No. 162
NBP Working Paper
NBP Working Paper No. 162
www.nbp.pl
A note on central bank transparency
and credibility in Poland
Tomasz Łyziak
NBP Working Paper No. 162
A note on central bank transparency
and credibility in Poland
Tomasz Łyziak
Economic Institute
Warsaw, 2013
Tomasz Łyziak – Narodowy Bank Polski, [email protected]
The views expressed in this note are of the author and do not necessarily reflect those of the
Narodowy Bank Polski. I wish to thank Joanna Mackiewicz-Łyziak and Ewa Stanisławska,
the co-authors of the article on central bank transparency and credibility in Poland [Łyziak
et al. (2007)], for discussing its updated results presented in this note. I am also grateful to
Jan Przystupa and Ewa Wróbel for helpful comments. All remaining errors are mine.
Print:
NBP Printshop
Published by:
Narodowy Bank Polski
Education & Publishing Department
ul. Świętokrzyska 11/21
00-919 Warszawa, Poland
phone +48 22 653 23 35
www.nbp.pl
ISSN 2084-624X
© Copyright Narodowy Bank Polski, 2013
Contents
Contents
1 Theoretical remarks on central bank transparency and credibility
7
5
2 Inflation targeting in Poland
8
6
3 Assessing central bank transparency
6
8
4 Assessing central bank credibility
8
10
5 Conclusions
10
12
1
NBP Working Paper No. 162
3
List of Figures
1
History of inflation targeting in Poland, 1998-2012 . . . . . . . . . . . . . . . . . . . . . . . .
13
15
2
Interest rate forecast errors based on financial market data . . . . . . . . . . . . . . . . . . .
15
17
3
Consistency of monetary policy decisions with the NBP staff inflation forecasts (4-quarter
horizon) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
15
17
4
12-month inflation expectations and the NBP inflation target . . . . . . . . . . . . . . . . . .
16
18
5
Credibility indices based on Cecchetti and Krause (2002) aproach
16
18
6
12-month inflation expectations of financial sector analysts vs. NBP inflation target and NBP
. . . . . . . . . . . . . . .
staff inflation projections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
17
19
List of Tables
1
Eijffinger and Geraats (2002) transparency index for Poland in 2000, 2005 and 2010 . . . . .
13
15
2
Mahadeva and Sterne (2000) transparency index for Poland in 2000, 2005 and 2010 . . . . . .
14
16
3
Transparency indices and measures of interest rate forecast errors . . . . . . . . . . . . . . . .
15
17
4
Inflation expectations and the NBP inflation target . . . . . . . . . . . . . . . . . . . . . . . .
16
18
5
Estimation results of the credibility parameter based on Bomfim and Rudebusch (2000) approach
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
17
19
2
4
Narodowy Bank Polski
Abstract
Abstract
This note extends the study by L
yziak et al. (2007), providing up-to-date assessment of central bank
transparency in Poland. We highlight the role of inflation projections prepared by the staff of the
National Bank of Poland in building transparency of monetary policy. The results suggest that central
bank inflation projections, published since 2004, have led to improvements in the predictability of interest
rate decisions.
The note updates also previous estimates of the degree of central bank credibility in Poland, using
survey-based measures of inflation expectations formed by consumers, enterprises and financial sector
analysts. It is confirmed that inflation expectations of enterprises and – especially – of financial sector analysts display a high degree of anchoring at the NBP inflation target, while consumer inflation
expectations are driven mainly by developments in subjectively perceived inflation.
JEL: D84, E52, E58.
Keywords: Transparency, Credibility, Expectations, Inflation Targeting, Poland.
3
NBP Working Paper No. 162
5
Introduction
“The mystery and mystique has given way to transparency and openness. . . The communication
of policymakers’ intentions with a view to enhancing their credibility has come to play a central
role in monetary policy”. (Mervyn King, 20001 )
“There may be circumstances in which it is justified to aim off the inflation target for a while
in order to moderate the risk of financial crises. (...) I do not see this as inconsistent with
inflation targeting because it is the stability of inflation over long periods, not year to year
changes, which is crucial to economic success. The key principles underlying flexible inflation
targeting are credibility, predictability and transparency of decision-taking, and they will remain
the cornerstone of successful monetary policy in the future”. (Mervyn King, 20122 )
Introduction
In the inflation targeting (IT) framework central banks attempt to establish credibility by conducting
a transparent policy, therefore the adoption of this framework is typically followed by major improvements
in central bank communication with the public and with the market [Mishkin and Schmidt-Hebbel (2001)].
Due to delays in the monetary transmission mechanism – which in the case of Poland are approximately
4-6 quarters [Demchuk et al. (2012)] – central banks with quantitative inflation targets must be able to
form views on future macroeconomic performance, especially on inflation prospects. Moreover, they should
have good understanding of the monetary transmission mechanism. For these reasons inflation forecasts,
capturing not only macroeconomic developments in general, but highlighting the effects of monetary policy,
are crucial in the process of building central bank transparency and credibility.3
Empirical studies show that transparent policy makes central banks able to reduce uncertainty about future interest rates, anchor inflation expectations, lower inflation variability and sacrifice ratios [e.g. Mishkin
and Schmidt-Hebbel (2007), Dincer and Eichengreen (2013)].4
By publishing detailed macroeconomic
forecasts central banks reduce inflation bias and become able to better control inflation [Geraats (2001),
Chortareas et al. (2002)]. It is due to the fact that central bank inflation forecasts – even if imprecise – exert
influence on private sector expectations [Tarkka and Mayes (1999), Crowe and Meade (2008), Ehrmann et al.
(2010)], supporting the process of building central bank credibility. Recent evidence suggests that during the
global financial crisis started in 2008 inflation-targeting central banks performed better than the others in
managing real interest rates5 and in terms of inflation volatility, although the assessment of output performance during the crisis does not point any differences between inflation targeters and remaining central banks
[Fouejieu (2013)]. The latter observation can be explained with the fact that IT economies are economically
and financially more open than the other ones, so they were more exposed to financial crisis disturbances.
1 Joint
luncheon of the American Economic Association and the American Finance Association, Boston, 7 January 2000.
Stamp Memorial Lecture, London School of Economics, 9 October 2012.
3 Inflation forecasts play a very important role in the inflation-targeting framework. See Svensson (2003) for a detailed
discussion of the concept of inflation-forecast targeting.
4 These benefits from inflation targeting seem to be greater in developing economies than in the developed ones [Mishkin and
Schmidt-Hebbel (2007)].
5 Fouejieu (2013) shows that inflation-targeting central banks had more room for monetary policy easing during the crisis,
so they reduced their short term interest rates by more than non-targeters.
2 The
4
6
Narodowy Bank Polski
Chapter 1
The aim of this note is to assess transparency and credibility of monetary policy in Poland in 1998/20012012. The note updates the results presented in the study L
yziak et al. (2007), in which we evaluated the
degree of central bank credibility and transparency during the initial stage of inflation targeting in Poland
(1998-2004) and found that inspite of intensified communication of the central bank with the public, the
predictability of interest rate decisions had not increased after the adoption of inflation targeting, while the
credibility of inflation targets among professional forecasters had been high inspite of inflation target misses.
The need to update those findings results particularly from the fact that the major improvements in the
communication of the NBP with the public – including the publication of the NBP staff inflation projections
(since August 2004) and minutes (since May 2007) – occurred after the end of the sample period analysed
in L
yziak et al. (2007).
1
Theoretical remarks on central bank transparency and credibility
Central bank transparency is a multidimensional phenomenon. Following literature, L
yziak et al. (2007)
distinguish two aspects of transparent monetary policy. The first one is related to central bank efforts to
make the decisions of monetary policy understood by the public. In this view monetary policy is transparent
if economic agents are provided with a full access to information necessary to understand goals and decisions
of monetary policy-makers. This aspect of central bank transparency can be measured with well-known
transparency indices, assessing central bank disclosure practices, such as those of Eijffinger and Geraats
(2006) or Mahadeva and Sterne (2000). Publication of central bank forecasts, models and explanations of
policy decisions are the most important elements taken into account while calculating transparency indices
of this kind.
Explanation of policy is a crucial, but insufficient condition for central bank transparency. Effectiveness of
communication is another important aspect to be considered in analysing central bank transparency [Winkler
(2002)]. From this point of view, monetary policy is transparent if it is predictable, i.e. if information released
by the central bank helps the private sector understand policy objectives and predict central bank decisions.6
Better understanding of central bank actions achieved via appropriate communication policy should
influence ability of the central bank to manage inflation expectations of the private sector. A central bank
