Transcript Presentation - The Cambridge Trust for New Thinking in Economics

Debt monetization, inflation, and the
“neutral” interest rate
by
Alfonso Palacio-Vera
Universidad Complutense de Madrid, Spain
Conference “Economic Policies of the New Thinking in
Economics”
St. Catharine´s College, Cambridge, UK
Thursday 14 April 2011
Introduction (I)
• A number of central banks (ECB, US Fed, Bank of Japan, and
Bank of England) have resorted, mainly from 2008 onwards, to
large-scale purchases of sovereign debt in an attempt to:
– Provide liquidity to the financial sector and help compress term and risk
premiums
– By-pass malfunctioning financial markets like the inter-bank market
– Alleviate the financial burden to the Treasuries of the massive fiscal
expansions that were engineered to counteract the collapse of private
expenditure in the wake of the financial crisis of 2007-9
• The US Fed has also purchased large quantities of mortgagebacked securities.
Introduction (II)
• The empirical evidence suggests that these “unconventional”
monetary policy measures succeeded, to some extent, in:
– Flattening the money market yield curve
– Containing risk/liquidity premiums in dysfunctional money markets
– Preventing the collapse of the inter-bank market
• Some examples are:
– US Fed: Gagnon et al. (2010), D´Amico and King (2010), Hamilton and
Wu (2011), Del Negro et al. (2010).
– Bank of England: Joyce et al. (2010)
– Bank of Japan: Ueda (2010)
– ECB: Lenza et al. (2010)
Introduction (III)
• Nevertheless, the effectiveness in stimulating aggregate demand
of debt monetization or “quantitative easing” remains unclear.
– Baumeister and Benati (2010) find that the compression in the long-term
bond yield spread during the period 2007-2009 in the US, the Euro area,
Japan, and the UK exerted a “powerful” effect on output growth.
– Fullwiler and Wray (2010) question the efficacy of debt monetization:
‘If we consider the possible negative impact on income and spending
resulting from lower interest income received on savings, a plausible case
can be made that QE2 [Quantitative Easing] will actually be deflationary
if the policy is successful in lowering rates’ (Fullwiler and Wray, 2010, p.
10)
Introduction (IV)
• As shown in Auerbach and Obstfeld (2005), there still remains
an argument for large-scale open market purchases of public
debt as a “fiscal” policy tool when short-term nominal interest
rates are zero.
• This is because, provided long-term interest rates are positive,
trading money for interest-bearing public debt reduces future
debt-service requirements thus allowing fiscal policy to become
more expansionary.
The ‘decoupling’ principle (I)
• Several authors emphasize that CBs are able to fully control the
supply of reserve balances and this, in turn, allows them to set
the price at which they are supplied by simply announcing the
target short-term interest rate (Friedman and Kuttner, 2010).
• Consequently, there arises a strong disconnection between the
level of short-term nominal interest rates and the quantity of
liquidity actually supplied.
The ‘decoupling’ principle (II)
• Such disconnection is most evident when the CB adopts a ‘floor
system’ whereby the former can target independently both the
amount of bank reserves and the overnight interest rate in the
inter-bank market.
• In a ‘floor-system’ the CB manipulates short-term interest rates
by varying the interest rate it pays on reserves instead of varying
the quantity of reserves it supplies.
• Under this system, the interest rate paid on reserves effectively
becomes a floor below which the market rate cannot fall.
The ‘decoupling’ principle (III)
• By adopting a ‘floor-system’, a CB actually divorces the quantity
of reserves from the interest rate target and in the process it gains
one degree of freedom insofar as it can now employ interest on
reserves together with open market purchases of assets to target
independently short-term nominal interest rates and the quantity
of reserves.
• Borio and Disyatat (2009) have labelled this feature of a ‘floor
system’ the ‘decoupling’ principle (see Figure 1).
The ‘decoupling’ principle (IV)
Figure 1: A ‘floor system’ of monetary implementation
The ‘decoupling’ principle (V)
• The main aim of the study is to explore the potential for macrostabilization purposes of the utilization by the CB of its ability to
manipulate both the short-term interest rate and the size of the
money base.
• For that purpose, we construct a closed-economy macro model
with Keynesian features that is made up of three sectors:
– Debtor and creditor households
– Businesses
– Public sector: Treasury and the Central Bank
The model (I)
• Assumptions:
– Financial intermediaries channel saving flows among the three sectors.
– Creditor households purchase all the securities issued by the other sectors
except those ones that are “monetized”.
– Securities issued by different sectors (e.g. debtor households, businesses,
Treasury) exhibit different risk premiums.
– Debtor households, businesses, and the Treasury adjust spending so as to
hit a “target” debt-to-capital ratio in the long run.
– All securities are perpetuities, i.e. they pay coupons forever and issuers do
not have to redeem them.
– Treasury revenue comes from both taxes and CB transfers.
– Interest income accruing to creditor households is paid in arrears so that
the debt service depends on last period´s debt and nominal interest rate.
– The CB adopts a ‘floor’ system.
The model (II)
• Supply-side:
– Potential output in real terms: y t    N t  v  k t
– Capacity utilization: u t 
– CIER: u t 
pt  yt
vKt
 pt 
yt
v  kt
1
  t   e  Lt 
t Nt

