voorlopige versie

8/6/2019 voorlopige versie

1/27

Table of contents

1. Introduction ................................ ................................ ................................ ................................ .............. 3

2. Inflation targeting in emerging market economies ................................ ................................ ..................... 3

3. Structure ................................ ................................ ................................ ................................ ................... 4

4. Justification of the structure ................................ ................................ ................................ ...................... 6

4.1. Introduction ................................ ................................ ................................ ................................ ..... 6

4.2. Goals of macro-economic policy ................................ ................................ ................................ ....... 6

4.3. Monetary policy: alternatives ................................ ................................ ................................ ........... 6

4.4. What is inflation targeting? ................................ ................................ ................................ .............. 7

4.4.1. Constrained discretion ................................ ................................ ................................ ................. 8

4.4.2. Flexible versus strict inflation targeting ................................ ................................ ........................ 8

4.4.3. Definition of inflation targeting ................................ ................................ ................................ .... 8

4.4.4. Characteristics of the framework ................................ ................................ ................................ . 8

4.4.5. Preconditions................................ ................................ ................................ ............................... 9

4.4.6. Features of inflation targeting countries................................ ................................ ..................... 10

4.4.7. Factors influencing the adoption of inflation targeting ................................ ............................... 11

4.4.8. Advantages ................................ ................................ ................................ ................................ 12

4.4.9. Criticism ................................ ................................ ................................ ................................ .... 12

4.5. Operational framework ................................ ................................ ................................ .................. 15

4.5.1. Results ................................ ................................ ................................ ................................ ....... 15

4.5.1.1. Inflation ................................ ................................ ................................ ............................ 15

4.5.1.2. Real economy ................................ ................................ ................................ .................... 15

4.5.1.3. Cost of disinflation................................ ................................ ................................ ............. 16

4.5.1.4. Interest rates and exchange rates ................................ ................................ ...................... 16

4.5.2. Defining the target ................................ ................................ ................................ ..................... 16

4.5.3. Decision-making process ................................ ................................ ................................ ............ 17

4.6. Emerging market economies ................................ ................................ ................................ .......... 17

4.6.1. Characteristics ................................ ................................ ................................ ........................... 18

4.6.2. Implementation of inflation targeting ................................ ................................ ........................ 19

4.7. Inflation targeting during the financial crisis of 2008................................ ................................ ....... 20

4.7.1. Performance ................................ ................................ ................................ .............................. 20

4.7.2. Empirical research ................................ ................................ ................................ ..................... 20

4.8. Case study: Chile ................................ ................................ ................................ ............................ 21

4.9. Case study: Turkey ................................ ................................ ................................ ......................... 22

4.10. Comparison ................................ ................................ ................................ ................................ ... 24

4.11. Conclusion ................................ ................................ ................................ ................................ ..... 24


2/27

2 | P a g e

5. Conclusion ................................ ................................ ................................ ................................ .............. 24

6. References ................................ ................................ ................................ ................................ .............. 25

6.1 Scientific articles ................................ ................................ ................................ ............................ 25

6.2 Non-scientific articles ................................ ................................ ................................ ..................... 27


3/27

3 | P a g e

1. IntroductionIn this paper a comprehensive overview of the content and the main structure of the master thesis on inflation

targeting in emerging market economies, to be written in the next academic year, will be given to the reader.

To be able to set up this structure, it was necessary to consult the extensive literature on inflation targeting.

After reading a selected part of the literature, a first insight in inflation targeting was obtained and different

topics within the field of inflation targeting in emerging market economies could be identified.

Based on these topics, it was possible to build a structure. The result is a well-balanced structure, which starts

by giving an overview of existing monetary policy frameworks, continues with an in-depth study of inflation

targeting in general and, later on, particularly of inflation targeting in emerging market economies. In a further

stage, a case study will be made on the Chilean and Turkish experience with inflation targeting. To conclude, a

comparison will be made between the two experiences in an attempt to identify the success factors and pitfalls

of inflation targeting in emerging market economies.

The structure of this paper is as follows. In the second paragraph, the ideas covering inflation targeting will be

given and the research questions will be identified. This part will also include a description of the methodology

that will be used to write the master thesis. In the third paragraph the structure of the master thesis will be

given. The extensive fourth paragraph justifies the structure presented in the third paragraph. In thisparagraph, a discussion of each major topic is included. In this way, the paper can already be seen as a concise

overview of inflation targeting in emerging market economies, of which the master thesis will be an extension.

The last paragraph concludes.

2. Inflation targeting in emerging market economiesAt the beginning of 2011, 27 countries had adopted inflation targeting and only three countries have

abandoned the regime (Hammond, 2011, p. 1). Crucially these exits are not due to a bad performance of the

framework. Finland, Spain and Slovakia are no longer inflation targeters because they entered the euro zone

(o.c., 2011, p. 7).

In the search for a nominal anchor for inflation expectations, three possible frameworks can be distinguished.

Central banks can decide to anchor inflation expectations by targeting the exchange rate, inflation or money

supply growth. Influencing inflation expectations of private agents is very important, as private agents are not

interested in the inflation level of next week. Instead they are more interested in knowing the inflation level on

the medium term (Svensson, 2010b, p. 3). Moreover, current inflation influences the inflation expectations, as

private agents adapt their expectations based on current inflation(Meeusen, 2006, p. 286). If the inflation level

is very high, private agents will expect inflation to rise even more. Therefore, it is of great importance to

influence the inflation expectations (Meeusen, 2006, p. 272).

The money supply growth was targeted under the regime of monetary targeting. In this framework, the

inflation was controlled by targeting the money supply growth. However, due to financial innovation and

deregulation, the framework became less attractive, as the relationship between money supply growth andinflation had become unstable in the 1980s. Afterwards, most countries applied exchange rate targeting. But in

the beginning of the 1990s fixed exchange rates seemed unbearable. Therefore countries preferred to abandon

a pegged exchange rate, opted for a floating exchange rate, and adopt inflation targeting (Hammond, 2011, pp.

5-6).

Most countries adopt nowadays inflation targeting. Countries adopt the framework for different reasons

(Ptursson, 2005). Some countries are forced to switch from an exchange rate peg towards inflation targeting

due to speculative attacks on their currency, which makes it difficult for the central bank to maintain the


4/27

4 | P a g e

exchange rate peg. Others adopt inflation targeting because they encounter increased incompatibility between

the ultimate goal of price stability and the official anchor. Other central banks implement inflation targeting

gradually.

However, in the literature it is unclear whether inflation targeting is the best option. As set out in Taguchi et al.