that is able to anchor inflation expectations at the level consistent with monetary policy goals is perceived as
credible.7 Therefore existing measures of central bank credibility refer to a gap between inflation expectations
of economic agents and the central bank inflation target [e.g. Faust and Svensson (2001); Hutchison and
Walsh (1998); Cecchetti and Krause (2002)] or to the weight attached to the inflation target in the formation
6 Empirical evidence supports this view. E.g. Fracasso et al. (2003) show that the quality of Inflation Reports reduces
interest rate surprises.
7 According to Blinder (2000), “a central bank is credible if people believe it will do what it says”. This is in line with the
concept of Cukierman and Meltzer (1986), who define credibility as “the absolute value of the difference between policy-makers’
plans and the public’s beliefs about those plans”.
5
NBP Working Paper No. 162
7
long-term
Chapter 2 / of
Chapter
3
inflation expectations by economic agents [Bomfim and Rudebusch (2000)].
There are different factors influencing central bank credibility. The most important ones – suggested
of long-term inflation expectations by economic agents [Bomfim and Rudebusch (2000)].
by Blinder (2000) on the basis of surveys conducted among central bankers and academic economists and
There are different factors influencing central bank credibility. The most important ones – suggested
confirmed empirically in the cross-country analysis by Mackiewicz-‚Lyziak (2010) – comprise: a history of
by Blinder (2000) on the basis of surveys conducted among central bankers and academic economists and
achieving central bank goals, central bank independence, a history of fighting inflation, central bank transconfirmed empirically in the cross-country analysis by Mackiewicz-‚Lyziak (2010) – comprise: a history of
parency and the fiscal discipline by the government that conditions the effectiveness of monetary policy.
achieving central bank goals, central bank independence, a history of fighting inflation, central bank transparency and the fiscal discipline by the government that conditions the effectiveness of monetary policy.
2
2
Inflation targeting in Poland
Before
1998 the strategy
of the National
Bank of Poland combined elements of the exchange rate targeting,
Inflation
targeting
in Poland
inflation targeting, and money supply targeting applied to proceed with a gradual disinflation process. The
Before 1998 the strategy of the National Bank of Poland combined elements of the exchange rate targeting,
integration with the global economy created problems for such strategy, making it impossible to control both
inflation targeting, and money supply targeting applied to proceed with a gradual disinflation process. The
the monetary aggregates and the exchange rate. Therefore in 1998 the Monetary Policy Council (MPC)
integration with the global economy created problems for such strategy, making it impossible to control both
decided to introduce inflation targeting. Another important reason for adopting inflation targeting in Poland
the monetary aggregates and the exchange rate. Therefore in 1998 the Monetary Policy Council (MPC)
was the need to anchor inflation expectations, perceived as one of the main obstacles for the continuation of
decided to introduce inflation targeting. Another important reason for adopting inflation targeting in Poland
the disinflation process.
was the need to anchor inflation expectations, perceived as one of the main obstacles for the continuation of
Over the 15 years of inflation targeting, monetary policy of the National Bank of Poland has been
the disinflation process.
quite successful (Figure 1). The initial stage of this strategy was aimed at completing disinflation in the
Over the 15 years of inflation targeting, monetary policy of the National Bank of Poland has been
Polish economy. While adopting inflation targeting the MPC set the medium-term inflation target at a
quite successful (Figure 1). The initial stage of this strategy was aimed at completing disinflation in the
level below 4% at the end of 2003, but there were also short-term targets for the ends of subsequent years
Polish economy. While adopting inflation targeting the MPC set the medium-term inflation target at a
announced. Even if the National Bank of Poland faced difficulties in achieving short-term inflation targets
level below 4% at the end of 2003, but there were also short-term targets for the ends of subsequent years
and its monetary policy in some periods was too restrictive or too lax [Kokoszczyński et al. (2006)], inflation
announced. Even if the National Bank of Poland faced difficulties in achieving short-term inflation targets
went down: while at the beginning of 1998 it was about 13%, by the end of 2003 it had fallen to 1.5%.
and its monetary policy in some periods was too restrictive or too lax [Kokoszczyński et al. (2006)], inflation
After completing the disinflation process monetary authorities in Poland were confronted with the goal of
went down: while at the beginning of 1998 it was about 13%, by the end of 2003 it had fallen to 1.5%.
strengthening price stability and making inflation expectations of economic agents firmly anchored. Since
After completing the disinflation process monetary authorities in Poland were confronted with the goal of
2004 the central bank has targeted inflation at 2.5% (with a tolerance band of 1 pp.) and in fact inflation
strengthening price stability and making inflation expectations of economic agents firmly anchored. Since
and inflation expectations in the Polish economy stabilized at a relatively low level. The IT framework
2004 the central bank has targeted inflation at 2.5% (with a tolerance band of 1 pp.) and in fact inflation
performed quite well when Poland faced external shocks, including its entry to the European Union (May
and inflation expectations in the Polish economy stabilized at a relatively low level. The IT framework
2004), oil and food prices shocks (2007-2008), and the sharp phase of the financial crisis (2008-2009).
performed quite well when Poland faced external shocks, including its entry to the European Union (May
Figure
2004), oil and food prices shocks (2007-2008), [here:
and the
sharp1]phase of the financial crisis (2008-2009).
[here: Figure 1]
3
3
Assessing central bank transparency
Over
the last 15 years,
the National
Banktransparency
of Poland has been putting more and more stress on developing
Assessing
central
bank
effective communication with the public. Transparency indices proposed by Eijffinger and Geraats (2006)
Over the last 15 years, the National Bank of Poland has been putting more and more stress on developing
(EG) as well as Mahadeva and Sterne (2000) (MS) show that the degree of information openness of the NBP
effective communication with the public. Transparency indices proposed by Eijffinger and Geraats (2006)
has become relatively high after the Polish central bank started publishing its staff inflation projections with
(EG) as well as Mahadeva and Sterne (2000) (MS) show that the degree of information openness of the NBP
a formal risk assessment (Table 1, Table 2). It took place in August 2004 and coincided with significant
has become relatively high after the Polish central bank started publishing its staff inflation projections with
changes in the structure of Polish Inflation Reports. Previously the public did not have access to inflation
6
forecasts prepared at the bank and the forward-looking orientation of monetary policy was significantly
less pronounced than in other central banks pursuing
6 inflation targeting. Also the quality of information
released was insufficient [e.g. Fracasso et al. (2003)].8 Central bank transparency, as measured with both
8
Narodowy Bank Polski
transparency indices, increased again in 2007, when the NBP started releasing minutes that present the main
issues discussed, arguments put forward and the decisions taken at the MPC meetings.9
released was insufficient [e.g. Fracasso et al. (2003)].8 Central bank transparency, as measured with both
changes in the structure of Polish Inflation Reports. Previously the public did not have access to inflation
transparency indices, increased again in 2007, when the NBP started releasing minutes that present the main
forecasts prepared at the bank and the forward-looking orientation of monetary policy was significantly
issues discussed, arguments put forward and the decisions taken at the MPC meetings.9
less pronounced than in other central banks pursuing inflation targeting. Also the quality of information
aa formal
formal risk
risk assessment
assessment (Table
(Table 1,
1, Table
Table 2).
2). It
It took
took place
place in
in August
August 2004
2004 and
and coincided
coincided with
with significant
significant
8
Assessing
bank transparency
[here: Table
1] bank transparency, as measured with both
released central
was insufficient
[e.g. Fracasso et al. (2003)].
Central
changes
changes in
in the
the structure
structure of
of Polish
Polish Inflation
Inflation Reports.
Reports. Previously
Previously the
the public
public did
did not
not have
have access
access to
to inflation
inflation
transparency indices, increased again in 2007, when the NBP started releasing minutes that present the main
forecasts
of
[here: Tableorientation
2]
forecasts prepared
prepared at
at the
the bank
bank and
and the
the forward-looking
forward-looking
orientation
of monetary
monetary policy
policy was
was significantly
significantly
issues discussed, arguments put forward and the decisions taken at the MPC meetings.9
less
less pronounced
pronounced than
than in
in other
other central
central banks
banks pursuing
pursuing inflation
inflation targeting.
targeting. Also
Also the
the quality
quality of
of information
information
Analysing effectiveness of central bank communication
with the public we analyse predictability of short8
8 Central bank transparency, as measured with both
released
Central
released was
was insufficient
insufficient [e.g.
[e.g. Fracasso
Fracasso et
et al.
al. (2003)].
