A R G U M E N T A OECONOMICA N o 2 (16) 2004 PL ISSN 12.H-5K VS
Urszula Ziarko-Siwek*
TRANSPARENCY OF MONETARY POLICY -
THEORY AND PRACTICE
T h e m ain goal o f this article is the testin g of the transparency o f m onetary policy in Poland by th e reaction of financial m ark ets to one event: the M o n e tary Policy Com m ittee’s decision c o n ce rn in g changing or not c h an g in g the reference rate.
K e y w o r d s : transparency o f m onetary policy, event studies m eth o d o lo g y
INTRODUCTION
T he p resent article concerns announcement effec t in Poland in 1 9 9 9 -2 0 0 3 . The main goal o f this paper is to find the answ er if the yield curve re acts to some inform ation published in the ec onom y in connection with the financial market efficiency. We chose to investig ate one event: M onetary Policy Committee (M P C ) decisions co n cern in g the reference rate. W e will try to interpret the yield curve reaction concerning transparency of monetary policy. The financial m a rk e t’s reactions, particularly to the MPC d ecision, prove if the m onetary policy is transparent or not.
W e w ill use the event study m ethodology. The m ain goal o f this test is to answ er how fast market participants react to new inform ation.
T he initial analysis of yield curve determ inants and announcements effects allow s us to formulate the following hypothesis: reactions of the short part o f the yield curve to the MPC decision abo u t changing the reference rate in 1999-2003 indicate that Polish m onetary policy was transparent.
1. TRANSPARENCY OF MONETARY POLICY IN THEORY
C en tral bank transparency could be defined as the absence of asym m etric information b etw een monetary p o licy m ak ers and other
*
eco n o m ic agents or as reducing asym m etric inform ation by publication of the cen tral bank’s private inform ation which is re le v an t for the policy m aking process (Hahn 2002, p. 430). The d efin itio n of transparency co n cern s both information to be in the possession of econom ic agents and its d eliv erin g . In M. G o o d frien d ’s opinion revealing accurate inform ation by the central bank reduces asym m etric inform ation and uncertainty in the m arket, because the m ark et acts in accordance with the efficient m arket hypothesis and the rational expectations theory (Goodfriend
1986).
Table I
Some definitions of transparency of monetary policy
A u th o r D efinition
R. E. K eleher (K eleher 1997, p. 2)
monetary p o licy is transparent when it is: • w ithout secrcts,
• u nam biguous,
• understood outside; by people, banks, com panies. H. D ilen, J. Nilsson
(D ilen , N ils s o n 1998, p. 206-207)
transparent m onetary policy:
• guarantees appropriate style o f co m m u n icatio n , • uses ap p ro p riate tools to comm unicate. B. W in k ler
(W inkler 2000, p. 8)
features o f m onetary policy: • openness, • clarity, • com m on understanding, • honesty. K. S zeląg (Szeląg 2003. p. 32)
• in narrow sen se - transparency o f m o n etary policy concerns revealing o f inform ation about internal d ecisio n taken by central bank,
• in w ide sen se - explaining how m onetary policy achieves objectives o f m onetary policy.
T h e above definitions and features of tran sp aren cy indicate that tran sp aren cy concerns o b jectiv es of monetary p olicy, tools of m onetary p o licy and procedures used to achieve these objectives.
E xplanation of m onetary policy objectives is a very im portant co m p o n en t of transparency, because it perm its to identify tools and p ro c ed u re s which help reach this objective. F inancial markets are fu n c tio n in g efficiently w hen information is w idely accessible, w hen
objectives are known and w hen central bank reactions (decisions) are m ore u n derstandable. If the ce n tral bank wants to c o n d u c t a transparent m onetary p o licy , it has to bring into effect clear ru les and procedures (which can elim inate uncertainty o f financial m arkets an d m inim ize the cost o f re d u cin g inflation). T h an k s to transparency, the financial market is m ore e ffic ie n t which is very useful for market p artic ip a n ts and the central b an k .
T he tim e in which some inform ation is published is also connected with a transparent m onetary policy. For m onetary policy to be tran sp aren t, the central bank sho u ld disclose im p o rtan t information punctually.
1.1. Determ inants of transparency
In o rd e r to improve tran sp aren cy of m onetary p o licy , the central bank’s inform ation policy should aspire to maintain a balance between openness o f m onetary policy and clarity (Szeląg 2 0 0 3 , p. 32). The am ount o f inform ation which is delivered by the central bank should be prudent. A way of transm itting this information sh o u ld not be too co m p licated in order not to blur the essence of info rm atio n . The central bank sh o u ld not show outside only fundamental and sim p le information, when in sid e m ore com plicated procedures and inform ation are used. The lack o f b alan c e between inform ation possessed by the cen tral bank and the ec o n o m ic agents can lead to lim ited transparency.