(2010, p. 2), it is generally accepted that advanced countries perform better under this framework. However

for emerging market economies the outcomes are unclear, as they face higher inflation levels, which are moredifficult to control. Moreover, output is more volatile (Fraga et al., 2003, p. 4). Issing (2004, p. 171) states that

the performance of emerging market economies is often worse than in advanced economies due to bad fiscal

policies.

The master thesis will try to provide an extensive overview of inflation targeting for emerging market

economies and an attempt will be made to formulate a clear answer on the questions raised hereafter: Is the

inflation targeting framework the most preferable monetary policy for emerging market economies? Does

inflation targeting have the same outcome in emerging market economies as in advanced economies? The

adoption of inflation targeting and the resulting effects in two emerging market economies will be compared,

namely in Turkey, where inflation targeting was not always a success and Chile, where the framework

performed quite well. Which effects were responsible for the success in Chile and the failure in Turkey? Based

on these outcomes, a general conclusion will be drawn on the success factors and pitfalls of the framework.

The methodology to be followed is mainly an extensive literature study, combined with two case studies on

inflation targeting in Chile and Turkey. Moreover, a small empirical research will be performed on the

performance of emerging market economies that were inflation targeters during the financial crisis of 2008.

The outcomes will be compared with the performance of inflation targeting advanced economies during the

same period. In this way, a conclusion can be made on whether the emerging market economies performed

better than the advanced economies during the financial crisis of 2008. More details on the proposed

methodology can be found in paragraph 4.7.2.

3. Structure1. Introduction2. Goals of macroeconomic policy3. Monetary policy: alternatives

3.1 Exchange rate targeting3.2 Monetary targeting3.3 Other

4. What is inflation targeting?4.1 Constrained discretion4.2

Flexible versus strict inflation targeting4.3 Definition of inflation targeting

4.4 Characteristics of the framework4.4.1 Announced numerical inflation target4.4.2 Price stability4.4.3 Intermediate target using all relevant information4.4.4 Forward looking4.4.5 Transparency4.4.6 Accountability


5/27


6/27

6 | P a g e

11. Conclusion12. References

4. Justification of the structure4.1.IntroductionIn the introduction the main structure of the master thesis is built up. Moreover, the research questions are

listed. To introduce the concept of inflation targeting the importance of low and stable inflation will be

highlighted, as this promotes economic growth and efficiency in the long run. In addition, an attempt will be

made to give a comprehensive definition of inflation targeting.

4.2. oals of macro-economic policyThis chapter will be very short and will discuss the goals of macroeconomic policy. Krugman & Obstfeld (2009,

pp. 503-505) state that countries pursue both internal and external equilibrium. To realize the internal

equilibrium, the government wants employment as close to its full employment level as possible. For example,

if the labor market is being characterized by an excess, this will result in an overheating of the economy. On theother hand, the central bank tries to ensure price stability. Therefore it has to control the growth of the money

supply. In this way, prices will remain at a moderate level.

Concerning the external equilibrium, the government has to guarantee a sustainable current account, not

necessarily requiring a current account in equilibrium. An example will explain this more clearly. A shortage can

be preferable if the country has attractive investment opportunities, as such investments generate revenues.

These revenues can be used to repay the loans afterwards. Inflation targeting countries set price stability as the

primary goal of macro-economic policy. Thus, the inflation and the monetary targeting frameworks pursue

primarily an internal equilibrium. Exchange rate targeting focus more on obtaining a sustainable current

account.

4.3.Monetary policy: alternativesIn this chapter, alternatives to inflation targeting will be discussed. Two main alternatives exist, namely

monetary targeting and exchange rate targeting. Both have their own rules, advantages and disadvantages. In

the master thesis a more detailed discussion will be provided on this topic. A strong focus will be put on the

shortcomings of these frameworks, which cause inflation targeting to be the most widely adopted monetary

policy.

Kadiolu et al. (2000, pp. 5-9) give an elaborate overview of the existing alternatives. First of all, exchange rate

targeting will be explained. In this approach, the central bank pegs it exchange rate to the exchange rate of a

large country which has low and stable inflation and conducts an independent monetary policy. This framework

results in a lower inflation level in the exchange rate targeting country and will make it easier to sustain this

lower inflation level. This regime involves a very clear monetary policy. If the central bank fears the currencywill depreciate (appreciate), the monetary policy will be strengthened (loosened). Further on, fixing the

exchange rate effectively anchors the inflation expectations, which are an essential determinant of the interest

rates. Moreover, this regime avoids the time-inconsistency trap. The time-inconsistency trap means that

policymakers pursue an overly expansionary monetary policy in the short-run, which will in the long-run lead to

an increase in input prices.

However, some disadvantages can be recognized as well. By pegging the exchange rate, the central bank

cannot adopt its own policy. It always has to undergo the monetary policies set out by the foreign central bank.


7/27

7 | P a g e

Thus, if an external shock disturbs the foreign economy, the central bank of this foreign country will take

appropriate action to mitigate the effects of the external shock. These actions will also influence the economy

in the exchange rate targeting country, even if this country is not affected by this external shock. Furthermore,

a speculative attack on the currency leads to increased financial fragility. Such an attack results in a devaluation

of the currency. This will increase inflation and inflation expectations, which will lead to higher interest rates

and lower cash flows for the firms. This will in the end cause a deterioration of the firms balance sheets. A

speculative attack makes it often difficult to maintain the exchange rate peg. In the past, this was often amotive for countries to adopt inflation targeting, as already pointed out in paragraph 4.1. Furthermore,

investors will encounter a lower risk as the exchange rate is fixed to another exchange rate. This will lead to

more lending. Exchange rate targeting performs poorly in stabilizing the inflation (Svensson, 2007, p. 3). Finally,

the domestic exchange rate will fluctuate less. This means that it is more difficult for the central bank to

recognize an overly expansionary monetary policy. If the central bank has serious concerns that it pursues an

overly expansionary policy, it can correct its policy. However, as the exchange rate is far less fluctuating, this

means that the central bank is not able to detect an overly expansionary monetary policy. Exchange rate

targeting has been abandoned as the international capital flows increased and a misaligned exchange rate

made it difficult to defend the currency against speculative attacks on the currency. Finally, this framework was

less able to stabilize inflation (Svensson, 2010b, p. 6).