(2003)].
[here: Table
1] bank transparency, as measured with both
term interest rates, influenced by the NBP interest rate decisions. Interest rate predictability is measured
transparency
indices,
increased
again
in
2007,
when
the
NBP
transparency indices, increased again in 2007, when the NBP started
started releasing
releasing minutes
minutes that
that present
present the
the main
main
10
on the
basis
ofinflation
forward projections,
rate agreement
(FRA)
contracts.
Alternatively
wethis
follow
L
Ž yziak
et al.a(2007)
[here:
Table
2]2008.
the
NBP
staff
as
observed
since
late
To
analyse
effect
we
define
given
9
issues
issues discussed,
discussed, arguments
arguments put
put forward
forward and
and the
the decisions
decisions taken
taken at
at the
the MPC
MPC meetings.
meetings.9
11
using
Reuters
survey
data
on
interest
rate
expectations
of
financial
sector
agents
in this case
– due to
interest rate decision, i.e. either a change of the NBP reference rate (i) or keeping ,itbut
constant,
as consistent
Analysing effectiveness of central bank communication with the public we analyse predictability of short[here:
Table
1]
f 12
changes
survey
questions
– we(πare
to cover
till April
[here:a shorter
Table
1]period,
with
thein
NBP
inflation
forecast
) able
if the
difference
between
this i.e.
forecasts
and2009.
the NBP inflation target
term interest rates, influenced by the NBP interest rate decisions. Interest rate predictability is measured
We
can observe
thatresponses
advancesofinthe
information
disclosure
(π tar
) implies
adequate
central[here:
bank,
i.e.132]
: practices were accompanied by an increase in
Table
[here:
Table 10
2] Alternatively we follow L
Ž yziak et al. (2007)
on the basis of forward rate agreement (FRA)
contracts.
the predictability of the NBP interest rate decisions (Table 3, Figure 2). Even if the initial stage of inflation
using Reuters survey data on interest rate expectations of financial sector agents11 , but in this case – due to
Analysing
effectiveness
central
bank
with
public
we
of
f
targeting
in Poland
did notof
improvements
intarpredictability
monetary
policypredictability
actions [ŽLyziak
et (1)
al.
Analysing
effectiveness
ofbring
central
bankπcommunication
communication
with the
theof
public
we analyse
analyse
predictability
of shortshortt+4 πt+4 ⇒ it it−1
changes in survey questions – we are able to
cover a shorter period, i.e. till April 2009.
term
interest
rates,
influenced
the
interest
rate
Interest
rate
predictability
(2007)],
publishing
bankby
inflation
projections
a substantial
difference
this respect. is
a result,
term
interest
rates,central
influenced
by
the NBP
NBP
interest made
rate decisions.
decisions.
Interest
rate in
predictability
isAsmeasured
measured
f
tar
We can observe that advances in information
disclosure
practices
were
accompanied
by
an
increase
in
=
π
⇒
i
=
i
(2)
π
10
t 10 Alternatively
t−1
t+4
t+4
on
basis
of
rate
contracts.
follow
L
ŽŽ yziak
al.
the the
mean
absolute
error (MAE)
of interest(FRA)
rate expectations
measured
on thewe
basis
of financial
market
data
on
the
basis
of forward
forward
rate agreement
agreement
(FRA)
contracts.
Alternatively
we
follow
L
yziak et
et
al. (2007)
(2007)
the predictability of the NBP interest rate decisions (Table 3, Figure 2). Even if 11
the initial stage of inflation
11 , but in this case – due to
using
Reuters
survey
on
rate
of
financial
sector
agents
f beforetar
declined
by 60%
withdata
respect
to the period
the
first
publication
of
inflation
projection
by the
NBP,
using
Reuters
survey
data
on interest
interest
rateπexpectations
expectations
of
financial
sector
agents
, but
in this case
– due
to
≺
π
⇒
i
≺
i
(3)
t
t−1
t+4
targeting in Poland did not bring improvements
int+4
predictability of monetary policy actions [ŽLyziak et al.
changes
survey
questions
–
able
to
aa shorter
period,
i.e.
April
while
thein
absolute
percentage
more than
40%.
In 2009.
the case of Reuters survey
changes
inmean
survey
questions
– we
we are
areerror
able(MAPE)
to cover
cover declined
shorterby
period,
i.e. till
till
April
2009.
(2007)],
publishing
central
bank
inflation
projections
made
a
substantial
difference
infinancial
this respect.
Asthe
a result,
The
results
of
our
analysis
(Figure
3)
indicate
that
since
the
beginning
of
the
crisis
MPC
We
can
observe that
that
advances
in information
information
disclosure
practices
were
accompanied
by
an increase
increase
in
dataWe
thecan
improvement
is even
stronger
in terms of disclosure
MAE, while
similar were
as faraccompanied
as MAPE is by
concerned.
observe
advances
in
practices
an
in
the mean
absolute less
errorattention
(MAE) of
interest
rate expectations
measured
on
the basis It
of can
financial
market data
has
paid
relatively
to
central
bank
inflation
forecasts
than
previously.
be
explained
with
the
the predictability
predictability of
of the
the NBP
NBP interest
interest rate
rate decisions
decisions (Table
(Table 3,
3, Figure
Figure 2).
2). Even
Even if
if the
the initial
initial stage
stage of
of inflation
inflation
[here:
Table
3]
declined
by
60%
with
respect
to
the
period
before
the
first
publication
of
inflation
projection
by
theweight
NBP,
the
fact
that
financial
crisis
disturbances
led
to
changes
in
the
monetary
policy
rule,
increasing
the
targeting in
in Poland
Poland did
did not
not bring
bring improvements
improvements in
in predictability
predictability of
of monetary
monetary policy
policy actions
actions [Ž
[ŽL
Lyziak
yziak et
et al.
al.
targeting
while the to
mean
absolute
percentage
error
(MAPE)
declined
by more
than
40%.
In the case
of(2012)].
Reuters survey
attached
output
stabilization,
even
though
inflation
exceeded
the
target
[Demchuk
et
al.
(2007)],
difference
[here: made
Figureaa substantial
2]
(2007)], publishing
publishing central
central bank
bank inflation
inflation projections
projections
made
substantial
difference in
in this
this respect.
respect. As
As aa result,
result,
data the improvement is even stronger in terms of MAE, while similar as far as MAPE is concerned.
the
mean
absolute
error
(MAE)
of
interest
rate
expectations
measured
on
the
basis
of
financial
market
data
the mean absolute error (MAE) of interest rate[here:
expectations
Figure 3]measured on the basis of financial market data
Interestingly, the introduction of the MPC minutes in 2007 was not accompanied by a further improvement
declined
first
declined by
by 60%
60% with
with respect
respect to
to the
the period
period before
before
the
first3]publication
publication of
of inflation
inflation projection
projection by
by the
the NBP,
NBP,
[here:the
Table
in the predictability of interest rate decisions and the relative forecast errors (MAPE) even increased. It
while
the
mean
absolute
percentage
error
(MAPE)
declined
by
more
than
40%.
In
the
case
of
Reuters
survey
while the mean absolute percentage error (MAPE) declined by more than 40%. In the case of Reuters survey
4
central
bankterms
credibility
seemsAssessing
that this result is
due to substantial
uncertainty
concerning
the behaviour of short-term interest rates
[here: Figure
2]
data
data the
the improvement
improvement is
is even
even stronger
stronger in
in terms of
of MAE,
MAE, while
while similar
similar as
as far
far as
as MAPE
MAPE is
is concerned.
concerned.
at the beginning of the financial crisis in 2008. However, there is another factor connected with monetary
Interestingly,
the introduction
of the
minutes
in 2007
was not accompanied
further improvement
To
assess central
bank credibility
inMPC
Poland
we refer
to 3]
survey-based
measuresby
of ainflation
expectations
[here:
Table
policy that can contribute to explaining this effect,
inconsistency of interest rate decisions with
[here:namely
Table some
3]
in the predictability
of interest
rateenterprises
decisions and
relative forecast
errors
(MAPE) even
increased.
It
formed
by financial sector
analysts,
and the
consumers.
The sample
of observations
available
for all
main
weaknesses
mentioned
included:
information
concerning
about
the8 The
NBP
staff
inflation
projections,
as incompleteness,
observed sinceinsufficient
late 2008.
To analyse
this committee
effect we discussions
define a given
[here:
Figure
2]
policy
and
objections
thatiscould
be
to decisions,
few
arguments
future
decisions presented.
[here:
Figure
2] for the
seems
that
this of
result
dueexpectations
to raised
substantial
uncertainty
concerning
behaviour
of short-term interest rates
the
time
series
inflation
starts
intooJanuary
2001.
9 In therate
interest
decision,
i.e. and
either
a change(2013)
of the
NBP reference
rate
(i) or
keeping
constant,
consistent
recent
study Dincer
Eichengreen
calculate
transparency
indices
based
on the it
Eijffinger
and as
Geraats
(2006)
at the
beginning
of 100
the economies
financial
crisissector
in 2008.
However,
there
is another
factor
connected
with
monetary
concept
for more
than
from
1998
through
2010.
Their
assessment
is based
on information
on
central
banks’
Inflation
expectations
of
financial
analysts
are
based
on
monthly
Reuters
surveys,
while
inflation
f the
12 MPC minutes in 2007 was not accompanied by a further improvement
Interestingly,
the
introduction
of
withInterestingly,
theand
NBP
inflation
forecast
(πofother
)the if
the minutes
difference
this
and by
theby
target
the
introduction
MPC
in between
2007
notforforecasts
accompanied
a NBP
further
improvement
websites
statues,
annual
reports and
published
documents.
Thewas
scores
Poland obtained
Dincerinflation
and
Eichengreen
14 interest rate decisions with
policy are
that
can
contribute
tocalculated
explaining
thispaper,
effect,
namely
some
inconsistency
of
expectations
of
Polish
enterprises
–
on
the
NBP
surveys
conducted
quarterly.
We
use
a
new
measure
of
(2013)
lower
than
EG
index
in
this
but
in
terms
of
their
evolution
in
time
they
confirm
a
positive
trend
tar
13
in
the) implies
predictability
of interest
interest
rate
decisions
and
thei.e.
relative
forecast errors
errors (MAPE)
(MAPE) even
even increased.
increased.
It
(π central
adequate
responses
of the
the
central
bank,
:
in
the
predictability
of
rate
decisions
and
the
relative
forecast