In o rd e r to im prove transparency o f monetary policy o n e should: • c le a rly form ulate objectives o f monetary policy an d disclose them publicly - the best way of im proving transparency is to publish strategic (long-term ) objectives of m onetary policy. Thanks to th is announcem ent m arket p articip an ts can u nderstand decisions taken by the central bank. The ce n tral bank can also d efin e short-term objectives that can help in reaching long-term objectives,
• ex p la in m onetary tran sm issio n mechanism w ith docum ents and articles pu b lish ed by the central bank on how m o n etary transm ission m echanism works - the central bank should explain th e influence of decisions co ncerning the changes o f base interest rates on m arket interest rates, in flatio n and output. T h e central bank should ju s tif y its decisions and ac tio n s taken, and of course reaso n s of missing its aim s,
• system atically publish statistical data and analy ses used by the cen tral bank in the course o f taking important d ec isio n s. The central bank attem p ts to increase transparency of m onetary p olicy by publishing a n a ly se s o f activities taken by the central bank and th e ir influence on the eco n o m y and objectives o f m onetary policy. A m o n g these publications are: quarterly and yearly inflation reports and m onthly statistical bulletin,
• o rganize press con feren ces after each m eetin g o f the MPC. The M o n etary Policy C om m ittee announces decisions taken during these c o n fere n ces. Immediate transm itting of the central b a n k ’s decision along w ith its explanation and the publication of this announcem ent via the In tern et assures suitable com m unication with m arket participants,
• publish MPC m inute-book and voting of M P C m em bers in order to red u ce the asymmetry o f inform ation, thus e n h a n cin g knowledge o f m ark et participants on m otives for central b a n k ’s decisions, their c o n n e ctio n s with objectives o f monetary policy and views of M onetary P olicy C om m ittee m em bers. Some authors think that publication o f v o tin g o f MPC m em bers is not a good instru m en t to increase the tran sp a ren c y of m onetary policy. Publication o f individual voting records will allo w the personalization of the decision m ak in g process with the p o ten tial implication that outside agents try to influence com m ittee m em b ers and adjust to such demands. The inform ation about voting b e h a v io u r might make it e a sie r for outside agents to forecast the future co u rse o f monetary policy (N eum ann 2002, p. 358),
• subm it annual m onetary policy guidelines to th e Sejm together with the subm ission by the C o u n cil of Ministers of the d ra ft Budget; report to the S ejm on the perform ance of monetary policy guidelines within five m o n th s o f the new fiscal year.
All o f the above determ inants increase the transparency of m onetary policy. But the question is whether monetary policy will be perfectly transparent, once the central bank implements all the above activities. In the real world, this problem is much complex. “ Perfect transparency” w ould imply that the central bank must disclose all the information which contributed to its decisions. But on purely practical grounds it cannot publish everything, such as the course and results o f all workshops, all the com m ittee meetings o f the central bank, every discussion, and so on. T here are som e natural limits to transparency (R em sperger et al. 1999, p. 10). In practice, transparency is invariably partial and can never be complete and perfect, so the central bank has to select inform ation to the public.