A second alternative to inflation targeting is monetary targeting. In this framework , the central bank targets themoney supply growth to anchor inflation expectations. Under this framework, it is far easier to deal with

external shocks in the domestic economy, whereas this is virtually impossible under the framework of

exchange rate targeting. Again, the time-inconsistency trap can be easily avoided in this framework. Monetary

targeting has been quite successful. However, two preconditions have to be distinguished, whose existence

determine whether the framework will be successful or result in a failure. First of all, it is of great importance

that the targeted monetary aggregate can be controlled by the central bank. If the monetary aggregate is not

within the central banks control, the central bank is una ble to take appropriate actions to mitigate the effect of

an external shock. The second precondition is that the target variable must be a determinant of the goal, which

can cause problems. As already stated in paragraph 4.1, the relationship between money supply growth (the

targeting variable) and inflation was not stable, which incited countries to adopt inflation targeting (Ptursson,

2005). The Bundesbank used the monetary targeting framework for a long time, but missed the target several

times to ensure meeting its inflation target (Svensson, 2007, p. 3). Svensson argues therefore that the

Bundesbank is more an indirect or implicit inflation targeter (Svensson, 2010b, p. 5). The problem encountered

by the Bundesbank in this case will be further discussed in paragraph 4.4.5.

Two other monetary policies exist that are variations on exchange rate targeting: crawling peg and target zone.

In those frameworks, the exchange rate is not fully floating and often other variables are targeted. As will be

discussed in paragraph 4.4.5, it is essential that the public can clearly distinguish which objective will be

pursued if a conflict exists between the objectives (Masson et al., 1998, p. 35).

Generally, one can argue that inflation targeting and the flexible exchange rate framework perform better than

fixed exchange rate targeting if the economy faces a real shock. However if the country is hit by a nominal

shock, the fixed exchange rate targeting regime performs better (Loayza et al., 2002, p. 8).

4.4.What is inflation targeting?This chapter will be the most extensive one and gives a broad overview of the concept. As already discussed in

paragraph 2, the purpose of the thesis is to determine which factors are responsible for the success or failure

of the policy. Therefore, an attempt will be made to point out the differences and the similarities in the

implementation, results of the framework in both advanced and emerging market economies.


8/27

8 | P a g e

4.4.1. Constrained discretionAs Bernanke et al. (1999, p. 298) already stated, inflation targeting cannot be classified as either rule or

discretion. Therefore, in the literature inflation targeting is being defined as a framework with constrained

discretion, in which the inflation target can be seen as the constraint. Inflation targeting is discretionary as the

central bank has the flexibility to determine how they react to short-term disturbances in output and financial

stability in its monetary policy (Bernanke et al., 1999, p. 298; Lucotte, 2010, pp. 2 -3; Hammond, 2011, p. 5).

4.4.2. Flexible versus strict inflation targetingFlexible inflation targeting means that the central bank tries to stabilize both inflation and the real economy

(more precisely output), whereas strict inflation targeting only stabilizes inflation. Flexible inflation targeting

performs better than strict inflation targeting if it focuses on forecasts of inflation and the real economy. These

forecasts take the time lags between the actions of the central bank and the final effect on inflation and the

real economy into account. Therefore flexible inflation targeting performs better and is often called forecast

targeting. Based on these forecasts, the central bank will choose an appropriate policy-rate path which will lead

to stable inflation and a real economy performing around a normal level (Svensson, 2010a, p. 1; Svensson,

2010b, p. 13).

In practice, strict inflation targeting is currently not adopted by any country (Svensson, 2010b, pp. 1-2).

4.4.3. Definition of inflation targetingThis paragraph will try to give a definition and the objective of inflation targeting.

Brdsen et al. (2001, pp. 452-453) define inflation targeting as the setting of the policy instrument, which is the

short-term interest rate for inflation targeting, to control the inflation forecast targeting. Therefore, the central

bank needs to understand how this short-term interest rate influences the inflation forecasts. Econometric

research in Norway performed by Brdsen et al. (o.c., p. 453) showed that the short-term interest rate

effectively influences the inflation rate. An increase in the short-term interest rate of 1 % leads to a decrease in

the inflation of 0.4%.

The objective of inflation targeting is defined by Svensson (1999a, p. 625; 2010b, p. 2) as a quadratic loss

function . This function is equal to the sum of the squared difference between inflation and the inflation

target (also called the inflation gap ) and a weight times the squared difference between actual

and potential output (also called the output gap ) or algebraically:

The goal is to minimize the deviation from the inflation target or .

Some functions take into account the policy rate changes as well. At the moment of initial adoption of the

framework, the central bank still needs to build credibility. Thus, the most weight will be put on stabilizing

inflation. Later on, if credibility has been won by the central bank by realizing disinflation, the central bank will

focus more on output stabilization.

4.4.4. Characteristics of the frameworkIn a first paragraph, the main features of inflation targeting are explained. In the literature a general consensus

exists on these characteristics. Bernanke et al. (1999), Mishkin (2000), Svensson (2007), Svensson (2010b),

Horvath (2011) all list the same characteristics. The central bank announces a numerical inflation target some

years in advance to pursue price stability, which is the central banks primary goal. Mishkin (2001, p. 1) cites a

definition of price stability. Price stability implies a rate of inflation that does not have to be taken into account


9/27

9 | P a g e

in making economic decisions. Therefore, the inflation rate needs to be relatively low. Mishkin (ibidem) states

that an inflation between 0 and 3 % satisfies this definition. However, more specific features need to be

defined, as even non-inflation-targeting central banks announce an inflation target. The European Central Bank

announces an annual inflation target of below, but close to 2 %. However, the ECB is not an inflation targeter

(Svensson, 2000b, p. 7).

One of the difficulties is that inflation is influenced by a lot of variables, which makes it more difficult todetermine an intermediate target, which is the determinant for monetary policy. Therefore, a second specific

characteristic is that models need to be designed that use all relevant information to make forecasts of the

target variables. Alternatives to inflation targeting, such as exchange rate targeting and money supply

targeting, do not face this problem, as they respectively peg an exchange rate or target the money supply.

Inflation targeting focuses therefore on inflation forecasts, which is more effective as a monetary policy

decision influences the target only after some time lag. This means that the monetary policy is forward-looking,

which is better than a reactive policy. This is consistent with the definition of Brdsen et al. (2001, pp. 452-453)

in paragraph 4.4.3. It is clear that the inflation forecasts are very important in order to control inflation.

Another characteristic of inflation targeting is that the transparency of the monetary system is higher because

all central banks publish reports on inflation forecasts. Furthermore, most countries give motivations for their

policy decisions and information on future policy decisions. In case of a deviation from the target, the centralbank needs to account for this deviation. In countries, such as the UK, Turkey, Thailand, Serbia the central

bank needs to formulate the actions it will take to reach the target in an open letter. However, this is the case

in few countries. Most countries does not have to account from deviations in an open letter (Svensson, 2010b,

p. 2; Hammond, 2011, p. 15).

All these reports make the central bank more credible and if the inflation expectations are close to the inflation

target, the credibility of the central bank will rise further. If the central bank has a desirable degree of

credibility, it succeeded already somewhat in anchoring the inflation expectations, meaning that in fact the

inflation has already been stabilized to some extent (Svensson, 2010, p. 3).

