy needs higher (goods) demand at any level
the Great Depression. Short term nominal interest rates of the interest rate (a rightward shift in the IS curve). If
are virtually zero and prices are falling. investment demand is to rise (that is ruling out changes
in taxes or government expenditure) then the economy
2 How does an economy end up in a liquidity trap? needs higher expected inflation for the real interest rate
to fall in the face of a zero bound on nominal rates. If
It has been argued that the fall in wealth associated with
current prices cannot fall the real interest rate cannot
the crashing of the 90s asset markets (stocks and land)
decrease by the amount necessary to clear markets.
bubble and the need to save for retirement in the face
Possible solutions (the IS has to shift):
of a shrinking population have depressed consumption.
This coupled with the fall in investment following the 1. The traditional solution proposed to exit a liquidity
5
 trap is an expansionary fiscal policy to increase ag- ity trap requires that the return on all assets which
gregate demand at any level of the real interest rate. can be used for such a purpose is zero.
The necessary increase in fiscal policy depends on In practice, there may be pretty compelling reasons
the size of the multiplier of government expenditure for not wanting a central bank or the government
and/or taxes. Japan has engineered a large fiscal ex- to conduct open market operations in shares as this
pansion over the past year, yet the result have not would be equivalent to nationalisation of private in-
been dramatic. We know that if agents are forward dustry.
looking consumption may not be much affected by
fiscal policy as individuals realize that taxes will go 3. Alternatively the required increase in expected infla-
up in the future to pay for current expenditure. tion can be achieved through a higher future price
Furthermore, these efforts have resulted in a dra- level. This requires a credible commitment by the
matic increase in the level of debt, inducing many Bank of Japan to create inflation, that is increase the
people to question the ability of Japan to continue future price level Pt+1. It has been suggested that the
with this policy without becoming insolvent. BoJ should therefore adopt an inflation target, that
is announce that it wants to keep the rate of inflation
2. Note that if there are more than two assets in the at a level high enough to engineer the necessary fall in
economy, monetary policy could still be effective as the real interest rate. The BoJ though has been ex-
long as there are assets which yield a positive return tremely reluctant to create inflation for fear of blem-
(e.g. long term bonds or stocks) which the central ishing its reputation for independence and inflation
bank can use for open-market operations. So a liquid- aversion (this is a blatant case in which discretion is
6
 better than rules). The final message is that getting out of a liquidity trap
may be extremely difficult since fiscal multipliers may be
4. There is an alternative possibility which involves tax-
small and because expectations may be difficult to alter
ing money. Up to now we have assumed that the ex-
(credibility). The best advice is to avoid ending up in
ogenous nominal return on money is zero. If a tax on
such a situation to start with.
money holdings is introduced, the opportunity cost
For this reason it is argued that central banks should
of holding money would not be the nominal rate on
not target a zero inflation rate, since by doing so they
bond it but it − iM M
t where it is the statutory fixed
give up the ability to stabilize the economy if a shock
nominal return on money balances. The LM curve
requires a negative market-clearing real interest rate.
would then be
Friedman’s rule which prescribes to target a rate of
Mt Yt
= . (8) inflation resulting in a zero nominal interest rate runs
Pt 1 + it − iMt exactly this kind of risk.
If iM
t is negative, the opportunity cost of holding
money could still be positive even if it were nega-
tive. Agents would still respond to an increase in the
money supply by switching to bonds pushing their
nominal rate of return below zero.
The problem is that taxing money may be difficult
since it is a bearer asset (that is why it is difficult to
tax the black economy).
7