1.2. Consequences of transparency of monetary policy
A m ore transparent policy approach would make a num ber of contributions to the central bank, to the economy, to financial markets and to nonfinancial sectors. Improving transparency, for example, would (K eleher 1997, p. 5-6):
• help to clarify the long-term policy objective - a more open policy would create incentives for m onetary policymakers to carefully outline the primary objectives of monetary policy and the central bank would create incentives to keep attention focused on such goals,
• im prove the workings and usefulness of financial m arkets - financial markets work better when inflation objectives are known and more timely and detailed information is readily available. A more open policy improves the workings o f financial markets because uncertainties are m inim ized and market volatility is reduced. More information enables m arket participants expectations to adjust faster to changes in monetary policy. This situation reduces premiums connected with uncertainty. The central bank helps market participants formulate inflation expectations by publishing the inflation objective. Inflation expectations and a creditable objective reduce the negative influences o f unexpected shocks. B etter informed financial market delivers to other participants more precise information about the future,
• im prove predictability o f monetary policy decisions and efficiency of financial markets - a more transparent policy enhances anticipation of the central bank decisions. When the central bank takes its decisions, it focuses its attention on some economic information, for example: inflation, output. If financial markets understand which factors influence inflation and know the function o f the central bank reaction, market interest rates should change immediately to this information. In accordance with the efficient market hypothesis, prices of financial instruments will always include all available information. But we have to rem em ber that it is not possible to completely eliminate surprises connected with central bank decisions, because the central bank does not possess full information about m onetary transmission mechanism,
• im prove central bank credibility - as monetary policy goals and procedures become well known and understood, the public more quickly learns about changes in policy. Additionally, as the central bank begins to achieve monetary policy goals with greater regularity, its credibility improves,
• im prove efficiency of m onetary policy:
- a m ore open and transparent explaining of the central bank decisions and understanding them by the public can reduce inflationary expectations,
which is especially important to efficiency of monetary policy in realizing its objective - to maintain price stability. Monetary policy will be efficient if inflationary expectations are stable and near inflation objective. If inflationary expectations change is essential, the nominal market interest rates will have to be changed in order to keep real interest rates at a suitable level. This situation will reduce the efficiency of monetary policy (Bernhardsen et al. 2002, p. 46),
- when financial markets understand and anticipate the actions of the central bank, the first step in the transmission m echanism between policy actions and economic activity and inflation work m ore smoothly (Freedman 2002, p. 155),
• improve the central bank responsibility to the public (Eijffinger et al. 2000, p. 2; Haan et al. 1998, p. 4) - responsibility demands explaining decisions and activities of the central bank. When all activities taken by the central bank are open, it is easier to evaluate the activities of the central bank and its responsibility. A. S. B linder thinks that the central bank should explain its decisions to the public in order for monetary policy to be more responsible. If the central bank cannot explain its decisions, it m eans that its decisions are not suitable (Blinder 2001, p. 97),
• improve conducting business activity - a more transparent monetary policy helps planning business activities. Companies better informed about future central bank decisions can take business decisions easier and faster,
• minimize the chances that policymakers would manipulate policy for political purposes - a more open monetary policy prevents politicians from using monetary policy and the central bank to realize short-term (political) objectives. Market participants would quickly react to such manipulation by im m ediately revising inflationary expectations and such actions would readily be obvious to everyone. Analyzing the yield curve delivers this information to others,
• improve monetary policy - a more transparent monetary policy would encourage and lead to more open debate and criticism. Such criticism would obligate the policymaker to defend its policy objectives, decisions and procedures.
Not everybody thinks that a fully transparent monetary policy has only strong points. Some authors think that with transparency some costs are connected, for example: costs of changing things, costs of delivering inform ation. In their opinion, too much transparency and too many publications, which cannot be suitably understood, can bring numerous bad outcom es (Issing et al. 2001, p. 134).
2. TRANSPARENCY OF MONETARY POLICY IN PRACTICE
W e can research transparency o f monetary policy in several ways. One of them is to analyse the reactions o f short-term interest rates to information published by the MPC concerning changing the main interest rate.
2.1. The NBP public information policy in 1999-2003
In the M edium -Term S trategy o f Monetary P olicy (1 9 9 9 -2 0 0 3 ) we can read a note that the central b an k taking decision a b o u t im plem enting D irect In fla tio n Targeting (D IT ) in Poland has to m o d ify the role of the NBP p u b lic inform ation policy (T h e M edium-Term S trateg y o f Monetary P olicy 1 9 99-2003, p. 12-13). In the MPC opinion, c le a rly presenting m onetary policy objectives and m onetary policy in stru m en ts are very im portant in the context o f cred ib ility of m onetary p o licy . The NBP public inform ation policy should be aimed at convincing the public about the cen tral bank com m itm ent to achieve a declared ta rg e t (The Medium- Term S tra te g y of M onetary P o licy 1999-2003, p. 13). The public inform ation policy will be particularly instrum ental for lowering inflatio n ary expectations.
T he N ational Bank of P olan d has improved its inform ation policy recently. A m ong essential elem en ts of information p o licy used by the central b an k to com m unicate w ith m arket participants are:
• p re se n tin g a strategy o f m onetary policy that c o n ta in s a long-term objective o f m onetary policy,
• p u b lish in g monetary policy guidelines for each y e a r that contains a short-term objective of m onetary policy,
• p u b lish in g the Inflation R eport quarterly and y ea rly , • M P C M inutes,
• p u b lish in g information ab o u t voting of MPC m em b ers in “M onitor S ądow y i G ospodarczy”,
• p re ss conferences organized by the MPC, • o p in io n on the subject o f fu tu re inflation, • o p in io n on the subject o f p ro ject of the Budget, • sp e e c h e s and opinions o f M P C members,
• ed u c atio n al materials pu b lish ed on the N B P w ebsite and in perio d icals.