The last generally accepted characteristic is accountability of the central bank. As the central bank announces

an inflation target, it is almost obliged to match the target. If the target is missed, the central bank isaccountable towards the public. However, missing the target does not necessarily mean that the credibility of

the central bank will decline. If the central bank gives sufficient disclosure on the reason of missing the target

to the public, the credibility of the central bank is not necessarily undermined (ibidem).

In Loayza et al. (2002, p. 3) the absence of fiscal dominance is assumed to be a characteristic of inflation

targeting. Not pursuing different monetary objectives at the time is seen as a characteristic as well. However, in

literature a general consensus exists that these two conditions should rather be seen as prerequisites to a

successful implementation of inflation targeting, as will be explained in paragraph 4.4.5.

In the literature, nowhere has a distinction been made between characteristics of the framework adopted in

advanced and emerging market economies.

4.4.5. PreconditionsTo design the ideal situation for adopting inflation targeting, it is of great importance to know the prerequisites

that are essential to the success of the inflation targeting regime. This paragraph will discuss the most

important preconditions to the framework based on literature from Mishkin (2000), Kadiolu et al. (2000),

Masson et al. (1998) and Svensson (2010b, p. 48).

It is of great importance that the central bank is both independent as well as accountable for missing a target.

Moreover, it should be clear that the primary goal of the central bank is to guarantee price stability (Lucotte,


10/27

10 | P a g e

2010, p. 3). Kadiolu et al. (2000, pp. 18-20) state that the central bank should have instrument independence,

so that it is free to determine the instruments which it uses to conduct its monetary policy. In the case of

inflation targeting, the central bank will reduce deviations from the target by changing the nominal interest

rate (Loayza et al., 2002, p. 7). Absence of fiscal dominance and a strong financial system are indispensable

elements for central bank independence.

Mishkin (2000, p. 107) points out some preconditions based on the disadvantages he discusses, of which anoverview can be found in paragraph 4.4.9. First of all, he thinks it is necessary that the government does not

control prices of some goods as this makes it even more difficult to control inflation. Another possibility is to

exclude the goods of which the prices are being controlled by the government from the basket of goods on

which the index price is calculated.

Furthermore, Mishkin states it is important to limit fiscal debt, for the reasons stated in paragraph 4.4.9.

Therefore, it can be desirable to establish an institution that controls the government expenditures. To avoid a

financial crisis caused by exchange rate fluctuations a strong financial system is needed. This idea is confirmed

by Loayza et al. (2002, pp. 9-10), as it is said that the central bank needs to make sure that the exchange rate is

not misaligned. Loayza et al. add that once the framework has been implemented fiscal imbalances need to be

avoided, because this leads to output instability.

Masson et al. (1998, p. 35) define that to avoid fiscal dominance the government should raise taxes and should

not rely on the revenues of seignorage. Seignorage is the revenue that arises for the government due to the

issue of extra money to monetize the fiscal debt via a devaluation of the currency. Moreover, well-developed

models to make forecasts are also needed to set the target efficiently (Svensson, 2010b, p. 48).

Furthermore, Kadiolu et al. (2000, pp. 18-20) and Svensson (2010b, p. 48) argue that it is possible that next to

the inflation target, other variables are targeted, such as GDP growth, the exchange rate, etc. However, if a

conflict arises between those targets, it should be clear which target is the primary target (Masson et al., p. 35),

which raises the transparency and credibility of the central bank.

Mishkin (2004, pp. 6-14) adds that a strong fiscal and financial system and reliable monetary institutions are

indispensable in making inflation targeting work. If the financial system is weak, investors will not invest in

securities issued by the emerging market economies, unless the investors are risk-lovers. However, the

majority of investors will dump these securities or simply not buy them, which is called a sudden stop. This

phenomenon will be discussed in detail in paragraph 4.6.

Lucotte (2010, p. 3) adds that to ensure good inflation forecasts well-developed econometric models are

needed to forecast inflation. In addition, he also states that it is better to adopt inflation targeting after some

disinflation has been achieved.

4.4.6. Features of inflation targeting countriesThis paragraph overlaps with the former paragraph. However, this section discusses characteristics of the

countries that adopt inflation targeting, while the previous section lists preconditions that need to be fulfilled

to be successful in inflation targeting. This means that these preconditions can be considered characteristicsof the inflation targeting countries. However, the following paragraph describes features that are not

necessarily prerequisites of inflation targeting. An example will explain this: most inflation targeting countries

have a high GDP, however this is not a precondition. It is possible that a country with a small GDP is successful

in inflation targeting as well.

Ptursson (2005, pp. 13-15) lists some features by which countries that adopt the framework of inflation

targeting are characterized. The countries are more open, making them more vulnerable for external shocks.

Furthermore, the government debt is usually lower. As stated in paragraph 4.4.5, the absence of fiscal


11/27

11 | P a g e

dominance is critical for the success of the regime. Therefore, it is logical that countries under inflation

targeting have a lower government debt to avoid fiscal dominance. Furthermore, these countries usually have

a higher GDP, except for the emerging market economies. Finally, Ptursson states that the more independent

a central bank is, the higher the pro bability of adopting inflation targeting.

Kadiolu et al. (2000, p. 21) add that these countries mostly apply fully floating exchange rates and use the

short-term interest rate as instrument. The monetary policies are forward-looking as the inflation target isdetermined by focusing on the inflation expectations. Furthermore, inflation targeting countries adopt the

framework only after they have already realized some disinflation. Masson et al. (1998, pp. 35-36) give an

economic rationale for this. If the countries adopt the framework when they face high inflation levels, the

inflation will need to be lowered substantially. Therefore, the central bank will announce a lower inflation

target, but it risks here that the target will be missed as higher inflation is more difficult to control. On the

other hand, the central bank could opt for lowering the inflation in little steps and announce a high target, but

the public will not accept this target and lose confidence in the central bank, which will lower the credibility of

the central bank.

Furthermore, inflation targeting countries typically have more domestic currency denominated bonds in their

portfolio, in contrast to countries that target for example money-growth (De Paoli et al., 2010, p. 3). The reason

is very simple. If the country is hit by an adverse real shock, which increases the inflation, this will lead to adomestic currency appreciation. This can be explained quite easily: as the inflation has risen, the central bank

needs to raise the nominal interest rate to offset the rise in inflation. A rise in the interest rate will lead to a

currency appreciation. If the domestic currency appreciates, the domestic currency denominated assets of the

investors will have a higher value. Therefore, the portfolio will consist of more domestic bonds denominated in

the domestic currency (De Paoli et al., 2010, pp. 19-20).