It
in
bank
transparency
in
Poland
in
period
under
analysis.
8
main
weaknesses
mentioned
included:
incompleteness,
insufficient
information
concerning
committee
discussions
about
10 The
consumer
inflation
expectations
quantified
on
the
basis
of
the
consumer
survey
conducted
every
month
by
the
We
consider
FRA
1x2
contracts
as
a
proxy
for
expected
WIBOR
1M
in
a
1-month
horizon.
seems
that
this
due
to
substantial
uncertainty
concerning
the
behaviour
of
policy
thatis
be
raised
to 3M
decisions,
too fewforarguments
formonth.
future
decisions presented.
seems
that
this result
result
iscould
due on
toWIBOR
substantial
uncertainty
concerning
the
behaviour
of short-term
short-term interest
interest rates
rates
11 Weand
useobjections
Reuters
survey
data
expectations
the current
15(2013)
9 In the recent study Dincer and Eichengreen
calculate
transparency
indices
based
on
the
Eijffinger
and
Geraats (2006)
Polish
Central
Statistical
Office
(GUS).
Quantifying
inflation
expectations
of
consumers
and
enterprises
we
f
tar
at the
the beginning
beginning
of 100
the economies
financial from
crisis1998
in π2008.
2008.
However,
there
is another
another
factor
connected on
with
monetary
π2010.
itthere
assessment
it−1
(1)
at
of
the
financial
crisis
in
is
factor
connected
with
monetary
concept
for more than
through
is based
on information
central
banks’
t+4 ⇒Their
t+4 However,
use
thethat
probability
method,
bynamely
CarlsonThe
and
Parkin
(1975)
and developed
by Eichengreen
Batchelor
websites
and
statues, annual
reports
and otheroriginally
published documents.
scores
for Poland
obtained
by Dincer and
policy
can
contribute
to proposed
explaining
this
effect,
some
inconsistency
of
rate
with
policy are
that
canthan
contribute
explaining
thisfpaper,
effect,
namely
inconsistency
of interest
interest
rate decisions
decisions
with
(2013)
lower
EG indextocalculated
in this
but
of their
evolution in time
they confirm
a positive trend
7 termssome
tarin
=
π
⇒
i
=
i
(2)
π
and
Orr
(1988).
In
the
case
of
consumers
we
use
the
subjectively
perceived
inflation
–
so-called
Consumer
t
t−1
8 The main
in central
bankweaknesses
transparency
in Poland
in the period
analysis.
t+4undert+4
mentioned
included:
incompleteness,
insufficient information concerning committee discussions about
8
main weaknesses
included:
incompleteness,
insufficient
information
concerning
committee discussions about
10 The
Weand
consider
FRA that
1x2mentioned
contracts
as
a proxy
for expected
WIBOR
1M in
a 1-month
horizon.presented.
policy
objections
could
be
raised
decisions,
too
few
for
future
policy
and
objections
that (CPPI)
could
be WIBOR
raised
toand
decisions,
too(2013)]
fewforarguments
arguments
formonth.
future decisions
decisions
presented.
11
Perceived
Price Index
[Ha¡lkato
L
¡fyziak
–
as
the
scaling
factor
in
the probability method.
9 We
use
Reuters
survey
data
on
3M
expectations
the
current
tar transparency indices based on the Eijffinger and Geraats (2006)
9 In
≺ πt+4
⇒ it ≺ it−1 indices based on the Eijffinger and Geraats (2006)
(3)
πt+4 calculate
In the
the recent
recent study
study Dincer
Dincer and
and Eichengreen
Eichengreen (2013)
(2013)
calculate
transparency
concept
for
more
than
100
economies
from
1998
through
2010.
Their
assessment
is
based
on
information
on
banks’
concept
for more
100 economies
from 1998
through
assessment
based on information
on central
central
banks’
It
exceeds
CPIthan
inflation
significantly
(5.1%
vs. 2010.
3.1% Their
on average
in is2001-2012)
due to the
fact that
in
websites
websites and
and statues,
statues, annual
annual reports
reports and
and other
other published
published documents.
documents. The
The scores
scores for
for Poland
Poland obtained
obtained by
by Dincer
Dincer and
and Eichengreen
Eichengreen
(2013)
are
lower
EG
index
calculated
in
paper,
but
in
terms
of
their
evolution
in
confirm
a
trend
The
ofofour
analysis
(Figure
3) indicate
that
since
oftime
the they
financial
crisis
the MPC
their
perception
price
developments
Polish
consumers
focus
on beginning
prices
of frequently
purchased
goods
and
(2013)
areresults
lower than
than
EG
index
calculated
in this
this
paper,
but
in
ofthe
their
evolution
in
time
they
confirm
a positive
positive
trend
7 terms
in central
central bank
bank transparency
transparency in
in Poland
Poland in
in the
the period
period under
under analysis.
analysis.
in
10 We consider
16
FRA
1x2
a
for
expected
WIBOR
1M
1-month
horizon.
10 We
has
paid
relatively
attention
central
forecasts
previously.
and disregard
negativeas
price
changes
ofinflation
these
items.
services
consider
FRA less
1x2 contracts
contracts
asto
a proxy
proxy
forbank
expected
WIBOR
1M in
in a
a than
1-month
horizon. It can be explained with
11 We use Reuters survey data on WIBOR 3M expectations for the current month.
11 We use Reuters survey data on WIBOR 3M expectations for the current month.
12 We
the
facttake
that
disturbances
led to
intofinancial
account acrisis
constant
forecast horizon
of 4changes
quarters.in the monetary policy rule, increasing the weight
13 This mechanical rule is consistent with the concept of inflation forecast targeting [Svensson (2003)]. In a few cases we made
attached to
output
evenrate
though
inflation
the inflation
target [Demchuk
al. (2012)].
exceptions
from
it, e.g.stabilization,
if keeping interest
constant,
despite exceeded
a gap between
forecast andettarget,
occurred after a
77
sequence of interest decisions consistent with the sign of this gap.
14 We use the main measure of enterprise inflation expectations proposed in L
€ yziak (2012).
15 GUS survey data on consumer opinions on future price
[here:developments
Figure 3] start in January 2004. To obtain a longer time series
of inflation expectations we extrapolate the balance statistic of GUS survey responses that is used in the quantification of
inflation expectations, exploiting the relationship between GUS survey data and Ipsos survey data, covering a longer sample
period.
16 It should be noted however that they take into consideration a relatively broad basket of goods and services, including food
9
NBP Working Paper No. 162
4
Assessing central bank credibility
To assess central bank credibility in Poland we refer to survey-based measures of inflation expectations
the
staff inflation
inflation forecast
projections,
as ifobserved
since late
2008.this
Toforecasts
analyse and
this the
effect
we inflation
define a target
given
withNBP
the NBP
(π f )12
the difference
between
NBP
interest
rate decision,
i.e.responses
either a of
change
of the bank,
NBP reference
rate (i) or keeping it constant, as consistent
(π tar ) implies
adequate
the central
i.e.13 :
with the NBP inflation forecast (π f )12 if the difference between this forecasts and the NBP inflation target
Chapter 4
f
tar i.e.13 :
(π tar ) implies adequate responses of the central
bank,
πt+4
⇒ it it−1
πt+4
(1)
tar
(2)
π ff = πt+4
tar ⇒ it = it−1
πt+4
it 2008.
it−1 To analyse this effect we define a given
(1)
the NBP staff inflation projections, as observed
late
t+4 ⇒
t+4 πsince
tar reference rate (i) or keeping it constant, as consistent
interest rate decision, i.e. either a change πofff the≺NBP
πt+4
(3)
tar ⇒ it ≺ it−1
=
π
⇒ it = it−1
(2)
πt+4
t+4
t+4
f 12
with
the NBP
(π ) as ifobserved
the difference
between
NBP
the NBP
staff inflation
inflation forecast
projections,
since late
2008.this
Toforecasts
analyse and
this the
effect
we inflation
define a target
given
The results of our analysis (Figure 3) indicate
that since
the beginning of the financial crisis the MPC
f
tar
πt+4
⇒
(3)
πoft+4
(π tar ) implies
adequate
the central
i.e.i13
: it−1
t ≺
interest
rate decision,
i.e.responses
either a of
change
the≺bank,
NBP
reference
rate (i) or keeping it constant, as consistent
has paid relatively less attention tofcentral
bank
inflation
forecasts
than
previously.
It
can
be
explained
with
with the NBP inflation forecast (π )12 if the difference between this forecasts and the NBP inflation target
The
of our analysis
(Figure 3) led
indicate
that since
themonetary
beginning
of the
financial
crisisthe
theweight
MPC
thetar
fact results
that financial
crisis disturbances
in the
policy
rule,
increasing
f to changes
tar
bank,
πt+4
⇒
(1)
πt+4
t (π ) implies adequate responses of the central
i.e.i13
: it−1
has
paid relatively
attention to
central
bank
inflation
forecasts
than
previously.
It can
be(2012)].
explained with
attached
to outputless
stabilization,
even
though
inflation
exceeded
the
target
[Demchuk
et al.
f
tar
the fact that financial crisis disturbances πled
the
monetary policy rule, increasing the weight
=changes
πt+4
⇒ in
it =
it−1
(2)
t+4to
f
tar
πt+4
⇒
i
i
(1)
πt+4[here:
t
t−1
Figure
3] the target [Demchuk et al. (2012)].
attached to output stabilization, even though
inflation
exceeded
f
tar
(3)
πt+4 ≺ πt+4 ⇒ it ≺ it−1
f
tar
= πt+4
⇒ it = it−1
(2)
πt+4
[here: Figure 3]
results of our central
analysis (Figure
3) indicate
that since the beginning of the financial crisis the MPC
4 The
Assessing
bank
credibility
f
tar
≺ πt+4
⇒ it ≺ it−1
(3)
πt+4
has paid relatively less attention to central bank inflation forecasts than previously. It can be explained with
ToAssessing
assess central bank
credibility
in Poland
we refer to survey-based measures of inflation expectations
4 The
bank
credibility
the
fact results
that financial
crisis disturbances
to changes
in the
policy
rule,
increasing
of our central
analysis
(Figure
3) led
indicate
that since
themonetary
beginning
of the
financial
crisisthe
theweight
MPC
formed by financial sector analysts, enterprises and consumers. The sample of observations available for all
attached
to outputless
stabilization,
even
though
inflation
exceeded
the
target
[Demchuk
et al.
has paid relatively
attention to
central
bank
inflation
forecasts
than
previously.
It can
be(2012)].
explained with
assess
bank credibility
in starts
Polandinwe
refer to2001.
survey-based measures of inflation expectations
the To
time
seriescentral
of inflation
expectations
January
the fact that financial crisis disturbances led to changes in the monetary policy rule, increasing the weight
formed
by financial