2.2. Event studies methodology
W e are going to conduct o u r research using ev e n t study m ethodology. T h e m ain goal of these tests is to answer how fast m arket participants react to new inform ation. T he main goal of ev en t studies is to check if so m e ev ents have influenced prices of some fin an cial instruments (how p rice s o f some financial instrum ents react to new inform ation). C o n d u ctin g this research we assume that investors are rational. T h erefo re, the suitable d irectio n o f these reactions should accompany the fast reaction of prices. If we w ant to use this m ethodology to investigate tran sp aren cy o f m onetary policy, we have to assu m e that the financial m ark et (m arket participants) works in accordance w ith efficient m arket hy p o th esis.
Stage /. Event definition
T h e initial task of co n d u c tin g an event study is to define the event of in tere st and identify the tim e o f beginning this ev e n t. The main goal of ev e n t studies m ethodology is to investigate how prices of financial in stru m en ts (interest rates, prices of bonds) react to some events. W e sh o u ld observe the trends o f interest rates around the time of this event ap p e arin g . This event has to be defined explicitly. T his event is defined as „th e M PC decision on ch a n g in g official interest ra te s” .
M P C members take d ecisio n s concerning the re feren ce rate during the M P C m eeting. M eetings o f the MPC are convened by the Chairperson at least once a month. The d ate o f the next m eeting should be known one m onth ahead. This is very im portant because o f the behaviour of m arket in te re st rates expectations. T h e press conference tak es place after each M P C m eeting. MPC m em bers try to explain th e ir monetary policy d e c isio n s during this conferen ce. At the sam e tim e the central bank pu b lish es an announcem ent on the NBP web site and the press agency w eb site. The positions taken by MPC m em bers d uring votes shall be p u b lish ed within six w eeks, not later than three m onths after the date the reso lu tio n is adopted.
D uring 1999-2003 the M P C took different d ecisions concerning official interest rates: increase o f interest rates (4), reduction of official interest rates (21) and no change (33). Characteristic for P olish monetary policy d u rin g this period were the so-called periodic changes. One of these perio d s (1999-2000) was characterized by increases in interest rates and
another p eriod (2001-2003) included the MPC d ec isio n s concerning a reduction o f official interest rates.
Stage 2. Identifying the event window
T he n ex t step is to identify the period over w hich the security (financial instrum ent) prices in v o lv ed in this event w ill be exam ined - the event w indow . This event w indow begins before th e event and ends after it. S o, the event w indow is expanded to a few d ay s (m inutes), the day b efo re the announcem ent and the day after the announcem ent. The length o f th e event window is giv en as follows by tw o points:
• day T 0+(.j), the first day, th e day before this e v e n t; the first day to capture th e price effects of announcem ents which o cc u r before this event - 7 - 7 ,
• day To+(+j), the last day, th e day after this ev e n t; the last day to capture th e price effects of announcem ents which o c c u r a fte r this event - 7+ 6.
S om e foreig n authors use very short event w in d o w , which begins several d a y s before the M PC d ecisio n and ends sev e ral days after this decision. O th ers use very long ev en t window, which b eg in s for example one day a fte r the previous M P C decision and ends o n e day after the current M P C decision. In ac co rd an ce with the second approach, market p artic ip a n ts form ulate their expectation concerning the future MPC decision a fte r the previous M PC m eeting. M oreover, d u rin g this period in the ec o n o m y some m acroeconom ic data are published, w hich influence the ex p e ctatio n s. In some d ev elo p ed countries, such as the US and UK, financial m arkets react to new s in several seconds (E derington et al. 1993; C la re et al. 2001a and 2001 b ). But in co u n trie s like the Czech R epublic (P odpiera 2000) and Poland, this reaction can take several hours.
iho first day of research
the previous decision
Ti=T-7
the last day of research
l he MPC decision
r2=T+6
event window F ig u re I . Event window
S ource: author’s own
Stage 3. Calculate relative changes in each clay in the event window, and average and cumulated relative average changes
W e calculated relative changes of market interest rates for each day: . “ riTn.. . )
j e < - 6 ,6 > , ( i )
w z ..
where:
W Z iT^ - relative change o f the market interest rate in the event window
w e in a period <7’0+y./ » T0+j>,
r jT - the market interest rate in the event w indow we in a day T0+j,
riTu+H ~ ^ie market interest rate in the event w indow we in a day Tn+j.t .
