Liquidity Trap in a Simple AD-AS Model

Siddhartha Chattopadhyay
Department of Humanities and Social Sciences
IIT Kharagpur

July 14, 2015

Abstract

This paper explains the dynamics under liquidity trap using the simple AD-AS
model of Mankiw (2012). The model is solved and impulse response is analyzed
under large and persistent demand shock. Following Mankiw (2012), the model is
solved under non-rational expectation so that it remains accessible and still be able
to communicate key results easily to the beginners.

JEL Classi…cation: E63, E52, E58

Keywords: New-Keynesian Model, In‡ation Target, Liquidity Trap
1 Introduction
Chapter 15 of Mankiw (2012) presents a very simple dynamic AD-AS model for the
beginners. Mankiw (2012) provides an analytical solution of the model under both demand
and supply shock. Moreover, impulse response of the model has also been analyzed in
Mankiw (2012) under demand shock. However, zero lower bound of interest rate and
the impact of liquidity trap has not been analyzed in Mankiw (2012). Buttet and Roy
(2014) has shown how to introduce zero lower bound to nominal interest rate and analyze
the impact of liquidity trap using the simple AD-AS model. They have shown that
liquidity trap produces an additional long-run equilibrium. The equilibrium is unstable
and can lead to de‡ationary spiral. A full blown dynamics of the model under liquidity
trap is analyzed here. Following Mankiw (2012), the model is solved under non-rational
expectation to keep it accessible to the beginners.
We know liquidity trap as a Keynesian policy prescription from our undergraduate
days. It produces a de‡ationary equilibrium with zero nominal interest rate and output
below its full employment level. Simple static AD-AS model suggests that, monetary
policy is completely impotent and we have to take help of …scal policy when economy is
in liquidity trap. Liquidity trap has been considered as a mere theoretical curiosity for
long. However, it becomes a hard-core reality after the experience Japan since 1995 and
US since 2008. As a result, liquidity trap is a key area of research in modern monetary
economics.
Sticky price New Keynesian DSGE model is the work horse of modern monetary
economics.1 However, a full blown analysis of liquidity trap under sticky price New
Keynesian DSGE model is beyond the scope and also not the objective of the paper.2
The paper is written for the beginners, intending to pursue higher studies and research
in monetary economics. As a result, entire analysis has been done using simple AD-AS
model under non-rational expectation so that the it remains accessible to the beginners.
This paper seems to be a good introduction to the main stream cutting edge research
in monetary economics. The paper proceeds as follows. Section 2 explains the model.
Section 3 gives the solution of the model. Impulse response is analyzed Section 4 and
Section 5 concludes.
1
See, Woodford (2003) and Walsh (2010) for a detail description of sticky price New Keynesian DSGE
model.
2
See, Cochrane (2013) for a detail analysis of liquidity trap under sticky price model.

1
2 The Model
The demand side of the economy is given equation (1) below.3

yt = [ t Et ( t+1 )] ut (1)

Here, yt denotes the deviation of output from its long-run value. Here, the long-run value
of output is assumed to be zero. t is in‡ation rate; t = it i denotes the deviation of
the nominal interest rate from its long-run value given as i = r = 1 ;with r de…ned as
the long-run real interest rate; represents the intertemporal elasticity of substitution
with 1: The demand curve shows that, output gap depends negatively on deviation
of real interest rate from its long-run value, given by rbt = t Et ( t+1 ). As a result, the
demand equation can also be called the IS equation. Moreover, 2 (0; 1) denotes the
discount factor; and ut represents demand shock. We assume, ut to follow a deterministic
dynamics as given below,
ut = t 1 u1 ; 2 (0; 1)

Here, Et is the conditional expectation operator and 2 (0; 1) is the persistence of the
demand shock. We assume that, economy gets hit by a large adverse demand shock of
magnitude u1 such that rtn = i 1 t 1
u1 < 0 for t = 1; 2; ::::; Tmin and rtn = i
1 t 1
u1 > 0 for t = Tmin + 1; Tmin + 2; :::::.
Note, the demand curve does not have any dynamics. The entire dynamics of the
model is governed by the supply equation given by equation (2) below.

t = Et 1 ( t ) + yt (2)

Here, > 0 is the slope of the supply equation. The supply is equation is also called as
Phillips curve in the literature.
The nominal interest rate is the instrument of the monetary authority, determined by
a Taylor rule (Taylor, 1993) given below.