sectorofanalysts,
and consumers.
The
sample Reuters
of observations
for all
[here:
Figure
3] on
Inflation
expectations
financialenterprises
sector analysts
are based
monthly
surveys,available
while inflation
attached to output stabilization, even though inflation exceeded the target [Demchuk et al. (2012)].
the
time series
inflation
expectations
starts
in January
2001.
expectations
of ofPolish
enterprises
– on the
NBP
surveys conducted
quarterly.14 We use a new measure of
Inflation
expectations
of financial
sector
are
based
on monthly
Reuters
surveys,
inflation
consumer
inflation
expectations
quantified
onanalysts
the
basis
of the
survey
conducted
everywhile
month
by the
[here:
Figure
3]consumer
4
Assessing central bank credibility
4
Assessing central bank credibility
14
expectations
Polish enterprises
– on15the
NBP surveys
conducted
quarterly.
We use and
a new
measure we
of
Polish
CentralofStatistical
Office (GUS).
Quantifying
inflation
expectations
of consumers
enterprises
consumer
inflation
expectations
quantified
on thewe
basis
of to
thesurvey-based
consumer
conducted
every month
by the
To
central
bank credibility
inoriginally
Poland
refer
measures
inflation
expectations
use
theassess
probability
method,
proposed
by
Carlson
and
Parkinsurvey
(1975)
andofdeveloped
by
Batchelor
15
Polish
Central
Statistical
Office
(GUS).
Quantifying
inflation
expectations
consumers
andavailable
enterprises
formed
by
financial
sector
analysts,
enterprises
andthe
consumers.
The
sample of
of
observations
for we
all
and
Orr
(1988).
In the
case
of consumers
we use
subjectively
perceived
inflation
– so-called
Consumer
use
theassess
probability
method,
proposed
originally
by
Carlson
and
developed
by
Batchelor
the To
time
series
ofIndex
inflation
expectations
starts
inwe
January
central
bank
credibility
Poland
refer
to
survey-based
measures
inflation
expectations
Perceived
Price
(CPPI)
[Ha¡lka in
and
L
¡ yziak
(2013)]
–2001.
as
theParkin
scaling(1975)
factorand
in of
the
probability
method.
and
Orrby
(1988).
In the
case
of consumers
weanalysts
use
subjectively
perceived
inflation
– to
so-called
Consumer
Inflation
expectations
ofanalysts,
financial
sector
are
based
on
monthly
Reuters
surveys,
while
inflation
formed
financial
sector
enterprises
andthe
consumers.
The
sample
of
observations
available
for all
It
exceeds
CPI
inflation
significantly
(5.1%
vs.
3.1%
on
average
in
2001-2012)
due
the
fact
that
in
14 in the probability method.
Perceived
Price
Index
(CPPI)
[Ha¡l–kaonand
L
¡ NBP
yziak
(2013)] focus
–conducted
as the
scaling
expectations
of ofPolish
enterprises
the
surveys
quarterly.
We use
a new measure
of
the
time
series
expectations
starts
in January
2001.
their
perception
ofinflation
price
developments
Polish
consumers
on prices
offactor
frequently
purchased
goods and
16 inflation
It
exceeds
CPI
inflation
significantly
(5.1%
vs.
3.1%
on
average
in survey
2001-2012)
due
to
thewhile
factinflation
that
in
consumer
expectations
quantified
onanalysts
the
ofitems.
the
consumer
conducted
every
month
by the
Inflation
expectations
of financial
are
based
on monthly
Reuters
surveys,
and
disregard
negative
pricesector
changes
of basis
these
services
15
14
their
of price
developments
Polish
consumers
focus
on
of frequently
purchased
goods
and
Polish
Central
Office forecast
(GUS).
Quantifying
inflation
expectations
of consumers
and
enterprises
we
12 We
expectations
ofStatistical
Polish
– onhorizon
the
NBP
surveys
conducted
We use
a new
measure
of
take
account
constant
of
4 use
quarters.
Allperception
the into
measures
ofaenterprises
inflation
expectations
we
are
formed
in prices
a quarterly.
12-month
horizon.
However,
in testing
13 This
mechanical rule is consistent with the concept of inflation forecast targeting [Svensson (2003)]. In a few cases we made
13 This
15
mechanical rule is consistent with the concept of inflation forecast targeting [Svensson (2003)]. In a few cases we made
and
disregard
negative
price
changes
ofshould
these
services
use
the 16
probability
method,
proposed
originally
by
Carlson
and
Parkin
(1975)
and
developed
by follow
Batchelor
consumer
inflation
quantified
on the
basis
ofitems.
the
consumer
survey
conducted
every occurred
month
by
thea
central
bank
credibility,
long-term
expectations
considered.
To
solve
thisand
problem
we
exceptions
from
it, e.g.expectations
if keeping
interest
rate
constant,
despite
abe
gap
between
inflation
forecast
target,
after
sequence
of
interest
decisions
consistent
with
the
sign
of
this
gap.
12 We take into account a constant forecast horizon
15 we of
4 quarters.
and
Orr
(1988).
Inmeasure
the
of (2012),
consumers
use
the
subjectively
perceived
so-called
Consumer
Polish
Central
Statistical
Office
(GUS).
inflation
expectations
consumers
and enterprises
we
14 We
approach
by L
case
yziak
i.e.Quantifying
we
use
the
adjusted
ofofinflation
the
NBP– inflation
target.
It
use proposed
the main
of enterprise
inflation
expectations
proposed
inmeasure
L
€ yziak (2012).
GUS survey
data
onifconsumer
opinions
on
future
price
developments
startscaling
ininflation
January
2004.inTo
obtain
a longer
time
series
Perceived
Price
Index
(CPPI)
[Ha¡lkarate
and
L
¡ yziak
(2013)]
– asbetween
the
factor
the
probability
method.
exceptions
from
it,
e.g.
keeping
interest
constant,
despite
a gap
forecast
and
target,
occurred
after
a
use
the
probability
method,
proposed
originally
by
Carlson
Parkin
(1975)
and
developed
by
Batchelor
takes
into
account
the
of
monetary
policy,
which
implies
thatin
ifthethe
current
rate
of
inflation
expectations
wemedium-term
extrapolate
theorientation
balance
of GUS and
survey
responses
that
is used
quantification
of
sequence
of interest
decisions
consistent with
the
sign statistic
of this gap.
inflation
expectations,
exploiting
the relationship
between
GUS survey
data
survey data,
longerthat
sample
It
CPI
inflation
significantly
(5.1%
vs.
3.1%
oneconomic
average
inIpsos
2001-2012)
due–covering
to
thea fact
in
We
use
the
of shocks
enterprise
expectations
proposed
in and
L
€perceived
yziak
(2012).
and
Orr
(1988).
Inmeasure
case
of consumers
we
use target,
the
subjectively
inflation
so-called
Consumer
of14exceeds
inflation
– main
due
tothe
some
– inflation
exceeds
the
agents
setting
their
12-month
inflation
period.
15 GUS survey data on consumer opinions on future price developments start in January 2004. To obtain a longer time series
16 It should be noted however that they take into consideration a relatively broad basket of goods and services, including food
their
perception
ofbeprice
developments
Polish
consumers
focus
on prices
offactor
frequently
goods
and
Perceived
Price
(CPPI)
L
¡ yziak
(2013)]
– as
the
scaling
inused
thepurchased
probability
method.
of
inflation
expectations
we
extrapolate
the
balance
statistic
of GUS
responses
thattargeting,
is
in
the
of
expectations
canIndex
aware
that[Ha¡
inlka
lineand
with
the
principles
ofsurvey
flexible
inflation
thequantification
observed
gap
inflation 16
expectations, exploiting the relationship between GUS survey data and Ipsos survey data, covering a longer sample
and
disregard
negative
price
changes
of
these
items.
services
It
exceeds
CPI inflation
(5.1%
3.1% on
average by
in the
2001-2012)
due to
the infact
that in
between
inflation
and the significantly
inflation target
will vs.
be reduced
gradually
central bank,
hence
12 months
period.
16 It
12
should be noted however that they take into consideration a relatively broad basket of goods and services, including food
Weperception
take
into
account
aabove
constant
4 quarters.
their
ofstay
price
developments
Polish
focus
prices
of frequently
goods
and
inflation
can
still
theforecast
targethorizon
even ifofconsumers
the
central
bankon
takes
actions
to bring purchased
it back to the
target.
13
8
This mechanical rule is consistent with the concept of inflation forecast targeting [Svensson (2003)]. In a few cases we made
16 from it, e.g. if keeping interest rate constant, despite a gap between inflation forecast and
e target, occurred after a
exceptions
and
negative
price
of these items.
services
In the
firstdisregard
step of our
analysis
wechanges
assess deviations
of inflation expectations (πt+n||t
denotes inflation
sequence of interest decisions consistent with the sign of this gap.
14 We take
12
use the
main
measure
of n
enterprise
inflation
expectations
proposed
in L
€ yziak
(2012).
into
account
a constant
forecast
of t)
4 quarters.
expectations
for
period
t+
formed
inhorizon
period
from
inflation
target
of the National Bank of Poland
8 the
15 This
13
GUS mechanical
survey datarule
on consumer
opinions
on concept
future price
developments
in January
2004.(2003)].
To obtain
longer
time
is consistent
with the
of inflation
forecaststart
targeting
[Svensson
In aafew
cases
we series
made
tar
of
expectations
extrapolate
balance
statistic
of aGUS
responses
is used
in
the quantification
exceptions
from
it,calculating
e.g. ifwe
keeping
interestthe
rate
constant,
despite
gap
between
inflation
forecast
and
target,
occurred
after of
a
(πtinflation
). Except
standard
measures
of this
kind,
i.e.survey
absolute
and that
absolute
percentage
differences
inflation expectations,
exploiting
the relationship
between
GUS
sequence
of interest decisions
consistent
with the sign
of this
gap.survey data and Ipsos survey data, covering a longer sample
14 We useboth
between
variables,
credibility
indices proposed
and Krause (2002) that punish
period.
the main
measurewe
of derive
enterprise
inflation expectations
proposedby
in Cecchetti
L
€ yziak (2012).
16 GUS
15
It should
be noted
however
thatopinions
they take
consideration
a relatively
broad
basket of
goods
services,
including
food
survey
data on
consumer
oninto
future
price developments
start
in January
2004.
Toand
obtain
a longer
time series
only
for positive
deviations
of inflation
expectations
thesurvey
central
bank target:
of inflation
expectations
we extrapolate
the balance
statisticfrom
of GUS
responses
that is used in the quantification of
inflation expectations, exploiting the relationship between GUS survey data and Ipsos survey data, covering a longer sample