N ext, we calculated the relative average changes in each moments (for each T u+j) before and after event by using this form ula:
N
where:
W Z r^ - average return in a day To+j,
In the next step we calculated the cumulated relative average changes in the event window:
CWZ(Tl,T,) = f w Z r „ . J ,
(3)
7=7', where:
C W Z (7 ’1, T2) - cumulated relative average changes in the event window,
T, = T()+j for j = - 6 T2 = T0+i for j = 6.
All events were divided into 4 classes: class 0, class 1, class 2 and class 3. Separate classes were characterized in table 2.
Table 2 C lass o f events N u m b e r o f class 0 1 2 3 F e a tu re the reference rate was not changed
the reference rate was increased
the reference rate was reduced and relative change w as higher than 5% or 5%, so-called big reductions
the reference rate was reduced and relative change was low er than 5%, so- called small reductions A m o u n t
of events 33 4 14 7
P eriod 1999-2003 1999-2000 2001-10.2002 11.2002-06.2003 Source: author’s own on the basis o f N B P ’ data
We divided all events into 4 classes because the events analyzed are different. With som e of them big reductions are connected (for example: 2.50 percentage point), with others - small reductions (for example: 0.25 percentage point), and with others - increases of the reference rate. In our opinion we should not treat these different changes (increase-reduction) in the same way.
Stage 4. Testing procedure
Important considerations are defining the null hypothesis. W e focused on the null hypothesis that the given event has no impact on the behaviour of market interest rates. Hence, the null hypothesis says that: relative average changes of interest rates are zero.
H„: W Z r„.: = 0
H,: W ZT[UJ * 0 .
2.3. Short m arket interest rates used in research
We used in the research some market interest rates such as: one-month W IBOR (1M), three-month W IBOR (3M), six-month W IBO R (6M) and two- week W IBOR (2W). We had to choose the two-week W IBO R (the first day of quotations was 23 January 2003) because the NBP changed the reference rate of the open market operation from 28 days to 14 days. This change meant that the central bank started to directly influence the two-week money market interest rate. Data used in the research concerning 1M, 3M and 6M WIBOR cover the period from 4 January 1999 to 7 November 2003, and the data concerning 2W W IBO R cover the period form 23 January 2003 to 7 November 2003. It should be emphasized that WIBOR is quoted at 11.00 o’clock. If some events took place for example at 12.00 o ’clock, WIBOR could not react to these events the same day, but the next day. W e also used interest rates o f FRA contracts: FRA 1X2, FRA 3X6, FRA 6X9, FRA 9 X 12. The data used are average between bid and ask prices. These data cover the period from 29 August 2000 (FRA 1X2 from 2 January 2001) to 7 November 2003.
All the data used is from the over-the-counter m oney market. Market participants of this market have easy access to professional news, they are very close to very important information, once it is published. Their access is also suitably fast. All data concerning interest rates were raised from REUTERS through the workers of one of the bank.
2.4. Results
The first money market interest rate used in the research was WIBOR 1M (2W). Relative average changes of quoted market interest rates for each instrument (Figure 2) and cumulated relative average changes for every day in the event period (Figure 3) indicate days when substantial reactions occurred of chosen interest rate in each class.
In the case of class 0, which includes events “no change of the reference rate”, relative average changes are near zero. This reflects the lack of reaction of interest rates to MPC members decisions. These results are in accordance with our expectations - the reference rate unchanged caused that changes of WIBOR are near zero. Reactions of the short part of the yield curve in this class are in accordance with the principles o f the efficient market hypothesis. The statistical value of p each day in the event window allows us to reject null hypothesis, that relative average changes equal zero. The impossibility of rejecting null hypothesis indicates that when the MPC publishes their information about the reference rate, the market does not react.