If however the country encounters a monetary expansion, the domestic currency will depreciate. A monetary

expansion leads to a decrease in the domestic nominal interest rate. This decline increases the domestic

consumption, causing a higher demand for money. Due to the higher demand for money, the price level

increases in the long run, which leads to a domestic currency depreciation. As the portfolio of the investors in

inflation targeting countries consists mainly of domestic currency bonds, this depreciation will reduce the value

of the assets of the investors and leads to a decrease in the net external wealth position of the country. Hence,a monetary expansion can be seen as beggar-thy-self policy (De Paoli, 2010, pp. 21-22).

4.4.7. Factors influencing the adoption of inflation targetingLucotte (2010) makes an interesting study on the circumstances which influence the adoption of inflation

targeting. He argues that the more independent a central bank is, the more probable it is that the country will

adopt inflation targeting, which was already stated in paragraph 4.4.6 as well. In this way, the central bank will

not pursue an overly expansionary monetary policy, due to which the inflation will increase. In addition, an

independent central bank will be less tempted to finance the fiscal debt (Lucotte, 2010, pp. 4-5).

Another influencing factor is the number of people that participate in the decision process in politics, also

called the number of veto players. If this number is high, this means that a high number of people can block a

decision and that if a decision has been taken, that it cannot be reversed easily. This means that the power to

make decisions is more fragmented and therefore, the central bank will not suffer from a high degree of fiscal

dominance, which will incite the country to adopt inflation targeting (Lucotte, 2010, pp. 7-8). Moreover, if a

country is highly decentralized, the central bank will be more independent as well, because the different parts

can be seen as additional veto players (Lucotte, 2010, pp. 8-9).

Furthermore, a strong banking system is important. Lucotte (2010) gives 5 reasons put forth in the literature.

First of all, if the banking system is strong the banks will adopt the monetary policy set out by the central bank.


12/27

12 | P a g e

If however, the banking system is weak, it is possible that the banks misunderstand the monetary policies

conducted by the central bank and in this case, the monetary policy transmission mechanism does not work

properly. The transmission mechanism is the mechanism that describes how the conducted monetary policy

influences the inflation expectations and the aggregate demand via the effects on short-term real interest rates

and exchange rates (Svensson, 2000a, p. 158). Secondly, if the banking system is weak, it will be more difficult

for the central bank to raise the interest rates, as this tarnishes the profitability of the banks and could lead to a

collapse. Therefore, the central bank will opt not to raise its interest rates and will miss easily the target.Thirdly, a weak financial system raises the probability of a sudden stop. Together with the phenomenon of

liability dollarization, this will lead to the depreciation of the currency. Finally, if the financial system is strong,

no incentives to finance the public debt can be identified (Lucotte, 2010, pp. 5-6). So, according to the

literature, the stronger the banking and financial system, the more it is probable that the inflation targeting

framework will be implemented by the country concerned. However, Lucotte succeeds to weaken this

argument and could deduct out of his econometric research that this relationship does not hold true (Lucotte,

2010, p. 26).

Another misconception, according to Lucottes survey, is that a country will be more likely to adopt inflation

targeting if the political environment is more stable, as this would have a positive impact on the independence

of the central bank (Lucotte, 2010, pp. 6-7, 26).

4.4.8. AdvantagesIn this chapter an analysis of the advantages of inflation targeting will be made. It seems important to indicate

these advantages, as it is essential in promoting the inflation targeting framework.

Mishkin (2000, pp. 105-106) puts forth some advantages of inflation targeting. In fact, the two main

advantages are simply shortcomings of exchange rate and monetary targeting. As already mentioned in

paragraph 4.3, for central banks that peg their exchange rate it is impossible to pursue their own monetary

policy. Inflation targeting countries can set their own policy and act to mitigate the undesired effects of

external shocks in the domestic economy.

One can argue that monetary targeting has this advantage as well, but the success of monetary targeting

depends on the relationship between money growth and inflation. If this relationship is not stable, monetary

targeting will fail. In the framework of inflation targeting this problem is not encountered, as an intermediate

target is used that is determined based on as much available information as possible. Thus, as Ptursson (2005,

p.8) states, inflation targeting is a best of worlds-solution.

Mishkin (2000) argues that inflation targeting is very transparent because it can easily be understood by the

public by publishing inflation reports. In addition, this framework improves the accountability of the central

bank, as it clearly states that price stability is the primary goal and it announces a numerical inflation target. If

the central bank avoids a history of missing the target, inflation targeting can be named a success.

A last advantage is that the time-inconsistency trap can be avoided, as inflation targeters emphasize that they

want to control inflation in the long run. An overly expansionary monetary policy, due to pressures from

politicians for increasing output growth or diminishing unemployment, will not be carried out because the

primary goal under the inflation targeting framework is clearly price stability.

4.4.9. CriticismTo give a well-balanced exposition, it is important to also discuss criticism expressed on the inflation targeting

framework. A good overview of the disadvantages of inflation targeting can be read in Ptursson (2005, pp. 49-

56). Ptursson formulates each time a counterargument.


13/27

13 | P a g e

First of all, one can argue that the framework is not flexible because it does not pay enough attention to

evolutions in the real economy, as it should only focus on the inflation level. In this way inflation targeting

would be responsible for a decline in the growth rate of the economy and more pronounced variations in the

business cycle. Ptursson argues that since an intermediate target is being used and since the target is

announced on a medium-term basis, inflation targeting is in fact a very flexible framework.

On the other hand, Ptursson indicates that some people state that the framework is too flexible, due to thefact that for defining the target all relevant information is used. This would lead to higher inflation, because it is

too difficult to control inflation in a proper way. Ptursson formulates a counterargument relating to two

specific characteristics of the inflation targeting framework, namely accountability and transparency. These

two characteristics make that a central bank considers discipline to be very important, which facilitates

controlling inflation.

Secondly, some preconditions are indispensable to the success of the framework. Ptursson argues that some

countries adopted inflation targeting without satisfying these preconditions. So Ptursson is not convinced of

the value of these preconditions. As will be discussed later on, it is not necessary to fulfill all preconditions.

However, a major part of the preconditions named in paragraph 4.4.5 should be fulfilled.

Thirdly, it is said that due to transmission lags the effect of monetary policy on inflation can only be seen after

some time, which means that inflation is very difficult to control. In addition, the length of the lag is dependent

on the business cycle and on the credibility of the central bank. Ptursson points out that in other monetary

policies these transmission lags cause problems as well. However, he admits that central banks of emerging

market economies that target inflation face more difficulties to influence and forecast inflation. It is for

example possible that the government fixes the prices of some goods. If these goods are enclosed in the basket

of goods, which prices are used to calculate the price index, a bias in the price index can exist and the real rate

of inflation can differ significantly from the inflation reflected by the price index.