 p t     
  1
v Kt
 v   Kt 
– “Natural” growth rate: g n  n  a
– Change in the inflation rate:   t    u t 1  u t 1 
The model (III)
• Consumption:
– Debtor households´ consumption in money terms:
C D , t  c D  z (1   )(1  t w ) Y t 1  (  H  i t 1) D H , t 1    D H ,t
h H , t   D H , t / K t 1
 Yt
 

 Y t 1
H

ˆ
( d H , t 1  d H )   d H , t 1

– Creditor households´ consumption in money terms:
C c ,t  c c (1  z )(1   )(1  t w ) Y t 1  (1   t ) (1  t c )(1  t p ) Y t 1  CHII
CHII
RII
t
t
t

 ((  H  i t 1) D H , t 1  (  F  i t 1)( 1   MF ) D F , t 1  (  G  i t 1)( 1   GM ) D G , t 1  RII t )( 1  t c )
M
 M
F i t 1 D F , t 1   G i t 1 D G , t 1
where CHII = creditor households´ nominal interest income and
RII = interest on CB reserves accruing to creditor households.
The model (IV)
• Investment and businesses´ indebtedness in money terms:
*
I t / K t 1  g t    f  f u ( u t 1  u ( ))  
f  *  gn
RP t   t ( Y t 1  (  F  i t 1) D F , t 1)( 1  t p )
*
 t     d  ( d F , t 1  dˆ F )
 ,
*
d
 0
*
h F , t   D F , t / K t 1  f  f u ( u t 1  u ( ))     t ( v u t 1  (1   F ) i t 1 d F , t 1)( 1  t p )
where RP = retained profits
 = firms´ average retention rate
The model (V)
• Public sector:
G t  (1   G ) i t 1 D G , t 1  T t   D G , t  RCB
t
M
M
 D GM,t   D M
F , t  RCB t  (1   G ) i t 1 D G , t 1  (1   F ) i t 1 D F , t 1  CBRII
CBRII
t
t
 Ht
 i t 1  GM D G , t 1   M
F D F , t 1 
 H t   GM   G G , t   MF   D F , t
b t  b 0   ( d G ,t 1  dˆ G )  (1   )( u t 1  u t 1)
h G ,t

 D G ,t
K t 1
 b t  (1   G )   G  GM   i t 1 d G , t 1   F  MF i t 1 d F , t 1
where CBRII = flow of interest payments to reserve holders, b t  ( G t  T t ) / K t 1
and b0 captures the stance of discretionary fiscal policy.
The model (VI)
• Goods market equilibrium:
St  It  Gt  T t
 Y t

Ct
 1   t  (1  t p )   v u t 1 
  t (1  t p )(  F  i t 1) d F , t 1  g t    b t

K t 1
 Y t 1

• Monetary policy rule:
*
*
i t  r   t  1  a  ( t  1   )
where r * is the steady-growth “neutral” interest rate and
a  0
The model (VII)
• A liquidity trap (LT) is typically defined as a situation in which the real
interest rate (in a closed economy without a government sector) at which
aggregate saving at the level of output consistent with constant inflation and
investment would be equal is negative (Krugman 1998).
• If we denote by  the minimum (ex-ante) actual real interest rate that a CB
can set for a given expected rate of inflation, we may define a LT as a
situation where:
r 
n
• If we express  as the difference between the risk/term premium  and
the expected rate of inflation  e then we say an economy is mired in a LT
if:
e

 
r
n
Steady-growth equilibrium
• Steady-growth equilibrium:
  u  u   d H   d F   d G  0
*
 D t / D t 1   Y t / Y t 1  g t    g n
d H ,t  dˆ H
d F  dˆ F
d G , t  dˆ G
u  u  u * ( )
*
*
*
i  r
*
M
M
b 0  ( g n  r * ) dˆ G  (1   G )  G dˆ G   F  F dˆ F
 