1
t = it i = Et t+1 + ( t t) + y (yt yt ) ut (3)

The Taylor rule given above is assumed to follow Taylor principle such that @@i > 1. Taylor
principle is satis…ed when > 1 and y 2 (0; 1). t in the Taylor rule is in‡ation target
and output target is yt Note, at steady state when t = t , equation (2) determines
3
Mankiw (2012) assumes, long-run output Y t > 0. I assume Y t = 0 in my paper.

2
yt = 0. Therefore, the Taylor rule becomes,

1
t = it i = Et t+1 + ( t t) + y yt ut
it = rtn + Et t+1 + ( t t) + y yt (4)

This completes the description of the model.

3 Solution of the Model

Following Mankiw (2010) we solve the model under non-rational expectation by as-
suming Et ( t+1 ) = t , Et t+1 = t and Et 1 ( t ) = t 1 .4 We assume that, economy
is in liquidity trap till period T . This implies, it = 0 for t = 1; 2; ::::; T and iT +1 > 0 for
T + 1 onwards. The key equations of the model with above assumptions are,

yt = [ t t] ut
t = t 1 + yt
t 1
ut = u1 ; 2 (0; 1)
it = 0; for t = 1; 2; :::T
it = rtn + t + ( t t) + y yt ; for t = T + 1; T + 2; :::::

The model is solved in two parts. The time period when economy is in liquidity trap and
when economy is out of liquidity trap.

3.1 During Liquidity Trap

Appendix shows that, till period T the demand curve is positively sloped and given
by equation (5) below.
yt = rtn + t (5)

The supply curve is also positively sloped and governs the entire dynamics of the model.
The equation of supply is given by (6) below.

t = t 1 + yt (6)
4
Such simpli…cation of expectation formation is made to keep the model accessible to the beginners.
The model can be solved easily under rational expectation as well.

3
We have to solve equation (5) and (7) simultaneously with initial condition y0 = 0 = 0.
Appendix shows that the forward looking solution of equation (6) with 0 = 0 gives,

X
t

t = yk (7)
k=1

Solving equation (5) and (7) we have,

y1 = r1n
1
1 = r1n (8)
1

and,
j 1
X
yj = rjn + yk ; for j = 2; 3; :::; T (9)
1 1 k=1

X
T

j = yj ; for j = 2; 3; :::; T (10)

j=2

Note, rtn = i 1
ut = i 1 t 1
u1 is known for all t = 1; 2; :::::: As a result, we can
calculate y1 and 1 from equation (8). Moreover, both y1 < 0 and 1 < 0 since, rtn < 0 for
t = 1; 2; :::; T . So, there is recession and de‡ation when economy falls into liquidity trap
due to large adverse demand shock. Once, y1 is known, yt for t = 2; 3; :::; T can be solved
from equation (9). Moreover, with known yt for t = 1; 2; ::::; T , equation (10) solves t for
t = 2; 3; :::; T .

3.2 After Exit from Liquidity Trap

Economy gets out of the liquidity trap at t = T + 1. Now we solve the model for
t = T + 1 onwards. Note, nominal interest rate is determined by Taylor rule when
economy is out of trap. Equation (1) after substituting Taylor rule (equation (??)) for it
gives the demand curve as,

( 1) ( 1)
yt = t + t (11)
1+ y 1+ y

4
Also note, demand curve is negatively sloped when Taylor principle is followed with >1
and y 2 (0; 1). Now, substituting equation (11) to equation (6) gives,

t = t 1 +z t (12)

1 1
where, = ( 1) < 1 and z = 1+ y
. Since, < 1, we can solve equation (12)
1+ 1+ 1+ ( 1)
y
backward with given initial condition (yT : T ). The backward solution of equation (12)
gives,
j 1
X
j k
T +j = T +z T +j k ; for j = 1; 2; ::: (13)
k=0

Now, substituting equation (13) to equation (11) gives,

j 1
!
( 1) X ( 1)
j k
yT +j = T +z T +j k + T +j ; for j = 1; 2; ::: (14)
1+ y k=0
1+ y