Svensson (1996, p. 26) argues that controlling inflation is not easy due to the lags in the transmission

mechanism as well. The instrument reacts only after some time to the monetary policy. Therefore, it can be

difficult to reach the inflation target, as at the time the instrument reacts to the policy, other economic

circumstances may exist. Svensson states that this problem can be solved by using inflation forecasts. In thisway, the instrument will be set so that the inflation forecast equals the target and the target will be reached

more easily.

The last criticism treated in Ptursson (2005, pp. 51-54) is the high volatility of the exchange rate. This poses a

big problem in emerging market economies, as the financial system is often very weak and loans from foreign

investors are very common. The latter often means that the balance sheet is denominated in a foreign

currency, which makes the country very vulnerable for sharp fluctuations in the exchange rate. If the domestic

currency depreciates, the currency in which the foreign debt is denominated appreciates. This will make it

more difficult to repay the debt and can even lead to a financial crisis in the domestic country. Ptursson

admits this problem and warns for focusing too much on the exchange rate because this will evolve towards an

exchange rate targeting regime. Thus, he states that modest fluctuations in the exchange rate are preferable.

Mishkin (2000, pp.106-107) and Kadiolu et al. (2000, pp. 11-13) distinguish the following disadvantages:

(1) The framework is too rigid. However, one can argue that constrained discretion is a better term, as the

framework does not provide nice rules to apply in reaction to shocks. The central bank has to use all available

information to conduct its monetary policy (Kadiolu et al., 2000, p. 13).

(2) There is too much discretion in the framework, which will incite policy makers to pursue an overly

expansionary monetary policy. Again, one can refer to the term constrained discretion as transparency is one


14/27

14 | P a g e

of the main characteristics discussed in paragraph 4.4.3 and makes it tempting for the central bank to follow an

overly expansionary policy (Kadiolu et al., 2000, p. 13).

(3) There is increased output instability.

(4) Economic growth diminishes.

(5) As inflation is hard to control due to the transmission lags, the central banks accountability is weak. Thisdisadvantage is of great importance for emerging market economies, because these economies are

characterized by higher inflation levels, which involve a higher possibility of large forecast errors and frequent

target misses. Further on, in emerging market economies, prices of goods are more government controlled

than in advanced economies, which makes it even more difficult to control inflation, as the inflation level is

biased due to these government controlled prices (Kadiolu et al., 2000, p. 12).

(6) Inflation targeting does not prevent fiscal dominance, because the government is not prevented from

pursuing an irresponsible fiscal policy. In this case, the government has to repay the debt or erode the debt

through a depreciation of the exchange rate, which will lead to a higher inflation and missing the target.

(7) As inflation targeting involves floating exchange rates, the framework will lead to financial instability, if the

financial system is weak. However, Mishkin notes that the central bank should not pay too much attention onthe exchange rate; otherwise the central bank adopts the exchange rate targeting framework.

Mishkin argues that the first 4 disadvantages do not occur in reality if the framework is properly implemented.

It can be concluded that generally Mishkin, Kadiolu et al. and Ptursson recognize the same disadvantages.

Hammond (2011, pp. 16-17) gives some challenges for inflation targeting. The first challenge that has been

pointed out is that typically inflation targeting succeeds in guaranteeing stable and low consumer prices.

However, as the financial crisis in 2008 has dramatically shown, the asset prices, for example housing prices,

can vary extremely. Inflation targeting should in the future focus more on less volatile asset prices. This could

be done by targeting the inflation over a longer time horizon.

Secondly, related to the first challenge, it is questioned whether these asset prices should not be excluded from

the inflation measure, as they are more volatile. Furthermore, one can argue that only domestic inflation

should be included in the inflation measure, to avoid imported inflation.

Hammond adds that one should be aware of deflation as this implies interest rate cuts towards the zero bound.

If on top, the economy is in a recession, it is possible that the economy faces the liquidity trap. This occurs

when further expansionary monetary policy is needed due to the recession, but this is impossible as the

interest rate is set close to 0% (Loayza et al., 2002, p. 10). In this case, non-conventional monetary policy has to

be executed, such as a foreign-exchange intervention to avoid currency appreciation, fixed-rate lending at

longer maturities etc. (Svensson, 2010a, p. 5).

In the light of the recent financial crisis, Svensson (2010a, pp. 4-6; 2010b, pp. 52-60) argues that monetary

policy is not the cause of the financial crisis, but rather weak supervision and regulation of the banking sector.

He adds that one should clearly see the difference between financial stability and monetary policy, which have

different objectives and different instruments to conduct their policies. Svensson argues however that both are

closely interrelated. Financial factors are indicators of the state of the economy and have an impact on the

target variables, i.e. inflation and the real economy (Svensson, 2010a, p. 2). Therefore, more research should

be done on the influence of these financial factors on inflation and the real economy.

Moreover, instead of inflation targeting the price-level could be targeted as well. Advantages of this approach

are that the price-level can be forecasted with a lower risk of making a wrong forecast and that price-level

targeting automatically stabilizes the economy. If the price-level falls below the target, inflation expectations


15/27

15 | P a g e

increase, while the nominal interest rate remain the same. Due to the rise in inflation expectations, the real

interest rate will decrease, which will stimulate the economy. In this way, the price-level will return to the

target. This is an interesting feature of price-level targeting when the price-level is low and risk of deflation

exists. In this case, automatic stabilization will take place and the classic liquidity trap can be avoided

(Svensson, 2010b, pp. 50-51). Moreover, price-level targeting results in lower long-run variability of the price

level, but higher short-run output and inflation variability than inflation targeting. This can be explained very

easily as under the inflation targeting framework higher-than-average inflation need to be followed by averageinflation, while under price-level targeting (in which the objective is to stabilize the price-level) higher-than-

average inflation should be compensated by lower-than-average inflation, leading to higher inflation variability

in the short-run and also higher output variability (Svensson, 1999b, p. 277). However, in the literature almost

no papers exist on the disadvantages of price-level targeting.

4.5.Operational framework4.5.1. ResultsSvensson (2010b, p. 4) argues that inflation targeting has generally succeeded in stabilizing output and

inflation. In addition, growth, productivity, employment and other measures of economic performance have

not been negatively affected after the adoption of inflation targeting in the particular country.

This section in the master thesis will give an overview of the results of inflation targeting based on empirical

research performed by different authors.

4.5.1.1. InflationFraga et al. (2003, p. 4) and Ptursson (2005, pp. 21-35) argue that a decrease in inflation in the inflation

targeting countries is being observed from the period the moment the countries adopted inflation targeting

until 2005. However, most countries adopt inflation targeting after they successfully realised some disinflation,

which makes it very difficult to conclude whether the decline in average inflation is due to a general trend or

due to the framework. Therefore, Ptursson carried out some further research and he found out that in

industrialized countries inflation targeting was primarily able to lock in the disinflation already achieved in

previous stages. In emerging marketing economies inflation targeting is able to reduce the average level of

inflation. Further on, Ptursson (ibidem) argues that inflation targeting leads to inflation stabilization.