A average WZ class 0 5% 3% 1% -1% -3% -5% T-6 T-5 T-4 T-3 T-2 T-l TO T+l T+2 T+3 T+4 T+S T+6 A average WZ class 1 A A 5% 3% A A ■* A A T-6 T-5 T-4 T-3 T-2 T-l T A A * -1% A -3% -5% 0 T+l T+2 T+3 T+4 T+5 T+6 5 % A average WZ class 2 3% 1% A A 4 ' 1 1 ' * 1 * A ; > A A _|o/o A -3 % -5% T-6 T-5 T-4 T-3 T-2 T-l TO T+l T+2 T+3 T+4 T+5 T+6 A average WZ class 3 » . A A 5% 3% 1% A A * ^ 1“ "T A T-6 T-5 T-4 T-3 T-2 T -l T A I A * A A -1% A -3% -5% 0 T+l T+2 T+3 T+4 T+5 T+6 Figure 2. R e la tiv e average changes o f W IB O R IM in the event w in d o w Source: a u th o r’s own
Figure 3. Cumulative relative average changes o f WIBOR I M in the event window S o u rce: author’s own
In the case of class 1, w hich includes events „in creasin g the reference ra te ” , from day T-4 we can see that financial m ark et participants expect that d u rin g the next M PC m eeting MPC m em bers will increase the re fere n ce rate. Relative av erag e changes are p o sitiv e to day T + l. W e sh o u ld not treat these re actio n s like delayed, becau se W IBOR used in our researc h is quoted at 11.00 o’clock (fixing). In the day of the an n ouncem ent of this decisio n by the MPC, the financial market could not re act to new inform ation.
In the case of this class is very hard to conclude. Firstly, the sam ple in this class is very sm all, it includes only 4 ev e n ts, because the M PC in creased reference rates only 4 times during 1 9 9 9 -2 0 0 3 . Such a sm all sam p le can deliver incidental results. N one o f these reactions are sig n ifican t in the statistical sense. Secondly, 4 e v e n ts took place at the turn o f 1999 and 2000, w hen financial m arket participants started to u n d ersta n d how the M PC w orked and took d ecisio n s. Especially during 1999 the expected Y ear 200 0 Problem has influ en ced on the financial m ark et and WIBOR.
C lasses 2 and 3 are con n ected with reductions o f reference rates. T his w as th e period 2001-2003. W e can see from the figures that from day T-2 fin an cial market participants expect reduction o f reference rates. We can do so by observing relative changes of W IBOR. T h e adoption of m arket in tere st rates extends to day T+2 (class 2) and day T+l (class 3). E ssen tial reactions of short-term rates took place on days T-2 and T + l, that is before and one day a fte r the MPC m eeting.
In the ca se o f class 2, re actio n s indicate that this p art o f the financial m arket is not fully efficient, b ecau se the reactions are delayed. In the case o f cla ss 3, reactions are fast, so this segment is e ffic ie n t. It is worth saying th at class 3 covers the perio d from N ovem ber 200 2 to June 2003 and class 2 covers the previous p eriod. So, we can say th a t over the years the fin an cial m arket reacts b e tte r and faster to the n ew s. T he statistical analysis indicates that in the ca se o f class 3, the fin an cial m arket reacts on days T -2 and T + l. This m eans that the financial m ark et predicts a reduction o f reference rates, but not completely. T he financial market predicts reduction partly, and also partly reacts after announcem ent.
T h an k s to the results we can say that financial m ark e t participants do not p re d ic t com pletely MPC d ec isio n , but these p re d ic tio n s are better and better. So, m onetary policy in P olan d is not fully tran sp a ren t.
A n a ly z in g cum ulated relative average changes in fig u re 3, we can see clearly how the financial m arket reacts to MPC d e c isio n s in each class. G enerally, we can say that the financial market starts reactio n from day T-3 and en d s reaction on days T + l , T+2.
A n o th er m easure which we can use in our research, is m argin analysis for in terb an k interest rates (d ifferen ce between the sam e W IBOR and W IB ID ). T he increasing m argin we can we interpret in the context of financial m arket participants u ncertainty. When the m arg in is bigger, the u ncertainty concerning future ev en ts (future MPC d ecisio n s) is also bigger.
T he big g est margin is in the first class, so b efo re M PC meetings connected w ith increasing o f re fere n ce rates. We see th e ir strong increase before the M PC meeting (from day T-4) and reduction after the MPC m eeting. D uring day of TO o f the MPC announcem ent, uncertainty is stable.
In the case of classes 2 and 3 we can also see the increasing of the m argin, b u t this increase is sm aller. O ver the course o f tim e the margin is sm aller and sm aller. This is connected with the financial market developm ent.