Once, yt and t are known for t = T + 1; T + 2; :::: nominal interest rate for t = T + 1; T +
2; :::: is calculated as,
it = rtn + t + ( t t ) + y yt (15)

3.3 Determination of T
T is determined as follows. First, determine Tmin as the last period when rtn < 0 .
Second, solve the model numerically for di¤erent T Tmin . Third, exit time from liquidity
trap is the …rst time period such that iT +1 > 0 and Taylor rule becomes applicable for the
…rst time. This determines the exit time T . Note, exit time is endogenously determined
in the model.

4 Impulse Response
Note, in‡ation target, t is chosen by monetary authority.5 Moreover, solution of
the model given above shows that, we get di¤erent equilibrium solution of yt and t for
di¤erent values of in‡ation target. Therefore, multiple equilibria are indexed by in‡ation.
This is an important feature of New Keynesian model. We have analyzed impulse response
5
In‡ation target can be chosen to minimize loss function of the monetary authority. See, Jung,
Teranishi and Watanbe (2005), Werning (2013) and Chattopadhyay and Daniel (2014) for optimal policy
without uncertainty. See, Adam and Billi (2006, 2007), Eggertson and Woodford (2003) and Nakov
(2008) for optimal policy under uncertainty in a sticky price model.

5
of the model for t = 0 for all t. Following Taylor (1993), we also set, = 1:5; y = 0:5.
Following literature of New Keynesian model, we also set, = 1; = 0:99 and = 0:057
for our analysis. These parameterization is used to generate quarterly impulse response
in the literature. Moreover, we also set, = 0:9 and u1 = 2:4% so that impulse response
can capture experience of Japan and US who are in liquidity trap respectively since 1995
and 2008. The impulse response is portrayed in Figure 1.

Figure: 1

The impulse response shows that, liquidity trap initially puts the economy into recession.
Then, the economy goes to boom and then output gap gradually converges to zero long-
run equilibrium. The movement of output gap portrayed in Figure: 1 can be easily
explained from equation (8), (9) and (14). Note, equation (9) determines output gap as a
weighted sum of natural rate of interest rate and past values of output gap. The natural
rate of output changes sign at Tmin from negative to positive. As a result, output gap is

6
negative and economy goes to recession initially. However, the economy moves to boom
as natural rate of interest becomes positive after Tmin . Note, as soon as the natural rate
of interest becomes positive, the second term of equation (9) is also getting less negative.
This dual impact makes output gap positive and puts the economy into boom. Moreover,
the dynamics of output gap is determined by equation (14) as soon as the economy gets
out of liquidity trap. Since, we set in‡ation target zero, the output gap gradually moves
to zero with persistence as shown in equation (14).
However, Figure: 1 also shows that though economy moves from recession to boom
before converging to zero long-run equilibrium, it remains in de‡ation throughout the pe-
riod. The dynamics of in‡ation rate portrayed in Figure: 1 can be explained by equation,
(8), (10) and (13). Note, since dynamics of in‡ation rate given in equation, (8), (10) gets
determined by sum of current and output gap, it remains negative even if economy moves
to boom. Moreover, the dynamics in‡ation rate is determined by equation (13) as soon
as the economy gets out of the liquidity trap. As a result, since in‡ation target is set to
zero, in‡ation rate converges to long-run equilibrium with persistence as soon as the
trap gets over.
The dynamics of nominal interest rate portrayed in Figure: 1 is self explanatory. It
remains zero as long as the economy is in liquidity trap but follows Taylor rule given
in equation (15). Since, economy is in liquidity trap till T = 21, nominal interest rate
portrayed in Figure: 1 is zero till T = 21, and becomes positive after after that.

5 Conclusions
This paper analyzes liquidity trap using a simple AD-AS model of Mankiw (2010). The
objective of the paper is to introduce a dynamic analysis of liquidity trap to the beginners.
Therefore, the paper is written using non-rational expectation so that the analysis remains
accessible to the beginners. Though, the simple model under non-rational expectation
is used, the paper is able to communicate some important concepts of liquidity trap.
The impulse response shows that a large and highly persistent demand shock causes an
economy to be into liquidity trap for more than 5 years. The impulse response also shows
that, though economy goes to boom after initial recession, it remains in de‡ation through
out the time period before convergence to zero long-run equilibrium. The simple model,
solved under non-rational expectation cannot match the stylized facts of output gap and
in‡ation rate but the analysis presented in this paper is de…nitely a good beginning to
the main stream cutting edge research in monetary economics.