period.
16 It should be noted however

goods and services, including food
that they take into consideration a relatively broade basket of tar
10
ICt =




1−






1
1
tar
0.2−πt+n
8
e
tar
πt+n||t
− πt+n
0
8
if
πt+n||t ≤ πt+n
if
tar
e
πt+n
< πt+n||t
< 20%
if
e
πt+n||t
≥ 20%
Narodowy Bank Polski
(4)
e
In into
the first
stepthe
of our
analysis weorientation
assess deviations
of inflation
inflation
takes
account
medium-term
of monetary
policy, expectations
which implies(πthat
the current
rate
t+n||tif denotes
expectations
t+n
formed
in periodthe
t) target,
from theeconomic
inflationagents
target setting
of the National
Bank ofinflation
Poland
of
inflation –for
dueperiod
to some
shocks
– exceeds
their 12-month
(πttar ). Except
calculating
measures
of this
kind, i.e.
absoluteinflation
and absolute
percentage
differences
expectations
can
be aware standard
that in line
with the
principles
of flexible
targeting,
the observed
gap
between inflation
both variables,
derive credibility
proposed
by Cecchetti
and Krause
(2002)inthat
punish
and theweinflation
target will indices
be reduced
gradually
by the central
bank, hence
12 months
expectations we use are formed in a 12-month horizon. However, in testing
only
for positive
inflation
from the
central
target:
inflation
can still deviations
stay aboveofthe
target expectations
even if the central
bank
takes bank
actions
to bring it back to the target.
central bank credibility, long-term expectations should be considered. To solve this problem we follow the
e
In the price
first step
of our
assess
deviations
of inflation
expectations
perceived
changes,
while
in thewe
case
of inflation
expectations
of enterprises
and(πfinancial
sector inflation
analysts
 analysis
t+n||t denotes
approach proposed by L

yziak (2012), i.e. we use the adjusted measure of the NBP
inflation target. It
e
tar

1
expectations
period
t
+ nrespectively.
formed in period
t) from the if
inflationπtarget
ofπthe
it amounts tofor
79%
and 93%

t+n National Bank of Poland
t+n||t ≤


takes into account the medium-term
orientation
of monetary
policy, which implies that if the current rate
1measures
e of this tar
tar
e
(πttar ). Except calculating
kind, i.e.
absolute
ICt = standard
(4)
πt+n||t
if absolute
πt+n
1 − 0.2−π
− πt+n
< πand
< 20% percentage differences
tar
t+n||t

of inflation – due to some
shocks
– t+n
exceeds
the
target,
economic
agents
setting their 12-month inflation

[here:
Table
5]


between both variables, we
derive credibility indices proposed by Cecchetti
and Krause (2002) that punish

e
 that
0 the principles if
πt+n||t
≥ 20%
expectations can be aware
in line with
of flexible
inflation
targeting, the observed gap
onlyRemarkably,
for positive inflation
deviations
of inflation
expectations
theanalysts
central bank
target:closer to the NBP inflation
forecasts
formed
by financialfrom
sector
are usually
between
inflation
and that
the inflation
target
will be
reduced gradually
bydeviate
the central
hencetarget
in 12 significmonths
We
find on
this basis
consumer
inflation
expectations
in Poland
frombank,
the NBP
 inflation projections (Figure 6). Given that the NBP projections assume constant
target than to the NBP staff
inflation
canexpectations
still stay above
the remaining
target evengroups
if the of
central
bank agents
takes actions
to bring
it back to the target.

antly,
while
of the
economic
e – especially
tar financial sector agents –

≤ofπt+n
1 analysts
if
πt+n||t

interest rates it suggests 
that
financial
sector
observing
deviations
the
NBP
staff forecasts from
e

 analysis
thetofirst
of (Figure
our
we assess
deviations
Figure
of inflation expectations (πt+n||t denotes inflation
are In
close
the step
target
4, Table
4,
5).
1
e
tar
tar to the
e
(4)
t =NBP1actions
the target believe IC
that
will
bring
future
inflation
target
than it results from the
πt+n||t
if closer
πt+n
− 0.2−π
− πt+n
< πt+n||t
< 20%
tar
expectations for period t
+ n formed
in period
t) from
the inflation
target
of the National Bank of Poland
t+n




projection
currently available.
It confirms0 a high
degree ofif
credibility
financial sector
e
4]central
standard
πbank
20% among
(πttar ). Except calculating
measures [here:
of thisFigure
kind, i.e.
absolute
and≥absolute
percentage differences
t+n||t
analysts in Poland.18
between
both
weconsumer
derive credibility
proposed
by Cecchetti
Krause
(2002)
thatsignificpunish
[here: Table
4]
We
find on
thisvariables,
basis that
inflation indices
expectations
in Poland
deviateand
from
the NBP
target
only for
positive
deviations
inflation
expectations
from
the
target: financial sector agents –
antly,
while
expectations
of of
the
remaining
groups
economic
agentsbank
– especially
[here:ofFigure
6] central
[here: Figure 5]

are close to the target (Figure
4, Table 4, Figure 5).

e
tar


≤ πt+n
1
if
πt+n||t


In the second step we
estimate the equation suggested
by
Bomfim
and
Rudebusch
(2000), in which

4]
5 Conclusions
[here: Figure
1
e
tar
tar
e
IC
=
(4)
πt+n||t
if πinflation
1 − 0.2−πaverage
− πcentral
20%current price dynamics
tar
inflation expectationst are
a weighted
of the
bank
target<and
t+n
t+n < πt+n||t
t+n