F ig u re 4. Average margin betw een 1M W IBOR and 1M W IB ID in each day in the event w in d o w (in basis points)
The next way to measure these reactions, also u sin g event study m ethodology, is to calculate the difference (spread) betw een market interest rate and the reference rate in the event w indow . The case “unchanged” means that the M PC did not change the reference rate. The case „ red u ctio n ” means that the M PC reduced the referen ce rate. And the third case „increase” means that the M PC increased the reference rate. The spread in th e case of no changes can be treated as an in d icato r of normal behaviour o f interest rates. On the basis of its behaviour w e can compare it with the behaviour of spreads in case of „reduction” and „increase”. Thanks to this analysis we can indicate the behaviour o f financial market participants expectations concerning future MPC decision (before TO) and concerning reactions to this decision (after TO).
F ig u re 5. A verage differences (sp re ad ) betw een 1M W1BOR and th e reference rate for increases, re d u ctio n s and no changes o f rcfcren ce rates (in basis points)
D uring the period when the MPC did not change reference rates, we can see that the spread is at the stable level - about 60 basis points. The essential difference in behaviour of spread can be seen near the day TO (the day o f announcem ent of the M PC decision). Spread size in the event window before “reduction” is sm aller than before “unchanged” . Over the course of som e days, the difference betw een spread “reduction” and spread “without ch an g in g ” is bigger, and on day TO this difference changes. This is connected with the expectations of the reduction in the reference rate. F inancial market participants, before a reduction of the reference rate, want to lend others their money. T h e bigger supply of m oney causes a reduction o f interest rates. The nearer TO and T + l, the closer the spread back to its norm al level.
In the second case, “increase” spread widens from day T-3 to day T -l. T his is caused by expectations connected with increasing of the reference rate on TO. On day TO the spread equals 200 basis points After TO the spread com es back to the normal level.
R eactions of the short-term part of the yield curve to the MPC decision are the following:
• changes of interbank interest rates took place before and after events; reactions lasted to day T+2, but the market of FRA reacts before and during TO,
• direction of interest rates reactions were in accordance with our expectations and direction o f M PC decisions,
• the financial market better predicted events o f leaving interest rates w ithout change than of changing them,
• the financial market was not surprised that the M PC changed reference interest rates, but the level o f this changes was surprising. The confirmation o f that were reactions of interest rates to the new s several days after the decisions, especially in the case of 1M (2W) W IB O R , which is under the influence of the central bank decision. In the case of classes I and 2 reactions lasted two days. In 2003 (class 3) the financial market reacted to the new s only during the first day after the decisions. This means that the m arket reacted faster and faster,
• in the majority of cases, except IM WIBOR in class 3, there were no significant reactions in the statistical sense; this m eans that the current M PC decision was expected by the m arket before,
• the longer term to m aturity the weaker were reactions, because the influence of the MPC was sm aller.
T ab le 3 R esu lts o f event study
M ark et in te re s t rates Classes T-2 T -l TO T + l T+2 1M WIBOR 0 1 ~2 "" --- --- ---- ...— .. ... ...-3 - -3M WIBOR 0 1 2 - - -3 -6M WIBOR 0 1 2 - - - -3 FRA 1X2 0 2 3 FRA 3X6 0 2 -3 FRA 6X9 0 - + 2 -3 FRA 9X12 0 2 -3
+ o r - d e n o te s statistical significance (a = 0 ,0 5 ) and indicates the sig n o f param eter S o u r c e : a u t h o r ’s o w n
F ig u re 6. C h o sen reactions o f sh o rt-term interest rates to analyzed e v en t S ource: a u th o r’s own
O ur results indicate that monetary policy in Poland and decisions taken by the M PC were not fully predictable. When we compare our results with those obtained by J. Zieliński (Zieliński 2001a and 2001b), also in connection with the transparency of monetary policy in Poland, we can say that they are coincident. The detailed analysis of the results indicates that over the course of time the predictability of monetary decisions has improved. During first years of the MPC, after the medium-term strategy of monetary policy began to function, financial market participants learned how the central bank worked. The Polish central bank, conducting its monetary policy, tried to improve the transparency of monetary policy during this time. The confirmation of this is implementation of new instruments of public information policy. So, we can ask the question why the results indicate that the monetary policy was not fully transparent. One of the reasons can be the fact that the financial market in Poland is not efficient, not developed, and not big enough. Moreover, if the central bank policy were perfectly transparent, we could expect that the announcement of the MPC decision did not deliver any new information. But we know of course that process of taking that decision is very complicated and its result depends on the votes o f the MPC members. Each of the MPC members can understand econom ic information and their influence on economy in a different way.
W e also tried to com pare our results with the determinants of transparency of monetary policy that were presented in the first part of the paper. Thanks to this com parison we can say that if the National Bank of P oland begins to publish its inflation forecast, the transparency of monetary policy may be improved.