7
6 Appendix: Solution Hint
1. The Demand Curve:

yt = rt ut
= ( Et ( t+1 )) ut
1
= ii t + ut
= (ii rtn t)

rtn = 1
ut
1 t 1
= u1

For ii = 0 we have,
yt = rtn + t

t = Et 1 ( t ) + yt
= t 1 + yt

2. Forward Solution of Supply Curve:

X
1

t = yk
k=t+1

Therefore,
X
1

0 = yk
k=1

Therefore,
X
t

t 0 = yk
k=1

Setting, 0 = 0 gives,
X
t

t = yk
k=1

8
 3. The Taylor rule:

1
b = Et t+1 + ( t t) + y (yt yt ) ut
1
= t + ( t t) + y (yt yt ) ut
= t + y yt ( 1) t

4. In‡ation after Exit:

T +1 = T +z T +1

T +2 = T +1 +z T +2

= T +z T +1 +z T +2
2
= T +z T +2 + z T +1

T +3 = T +2 +z T +3
2
= T +z T +2 + z T +1 +z T +3
3 2
= T +z T +3 + z T +2 + z T +1

)
j 1
X
j k
T +j = T +z T +j k ; for j = 1; 2; :::
k=0

References
[1] Adam, K. and R.M. Billi, (2006), Optimal Monetary Policy under Commitment with
a Zero Bound on Nominal Intrest Rates, Journal of Money, Credit, and Banking
39(7), 1877-1905.

[2] Adam, K. and R.M. Billi, (2007), Discretionary Monetary Policy and the Zero Lower
Bound on Nominal Interest Rates, Journal of Monetary Economics 54(3), 728-752.

[3] Bullard, J. and K. Mitra, (2002), Learning about Monetary Policy Rules, Journal of
Monetary Economics, 63(3), 477-96.

9
 [4] Buttet, S. and U. Roy (2014), A Simple Treatment of the Liquidity Trap for Interme-
diate Macroeconomics Courses, The Journal of Economic Education, 45(1), 36-55.

[5] Calvo, G. A. (1983), Staggered Prices in a Utility Maximizing Framework, Journal

of Monetary Economics, 12(3), 983-98.

[6] Chattopadhyay, S. and B. Daniel (2014), In‡ation Target at Zero Lower Bound,
Working Paper, University at Albany, SUNY.

[7] Cochrane, J. H. (2013), "The New Keynesian Liquidity Trap," manuscript, University
of Chicago, Booth School of Business.

[8] Eggertson, G. and M. Woodford (2003), The Zero Bound on Interest Rates and
Optimal Monetary Policy, Brookings Paper on Economic Activity, 1, 139-211.

[9] Jung, Taehun, Yuki Teranishi, and Tsutomu Watanabe (2005), Optimal Monetary
Policy at the Zero-Interest-Rate Bound, Journal of Money, Credit, and Banking
37(5), 813-835.

[10] Mankiw, N. G. (2012), Macroeconomics (Chapter 15, 8th edition), Worth Publishers.

[11] Nakov, A. (2008), Optimal and Simple Monetary Policy Rules with Zero Floor on
the Nominal Interest Rate, International Journal of Central Banking, 4(2), 73-128.

[12] Taylor, J. B. (1993), Discretion versus Policy Rules in Practice, Carnegi-Rochester

Conference Series on Public Policy, 39, 195-214.

[13] Walsh, C. E. (2010), Monetary Theory and Policy, Third Edition, The MIT Press,
Cambridge, MA.

[14] Werning, I. (2012), Managing a Liquidity Trap: Monetary and Fiscal Policy, manu-
script, MIT.

[15] Woodford, M. (2003), Interest and Prices: Foundations of a Theory of Monetary

Economics, Princeton University Press, Princeton and Oxford.

10