Bernanke et al. (1999) state, in accordance to the arguments of Ptursson, that inflation targeting is only

responsible for a small proportion of the decrease in inflation. Inflation targeting lowers the inflation

expectations only after some time lag (Bernanke et al., 1999). This idea is confirmed by Svensson (2010b, p. 8,

10-11). However, Svensson adds that the effect in emerging market economies is significant and that inflation

expectations are more effectively anchored in those economies.

Hammond (2011, p. 7) points out that inflation targeting is mostly adopted when inflation was already low and

that inflation targeting is especially effective in anchoring the inflation expectations around the target and keep

the inflation low in this way.

4.5.1.2.Real economyPtursson (2005, pp. 35-38) and Svensson (2010b, p. 11) state that it is a general misconception that inflation

targeting is harmful for growth, because it would focus too much on inflation, thus undermining output

stability. Ptursson and Svensson argue that there is no evidence supporting this statement. Ptursson states

that it is difficult to express an opinion on this relationship as no country has yet adopted the framework for a

period of time that is long enough.


16/27

16 | P a g e

4.5.1.3. Cost of disinflationThe well-known Phillips-curve shows clearly the trade-off between unemployment and inflation. If an economy

is characterized by high inflation, unemployment will be low. Further on, it can be argued that inflation is

procyclical and unemployment anticyclical (Meeusen et al., 2006, p. 275). Based on these two facts, one can

conclude that a decrease in inflation will involve a decrease in output (as inflation is procyclical) and an

increase in unemployment in the short run. In the long run, the trade-off disappears (Issing, 2004, p. 170).Thus, the loss in output and increase in unemployment are the costs associated with disinflation, also called the

sacrifice ratio (Bernanke et al., 1999). As inflation targeting leads to higher credibility of the monetary policy,

the expectations of the public will adapt more easily. Therefore, it is argued that inflation targeting reduces the

cost of disinflation. However, empirical research shows that the sacrifice ratio is not lowered because of

inflation targeting. This means that realizing a decline in inflation always implies a loss in output and an

increase in unemployment in the short run (Bernanke et al., 1999, p. 266).

4.5.1.4. Interest rates and exchange ratesWhen a country adopts inflation targeting the credibility of the central bank increases. Ptursson (2005, pp. 38-

43) argues that this lower credibility will lower inflation expectations and the inflation risk premium in the

nominal interest rate, which in turn will lead to a decrease in the nominal interest rate. Ptursson shows that

the fall in the nominal interest rate is however higher than what can be expected based on the theory.

Furthermore, exchange rate volatility is lower.

4.5.2. Defining the targetIt is of great importance that the target is defined properly because an unrealistic target will not contribute to

the credibility of the central bank as the probability of missing the target is quite high. For example, if inflation

in a particular country is 15 %, it would be unrealistic to target an inflation rate of 2 % next year. This will

definitely lead to a failure and will decrease the credibility of the central bank. Therefore, it seems appropriate

to include a paragraph on defining an appropriate target in the master thesis.

First of all, the central bank needs to find out how it is going to measure the inflation: based on headline CPI or

another measure. It is possible that the central bank measures inflation based on the prices of a basket ofgoods in which food and energy prices are excluded to eliminate the most volatile prices, called core CPI.

Mostly, headline inflation is used to determine the target. However, in some developing economies the CPI

basket of goods consists of 40 % food. Excluding food from the basket would not give a good measure of

inflation. In this case, headline inflation is preferable. Another possibility is to use the GDP Deflator. H owever,

this measure is only published on a quarterly basis, while CPI is published on a monthly basis and is better

known by the public (Hammond, 2011, pp. 9-10). Mostly, the inflation targeting countries publish both core

and headline CPI.

Normally the target is announced for some years in advance and from time to time the target is revised. The

identity of the target setter is defined in institutional arrangements: in some countries the government sets the

target, in others the central bank is responsible for, like in Chile (Horvth, 2011, p. 4). It is also possible that the

central bank confers the target with the government.

The target is dependent on multiple determinants, such as past inflation, GDP growth, credibility... The central

banks define which determinant they judge to be important in setting the target. The main determinant in

Chile is deflation risk. If the Chilean central bank fears deflation, the target will be higher (Horv th, 2011, p. 4).

The target in emerging market countries is usually somewhat higher than in advanced economies.

Furthermore, the country needs to choose whether it will use a point target or a range. If the country opts to

target a point target, the central bank has to determine how much the actual inflation may deviate from the


17/27

17 | P a g e

target, also called tolerance bands (Dotsey, 2006, p. 11). In addition, using a point target needs a very precise

monetary policy not to miss the target. On top, if the tolerance band is symmetrical, the central bank gives a

clear signal that it puts equal weight to avoiding inflation and deflation (Hammond, 2011, p. 10). If the country

handles an inflation range (which is very common), the range is wider (typically central target +/- 1 %) in

emerging market countries, because the economy is more volatile. The width of the target range depends on

the level of inflation and the inflation volatility. Thus, if both determinants are high, the width will be higher as

well. The target range in Chile was [2%; 4%] in 2008. Thus, the central target is 3 % and the width of the targetis +/- 1 % (Horvth, 2011, p.5).

Moreover, it is important to establish the right numerical value for the target. As already explained in

paragraph 4.4.2, inflation between 0% and 3% is considered to be desirable. These values are good candidates

for the numerical value of the target in developed countries. However, central banks must be aware of the risk

of setting the target too low. If the economy is then hit by a shock which decreases inflation, it is possible that

the economy suffers from deflation, which leads to financial instability as the nominal value of assets decline,

while that of debt remains unchanged (Mishkin, 2001, p. 1; Hammond, 2011, p. 8). Therefore, it is better to set

a positive inflation target as this reduces the risk of deflation. Bernanke (2004, p. 166) calls this positive

inflation target a buffer zone against deflation. In addition, the inflation measures overstate the inflation

typically by 0.5 % points. Targeting a positive inflation level corrects for this bias (Hammond, 2011, p. 8).

In addition, the target horizon needs to be defined. This is the time within which the target needs to be

reached (Svensson, 2010b, p. 41). This horizon is dependent on multiple factors. If for example the inflation

deviates a lot from the target, it is better to use longer target horizons to avoid target misses (Loayza et al.,

2002, pp. 12-13).

Svensson (2010b, p. 41) points out that in the past this target horizon was fixed, i.e. 2 years, whereas nowadays

the target horizon is on the medium-term. The central banks realized that the target horizon is dependent on

the state of the economy to determine the target horizon. Therefore, the target horizon is set more abstractly.