1
*
 (1  t p ) v u  (  F    r )( 1  t p ) dˆ F
*
*
*
 (1  dˆ F )(  *  g n )  

Steady-growth analysis (I)
• The steady-growth “neutral” interest rate is given by:
 1 


   A 0  A   *  A g  g n 
r
 Ar 
*
where the sign of all the operators is a priori ambiguous and:
A r   t p c c (1  t c ) dˆ F  [ c D  c c (1  t c )] dˆ H  [1  c c (1  t c )] dˆ G
*
A   (1  v u ( ))  [1  ( c D  c c (1  t c ))] dˆ H  c c (1  t c ) dˆ G  c c (1  t c )[ 1  c c (1  t p )] dˆ F
 [1  c c (1  t c )]( 1  dˆ F )
Steady-growth analysis (II)
• The ambiguity of A r obeys to the fact that an increase in the debt
service of businesses (brought about by an increase in the real
interest rate) causes both an increase and a decrease in creditor
households´ interest and profit income respectively thus having
an uncertain impact on their consumption.
• Yet, under normal circumstances, we may assume A
r
 0
.
• The sign of A  is also ambiguous, albeit the higher the debt-tocapital ratios are, the more likely it is that it will be positive.
• However, the former may irrelevant when the economy is mired
in a LT since the increase in the inflation target may simply not
be credible.
Steady-growth analysis (III)
• The comparative dynamics results are:
Table 1: Impact on r * of changes in the parameters

G
M
F
M
z

dˆ H
dˆ G
dˆ F
H
G
F
cD
cc
gn

*
+
+
+
?
?
?
?
-
-
+
?
?
?
?
r
*
Note: (+) and (-) mean that an increase in the variable listed on the upper row gives rise to an increase
and a decrease respectively in the value of r * whereas (?) means that the effect is ambiguous
Steady-growth analysis (IV)
• In particular, we have that:
 r*

M
F
 r*

M
G

(1  c c (1  t c ))  F dˆ F
0
iff  F  0
0
iff  G  0
Ar

(1  c c (1  t c ))  G dˆ G
Ar
• Debt monetization entails a redistribution of income away from
creditor households and towards the Treasury so that aggregate
demand will increase insofar as the public sector sets the level
of the budget deficit so as to hit its target debt-to-GDP ratio.
Steady-growth analysis (V)
• And
 r*
 H
 r*
 G
 r*
 F



 ( c D  c c (1  t c )) dˆ H
 0
Ar
M
 [1  c c (1  t c )]( 1   F ) dˆ G
 0
Ar
[ (1  c c (1  t c ))  MF  t p c c (1  t c )] dˆ F
0
Ar
• The fall in  F leads to an increase and a decrease in creditor
households´ profit and interest income respectively but it also
leads to a decrease in the transfers of income from the CB to the
Treasury and, hence, in the public budget balance.
Conclusion (I)
• The monetization of public and private debt operates primarily
through its impact on the transfer of CB revenue to the Treasury
and, hence, it constitutes an important fiscal policy tool.
• Its effectiveness will be directly proportional to the size of the
risk/term premiums on different types of securities and on the
proportion of these securities that are purchased by the CB.
• If debt monetization is to be effective as a fiscal policy tool,
there is a need for a high degree of coordination between the CB
and the fiscal policy authorities.
Conclusion (II)
• The effectiveness of debt monetization when it brings about the
compression of risk and term premiums on different types of
securities is unambiguous in the case of public debt.
• However, the net impact on aggregate demand stemming from
the compression of risk/term premiums on securities issued by
businesses may be negative due to:
– The ambiguous impact on consumption by creditor households and
– The adverse impact on transfers from the CB to the Treasury
• Finally, an increase in the inflation target may have (provided it
is credible) an expansionary effect when the aggregate debt-tocapital ratio in the economy is relatively high.