17 results presented

e published by the National Bank of
that inflation
in this note suggest
(π 0 )The
, i.e.:
Table 4]
0 [here:
ifprojections
πt+n||t
≥ 20%
Poland have played the important role in making interest rate decisions more predictable. In the initial stage
We find on this basis that consumer inflation [here:
expectations
Figure in
5] Poland deviate from the NBP target significe
tar
of inflation targeting (1998-2004) – inspite
of= the
developments
tools used by the Polish
λπt+n
+ (1 − λ) πt0of+communication
εt
(5)
πt+n||t
antly, while expectations of the remaining groups of economic agents – especially financial sector agents –
central
bank
– thestep
predictability
of the
monetary
policy
decisions
limited
even deteriorating
yziak
In the
second
we estimate
equation
suggested
by was
Bomfim
andand
Rudebusch
(2000), in[”Lwhich
are close to results
the target
(Figure
Table
4, Figure
5).set by the National Bank of Poland has a strong impact
Estimation
confirm
that4,the
inflation
target
et al. (2007)].
Only theare
publication
inflationofforecasts
since
August
2004target
has contributed
improvements
inflation
expectations
a weightedofaverage
the central
bank
inflation
and currenttoprice
dynamics
on inflation expectations of financial sector analysts and enterprises, while a relatively small impact on
0
17
in this
[here: Figure 4]
) , respect.
i.e.:
(π
perceived price
changes,
while in the
case 5).
of inflation
expectations
of enterprises
financial
analysts
consumer
inflation
expectations
(Table
The weight
of the NBP
inflation and
target
in thesector
formation
of
The note confirms previous findings concerning central credibility in Poland [i.e. L
” yziak (2012)], showing
it amounts
to 79% andis93%
Table 4] whose expectations rely mostly on currently
inflation
expectations
onlyrespectively.
15% in the
case[here:
of tar
consumers,
e sector analysts
that inflation expectations of financial
and
enterprises
= λπt+n
+ (1
− λ)
πt0 + εt display a high degree of anchoring
(5)
πt+n||t
and
non-alcoholic
tobacco,
andofenergy
carriers,
medical products,
fuels, communication
services,
perceived
price beverages,
changes, while
inhousing
the case
inflation
expectations
of enterprises
and financial
sectornewspapers
analysts
at the
NBP
while
inflation
expectations
are driven mainly by price movements
Figure
5] for details].
[here:
Table
5]
and
articles
andinflation
products target,
for personal
careconsumer
[see Halka [here:
and
L
yziak
(2013),
it 17amounts
and inflation
93%that
respectively.
Estimation
results
confirm
the inflation
target
set by perceived
the National
Bank
of Poland
has
a strong
impact
In the caseto
of79%
consumer
expectations
we use
subjectively
inflation
(CPPI)
instead of
current
CPI inflation.
currently perceived.
In
the second
step weforecasts
the
equation
suggested
by Bomfim
and aRudebusch
(2000),
in
which
Remarkably,
inflation
formed
by financial
sector
analysts
arewhile
usually
closer to the
NBP
inflation
on inflation
expectations
ofestimate
financial
sector
analysts
and enterprises,
relatively
small
impact
on
[here: Table 5]
inflation
expectations
are
a weighted
average
the
central
target
and
current
price
dynamics
target than
to the NBP
staff
inflation
projections
(Figure
Given
that the
NBP projections
assume
constant
consumer
inflation
expectations
(Table
5). of
The
weight
the inflation
NBP
inflation
target
in the
formation
of
9 6).ofbank
All the
measures
of inflation
Assessing
central
bank credibility
0 17
)Remarkably,
, i.e.:
(π
interest
rates
it suggests
that
financial
sector
analysts
observing
deviations
of thecloser
NBPto
staff
from
inflation
expectations
is only
15% formed
in the
case
of consumers,
whose
expectations
rely
mostly
on currently
inflation
forecasts
by
financial
sector
analysts
are usually
the forecasts
NBP
inflation
and
non-alcoholic
beverages,
tobacco,
housing
andbring
energyfuture
carriers,
medical
products,
fuels,
communication
the
target
that staff
NBP
actions
will
inflation
closer
the
target
than itservices,
resultsnewspapers
from
the
target
thanbelieve
to the
NBP
inflation
projections
(Figure
6).
Given
thatto
the
NBP
projections
assume
constant
and articles and products for personal care [see
Ha
l
ka
and
L
yziak
(2013),
for
details].
e
tar
0
17 In the case
λπ
+observing
(1 of
−
λ)
πt deviations
+inflation
εt
(5)
πt+n||t
projection
currently
available.
It confirms
high
degree
central
bank
credibility
among
financial
sector
ofitconsumer
inflation
expectations
wea=
use
subjectively
perceived
(CPPI)
ofstaff
current
CPI inflation.
t+n
interest
rates
suggests
that financial
sector
analysts
of theinstead
NBP
forecasts
from
18
analysts
Poland.that NBP actions will bring future inflation closer to the target than it results from the
the targetin believe
Estimation results confirm that the inflation target set by the National Bank of Poland has a strong impact
9
projection currently available. It confirms a high degree
of central bank credibility among financial sector
on inflation expectations of financial sector analysts
and enterprises,
while a relatively small impact on
[here: Figure
6]
18
analysts in Poland.
consumer inflation expectations (Table 5). The weight of the NBP inflation target in the formation of
inflation expectations is only 15% in the case[here:
of consumers,
Figure 6] whose expectations rely mostly on currently
5 Conclusions
and non-alcoholic beverages, tobacco, housing and energy carriers, medical products, fuels, communication services, newspapers
and articles and products for personal care [see Halka and L
yziak (2013), for details].
17 The
In theresults
case of consumer
inflation
expectations
we usethat
subjectively
perceived
inflation
(CPPI) instead
of current
CPI Bank
inflation.
presented
in this
note suggest
inflation
projections
published
by the
National
of
18 The results obtained are consistent with the results reported in the literature on the basis of similar survey data [Lyziak
Polandorhave
played
important
roleofin
making interest
ratethat
decisions
predictable.
In theofinitial
(2012)]
on the
basis the
of the
NBP Survey
Professional
Forecasters
enables more
analyzing
different aspects
centralstage
bank
credibility with long-term probability inflation forecasts [Kowalczyk et al. (2013)].
5
Conclusions
9
of inflation
targeting
(1998-2004)
– inspite
of the
of communication
tools
by theBank
Polish
The results
presented
in this note
suggest
thatdevelopments
inflation
projections
published by
theused
National
of
central
bank played
– the the
predictability
decisions
was limited
even deteriorating
[”Lyziak
Poland have
important of
rolemonetary
in makingpolicy
interest
rate decisions
more and
predictable.
In the initial
stage
10
et
(2007)].
Only the
publication– of
inflation
forecasts
since August
2004 has contributed
to improvements
of al.
inflation
targeting
(1998-2004)
inspite
of the
developments
of communication
tools used
by the Polish
NBP
Working
No.predictability
162
in
this
respect.
central
bankPaper
– the
11
of monetary policy decisions was limited and even deteriorating [”Lyziak
The
note confirms
findings
concerning
central
credibility
Poland
[i.e. L
” yziak to
(2012)],
showing
et al.
(2007)].
Only theprevious
publication
of inflation
forecasts
since
August in
2004
has contributed
improvements
that
inflation
in this
respect.expectations of financial sector analysts and enterprises display a high degree of anchoring
the target believe that NBP actions will bring future inflation closer to the target than it results from the
projection currently available. It confirms a high degree of central bank credibility among financial sector
analysts in Poland.18
Chapter 5
[here: Figure 6]
5
Conclusions
The results presented in this note suggest that inflation projections published by the National Bank of
Poland have played the important role in making interest rate decisions more predictable. In the initial stage
of inflation targeting (1998-2004) – inspite of the developments of communication tools used by the Polish
central bank – the predictability of monetary policy decisions was limited and even deteriorating [”Lyziak
et al. (2007)]. Only the publication of inflation forecasts since August 2004 has contributed to improvements
in this respect.
The note confirms previous findings concerning central credibility in Poland [i.e. L
” yziak (2012)], showing
that inflation expectations of financial sector analysts and enterprises display a high degree of anchoring
at the NBP inflation target, while consumer inflation expectations are driven mainly by price movements
currently perceived.
18 The results obtained are consistent with the results reported in the literature on the basis of similar survey data [Lyziak
(2012)] or on the basis of the NBP Survey of Professional Forecasters that enables analyzing different aspects of central bank
credibility with long-term probability inflation forecasts [Kowalczyk et al. (2013)].
10
12
Narodowy Bank Polski
References
References
R. A. Batchelor and A. B. Orr, “Inflation expectations revisited,” Economica, vol. 55, no. 219, pp. 317–331,
1988.
A. S. Blinder, “Central-bank credibility: Why do we care? How do we build it?” The American Economic
Review, vol. 90, no. 5, pp. 1421–1431, 2000.