C O N C LU SIO N
T he main goal of this paper was to present the essence of transparency of m onetary policy, the determinants influencing improvement and the consequences of open and transparent monetary policy. Since 1999 the National Bank o f Poland has conducted transparent monetary policy using the Direct Inflationary Target strategy. The National Bank of Poland has been improving its information strategy and transparency all the time.
In this paper we presented the results of the research conducted using m ethodology of event study. W e tested if the short part of the yield curve reacts to the MPC decisions concerning changes of the reference rate. We cannot say that monetary policy is transparent.
REFERENCES
B ernhardscn T ., Klosler A., Transparency and predictability o f m onetary policy, Norges B ank, E co n o m ic Bulletin Q2, 2002.
B linder A .S ., Bankowość centralna w teorii i praktyce [Central hanking in theory and practice], C cD cW u, W arszawa 2001.
Clare A., C o u rten ay R., What can we learn about monetary policy transparency from fin a n cia l market data?, E conom ic R esearch Centre o f the D eu tsch e Bundesbank,
D iscu ssio n Papers no. 06/01, February 2001a.
Clare A. , C o u rten ay R., Assessing the impact o f macroeconomic new s announcements on securities prices under different m onetary policy regimes. B ank o f E ngland, W orking P ap ers no. 125, 2001b.
Dilen H., N ilsso n J., Transparency, uncertainty and monetary policy. R eserve Bank o f New Z ealan d , M onetary policy under u n certain ty workshop, June 1998.
E derington L. H., Lee J. H., How m arkets process information: news releases and volatility,
.J o u rn a l o f Finance*', vol 48, no. 4, 1993.
E ijffinger S. C. W., Hocberichts M. M ., Central bank accountability and transparency: theory a n d some evidence. E co n o m ic Research Centre o f the D eu tsch e Bundesbank, D isc u ssio n Papers no. 6/00, N o v em b er 2000.
Freedm an C h ., Transparency in the practice o f monetary policy. F ederal R eserve Bank o f St. L ouis, R eview , July-August 2002.
G oodfricnd M ., Monetary mystique: secrecy and central banking, “Journal o f M onetary E co n o m ic s” , no. 17, 1986.
Haan J., A m tcn b rin k F., Eijffinger S. C . W ., Accountability o f central bank: aspects and quantification, Tilburg U niversity, C e n te r o f Economic R esearch, D iscussion Papers no. 54, M ay 1998.
Hahn V ., Transparency in monetary policy: a survey, IFO Studicn, Jg. 48, no. 3, 2002. Issing O ., G a sp a r V., Angcloni I., T ristani O ., Monetary policy in the euro area. Strategy and
decision making at the European Central Bank, Cambridge U n iv ersity Press, Cambridge 2001.
K eleher R. E ., Transparency and Federal Reserve M onetaiy Policy, Joint Economic C o m m itte e Study, November 1997.
N eum ann M . J. M., Transparency in m onetaiy policy, AEJ, vol. 30, no. 4, D ecem ber 2002. Podpiera R ., Efficiency o f financial m arkets in transition: the case o f macroeconomic
releases, C E R G E-EI, W orking P ap ers no. 156, July 2000.
R em sp erg er H., W orms A., Transparency in monetaiy policy, C F S W orking Papers no. 1999/16, January 1999.
Szeląg K., Strategia jednolitej polityki pieniężnej w strefie euro - kluczowe elementy i zasady [Strategy o f a single monetary p o licy in the euro area - key elem ents and principles],
Średniookresowa strategia polityki pieniężnej na lata 1999-2003 [Medium-term strategy o f m onetary policy fo r 1 999-2 003], N arodow y Bank P olski, R ada Polityki Pieniężnej, W arsaw , September 1998.
W in k le r B., Which kind o f transparency? On the need fo r clarity in monetary policy making,
E C B W orking Papers no. 26, A ugust 2000.
Z ie liń sk i J., Przejrzystość i wiarygodność polityki pieniężnej w Polsce: reakcja krzywej dochodowości na zmiany stóp procentowych NBP [Transparency and credibility o f m onetary policy in Poland: yield curve’s response to the changes o f central h a n k ’s interest rates]. Materiały z sem in ariu m w NBP, 28th Ju n e 2001 a.
Z ie liń sk i J., Transparency o f m onetary policy in Poland: yield curve responses to changes in official interest rates, N ational B ank o f Poland, Paper No. 20, W arsaw , November 2001b.