A. N. Bomfim and G. D. Rudebusch, “Opportunistic and deliberate disinflation under imperfect credibility,”
Journal of Money, Credit and Banking, vol. 32, no. 4, pp. 707–21, 2000.
J. A. Carlson and M. Parkin, “Inflation expectations,” Economica, vol. 42, no. 166, pp. 123–138, 1975.
S. G. Cecchetti and S. Krause, “Central bank structure, policy efficiency, and macroeconomic performance:
exploring empirical relationships,” Federal Reserve Bank of St. Louis Review, vol. 84, no. 4, pp. 47–60,
2002, Federal Reserve Bank of St. Louis.
G. Chortareas, D. Stasavage, and G. Sterne, “Does it pay to be transparent? International evidence from
central bank forecasts,” Federal Reserve Bank of St. Louis Review, vol. 84, no. 4, pp. 99–117, 2002, Federal
Reserve Bank of St. Louis.
C. Crowe and E. E. Meade, “Central bank independence and transparency: Evolution and effectiveness,”
European Journal of Political Economy, vol. 24, no. 4, pp. 763–777, 2008.
A. Cukierman and A. H. Meltzer, “A theory of ambiguity, credibility, and inflation under discretion and
asymmetric information,” Econometrica: Journal of the Econometric Society, vol. 54, no. 5, pp. 1099–
1128, 1986.
O. Demchuk, T. L
¡ yziak, J. Przystupa, A. Sznajderska, and E. Wróbel, “Monetary policy transmission mechanism in Poland. What do we know in 2011?” NBP Working Papers, vol. 116, 2012, National Bank of
Poland.
N. N. Dincer and B. Eichengreen, “Central bank transparency and independence: updates and new measures,”
BOK Working Paper, vol. 21, 2013, Bank of Korea.
M. Ehrmann, S. Eijffinger, and M. Fratzscher, “The role of central bank transparency for guiding private
sector forecasts,” ECB Working Paper, vol. 1146, 2010, European Central Bank.
S. C. Eijffinger and P. M. Geraats, “How transparent are central banks?” European Journal of Political
Economy, vol. 22, no. 1, pp. 1–21, 2006.
J. Faust and L. E. Svensson, “Transparency and credibility: monetary policy with unobservable goals,”
International Economic Review, vol. 42, no. 2, pp. 369–397, 2001.
11
NBP Working Paper No. 162
13
A. Fouejieu, “Coping with the recent financial crisis: Did inflation targeting make any difference?” International Economics, vol. 133, pp. 72–92, 2013.
A. Fracasso, H. J. Genberg, and C. Wyplosz, “How do central banks write? An evaluation of inflation reports
by inflation targeting central banks,” Geneva Reports on the World Economy, vol. Special Report 2, 2003,
CEPR Publications.
P. M. Geraats, “Why adopt transparency? The publication of central bank forecasts,” ECB Working Paper,
vol. 41, 2001, European Central Bank.
A. Ha•lka and T. L
• yziak, “How to define the Consumer Perceived Price Index? The case of Poland,” NBP
Working Paper, vol. 160, 2013, National Bank of Poland.
M. M. Hutchison and C. E. Walsh, “The output-inflation tradeoff and central bank reform: Evidence from
New Zealand,” The Economic Journal, vol. 108, no. 448, pp. 703–725, 1998.
R. Kokoszczyński, T. L
• yziak, J. Przystupa, and E. Wróbe, “Analyzing monetary policy stance: The case of
Poland,” Monetary Policy And Issues: New Research, vol. 83, 2006.
H. Kowalczyk, T. L
• yziak, and E. Stanis•lawska, “A new approach to probabilistic surveys of professional
forecasters and its application in the monetary policy context,” NBP Working Paper, vol. 142, 2013,
National Bank of Poland.
T. L
• yziak, “Inflation expectations in Poland,” NBP Working Paper, vol. 115, 2012, National Bank of Poland.
T. L
• yziak, J. Mackiewicz, and E. Stanis•lawska, “Central bank transparency and credibility: the case of
Poland, 1998–2004,” European Journal of Political Economy, vol. 23, no. 1, pp. 67–87, 2007.
J. Mackiewicz-•Lyziak, Wiarygodność banku centralnego.
Difin, 2010.
L. Mahadeva and G. Sterne, Monetary policy frameworks in a global context.
Routledge, 2000.
F. S. Mishkin and K. Schmidt-Hebbel, “One decade of inflation targeting in the world: what do we know
and what do we need to know?” NBER Working Paper, vol. 8397, 2001, National Bureau of Economic
Research.
——, “Does inflation targeting make a difference?” NBER Working Paper, vol. 12876, 2007, National Bureau
of Economic Research.
L. E. Svensson, “Monetary policy and real stabilization,” NBER Working Paper, vol. 9486, 2003, National
Bureau of Economic Research.
J. Tarkka and D. Mayes, “The value of publishing official central bank forecasts,” Bank of Finland Research
Discussion Paper, vol. 22, 1999, Bank of Finland.
B. Winkler, “Which kind of transparency? On the need for effective communication in monetary policymaking.” Ifo Studien, vol. 48, no. 3, pp. 401–427, 2002.
12
14
Narodowy Bank Polski
Figures and tables
Figures and tables
Figure 1: History of inflation targeting in Poland, 1998-2012
2000 2005 2010
1. Political
a. Formal objectives
b. Quantitative targets
c. Institutional arrangements
3
1
1
1
3
1
1
1
3
1
1
1
2. Economic
a. Economic data
b. Policy models
c. Central bank forecasts
1
1
0
0
3
1
1
1
3
1
1
1
3. Procedural
a. Explicit strategy
b. Minutes
c. Voting records
2
1
0
1
2
1
0
1
3
1
1
1
4. Policy
a. Prompt announcement
b. Policy explanation
c. Policy inclination
3
1
1
1
3
1
1
1
3
1
1
1
5. Operational
a. Control errors
b. Transmission disturbances
c. Evaluation policy outcome
2
1
0.5
0.5
2
1
0.5
0.5
2
1
0.5
0.5
Total
11
13
14
Source: own calculations.
Table 1: Eijffinger and Geraats (2002) transparency index for Poland in 2000, 2005 and 2010
13
NBP Working Paper No. 162
15
Questions:
Category of answer:
2000 2005 2010
1. Explanation of policy decisions
Central bank provides explanation on day policy changes?
Yes
No
Yes
Explanations provided when policymakers meet and not change policy? Sometimes
No
Policy decisions discussed in standard bulletins and reports?
At least twice a year
At least annually
No
Within a month of meeting
More than a month after
No
Minutes of policy meetings published?
Yes
Voting patterns published?
No
2. Published forward-looking analyses
More than annually
Forward-looking analyses in standard bulletins and reports
At least annually
Unspecified
Otherwise
Words, one of numbers and graphs
One of words, numbers and graphs
Unspecified
None
Form of publication
Words, one of numbers and graphs
One of words, numbers and graphs
Unspecified
None
Risks to forecast published?
Yes
Sometimes
No
Discussion of past forecast errors
3. Assessment and analyses
More than annually
At least annually
Otherwise
Analyses in standard bulletins and reports?
At least monthly
At least quarterly
Less than quarterly / occasional
Frequency of speeches
Never, almost never
More than 10 each year
More than 5 each year
More than 2 each year / occasional
Never
Working papers and other research publications
Source: own calculations.
Table 2: Mahadeva and Sterne (2000) transparency index for Poland in 2000, 2005 and 2010
14
16
Narodowy Bank Polski
Figures and tables
Interest rate forecast errors
EG index MS index
WIBOR 1M (FRA)
WIBOR 3M (Reuters)
MAE (pp.) MAPE (%) MAE (pp.) MAPE (%)
2000:01-2004:08 11 (2000) 66 (2000)
0.24
21.9
0.18
16.2
2004:09-2007:05 13 (2005) 79 (2005)
0.09
12.6
0.05
10.0
2007:05-2012:12 14 (2010) 85 (2010)
0.09
17.5
–
–
Source: own calculations.
Table 3: Transparency indices and measures of interest rate forecast errors
Figure 2: Interest rate forecast errors based on financial market data
Figure 3: Consistency of monetary policy decisions with the NBP staff inflation forecasts (4-quarter horizon)
15
NBP Working Paper No. 162
17
Figure 4: 12-month inflation expectations and the NBP inflation target
Inflation expectations of
NBP inflation target
consumers enterprises financial sector analysts official
adjusted
mean (%)
std. dev. (p.p.)
4.8
2.0
3.2
1.1
2.9
1.0
2.9
1.0
2.9
1.0
mean abs. deviation from the adj. NBP target (p.p)
mean abs. perc. deviation from the adj. NBP target (%)
2.0
72.3
0.5
19.7
0.3
9.9
–
–
–
–
credibility index by Cecchetti and Krause (2002)
0.88
0.98
0.99
–
–
Note: In the case of inflation expectations of consumers and financial sector analysts we use monthly data, while in
the case of enterprise inflation expectations – quarterly data.
Table 4: Inflation expectations and the NBP inflation target
Figure 5: Credibility indices based on Cecchetti and Krause (2002) aproach
16
18
Narodowy Bank Polski
Figures and tables
weight of inflation target (λ) weight of current inflation (1-λ)
number of observations adj. R2
consumers
0.15
(0.05)
0.85
144
0.94
enterprises
0.79
(0.09)
0.21
48
0.62
financial sector analysts
0.93
(0.04)
0.07
142
0.98
Note: In the case of inflation expectations of consumers and financial sector analysts we use monthly data (2001:012012:12), while in the case of enterprise inflation expectations – quarterly data (2001Q1-2012Q4). Estimating the
equation for consumer inflation expectations we use currently perceived inflation (i.e. the Consumer Perceived Price
Index) instead of current inflation. Standard errors in parentheses.
Table 5: Estimation results of the credibility parameter based on Bomfim and Rudebusch (2000) approach
Figure 6: 12-month inflation expectations of financial sector analysts vs. NBP inflation target and NBP staff
inflation projections
17
NBP Working Paper No. 162
19
No. 162
NBP Working Paper
NBP Working Paper No. 162
www.nbp.pl
A note on central bank transparency
and credibility in Poland
Tomasz Łyziak