To conclude, the process to define the target is similar in all countries (Horvth, 2011, pp. 10-14).

4.5.3. Decision-making processUnder an inflation targeting regime, the central bank meets preferably on fixed dates to determine the

monetary policy and to decide on how to react on shocks by adapting the interest rate in such a way that the

inflation target is reached (Hammond, 2011, p. 10). The power of decision-making can be delegated to the

Board of the Central Bank or to a Monetary Policy Committee, consisting of experts on monetary policy. This

decision-making committee exists on average of 7 people and the number of people in this body is

independent on the size of the country. Normally, a decision is made based on voting or a consensus has to

exist in the committee (Hammond, 2011, pp. 10-11).

As stated in paragraph 4.4.2, one of the characteristics of the central bank is accountability. The central bank is

accountable towards the government and the public. To fulfill this characteristic, the central bank needs to

undergo governmental hearings and needs to publish open letters to account for its monetary policy and for

deviations from the target. Corrective actions need to be formulated in the open letters if the target is missed

(Hammond, 2011, pp. 14-15).

All publications and communications from the central bank towards the public include preferably an Inflation

Forecast Report, in which inflation, growth and other economic variables are predicted and justification for

policy decisions is given. On top, many central banks publish minutes of the meeting of the decision-making

body (Hammond, 2011, pp. 15-16).

4.6.Emerging market economies


18/27

18 | P a g e

4.6.1. CharacteristicsThis part will try to describe the general macro-economic situation in the emerging market economies.

However, it is clear that emerging market economies are not as homogenous as advanced economies. Giving

some key characteristics of emerging market economies will be difficult for this reason. Nevertheless, an

attempt will be made to define these characteristics. These characteristics can help identify how the

implementation of the inflation targeting framework will differ from that of advanced economies, based on thedifferences between both types of economies.

Despite the heterogeneity of the emerging market economies, these economies rely more on indirect

monetary policy instruments, need to reform their financial sector and appeal more to the international capital

market (Masson et al., 1998, p. 36).

Fraga et al. (2003, p. 4) observed larger deviations from the target and these deviations occur more commonly

in emerging market economies.

Mishkin (2004, p. 3) describes some extra features. He states that emerging market economies may suffer from

weak fiscal, monetary and financial institutions. Often the financial institutions are not very strictly supervised

and regulated. It is possible that advanced economies face these problems as well, but to a lower degree.

These shortcomings can easily cause high inflation and currency crises.

A currency crisis will incite the people to switch to a foreign currency, which is also called a currency

substitution (Mishkin, 2004, pp. 4-5). In addition, it is common that the liabilities on the balance sheets of the

domestic firms are denominated in USD. This phenomenon is also called liability dollarization. Currency

depreciation is a decline in the value of a currency and makes it harder to repay the loans denominated in USD.

Typically, the assets are denominated in the domestic currency and due to the currency depreciation. The

assets will increase at a slower pace than the increase in worth of the liabilities. In addition, a currency

depreciation leads to higher inflation because of imported inflation due to higher import prices (Miskhin, 2004,

p. 22). As a consequence of this negative shock, the level of investment will decrease and this can lead to an

economic recession. Important detail is that in the balance sheets of emerging market economies, the

proportion of debt is substantially higher than the proportion of assets, which means that such a shock has

even more negative effects on the balance sheets (Mishkin, 2004, p. 8). This is the reason why emerging

market economies do not like exchange rate flexibility and are often still reluctant to adopt inflation targeting.

The last typical characteristic described by Miskhin (2004, pp. 5-6) is the phenomenon of a sudden stop, which

is a dry up of capital inflows in the capital market. Typically the elements causing a sudden stop are external to

the emerging market economies, e.g. the financial crisis in 2008, a crisis in the world financial institutions

Such a crisis makes it less attractive for investors to invest in risky securities of the emerging market economies

and these investors are more likely to invest in more prudent investments if the world economy is in a

recession. Such a sudden stop will also lead to a currency depreciation and cause an economic contraction, due

to the dynamics explained above in this paragraph.

Finally, it is possible that some emerging market economies operate under an IMF conditionality program, e.g.

Brazil. In this case, the implementation of inflation targeting is hampered because this program implies anupper limit on the central banks net domestic assets and a lower limit for the volume of net international

reserves held by the central bank. This floor and ceiling pose some problems. First of all, the lower limit on the

net international reserves means that the central bank cannot easily react on fluctuations in the exchange rate

if these can imply a target miss by using the international reserves because the IMF has set an explicit quantity

for the level of net international reserves. In addition, the IMF fixes the maximum amount of net domestic

assets based on a backward-looking mechanism, while the inflation target is set using forward-looking models

that estimate the future inflation. Thus, the central bank cannot execute a monetary policy in complete

freedom in this case (Loayza et al., 2002, pp. 12 -13).


19/27

19 | P a g e

4.6.2. Implementation of inflation targetingBased on the characteristics analyzed above, it is clear that emerging market economies have weak institutions.

Therefore a reform program should be executed (Masson et al., 1998, p. 36).

A fiscal reform is necessary if fiscal dominance has a large influence on monetary policy. The government needs

to conduct a more transparent fiscal policy, in a way that the budget deficit can be controlled more effectively

and fiscal dominance can be avoided (Mishkin, 2004, p. 9).

As explained above, the financial system is often weak in these countries as well, therefore it is essential that

the banking sector are regulated and supervised effectively. A well-known problem in the banking sector is

moral hazard, which incites the bankers to take too much risk. This problem exists due to full deposit insurance,

which means that in case of a bank run, the deposit holders are insured for the full amount on their bank

deposit. Therefore, Mishkin (2004, p. 9) thinks the regulations concerning such insurance should be limited to

avoid moral hazard by the bankers, because this increases the possibility on a bank run. He adds that a policy is

needed to limit liability dollarization.

A policy that makes the economy more open is needed according to Mishkin (2004, pp. 9-11). The economic

rationale is as follows: if the economy is more open to international trade, this means that the country trades

more with other foreign economies. To facilitate trade, the assets on the balance sheets of the domestic firms

are therefore denominated in the foreign currency as well, usually USD. If in that case the currency depreciates,

the value of the liabilities as well as that of the assets will increase, which makes the shock less severe. If then a

sudden stop occurs, the currency need to be depreciated less to move towards the equilibrium state.

Lastly, the instrument independence of the monetary institution, as well as the primary goal of monetary

policy, i.e. price stability, should be guaranteed by law and clearly communicated towards the public (Mishkin,

2004, pp. 11-14). If the central bank had in recent years very low credibility, the private agents will doubt that

the central bank will meet the target now. Thus, the private agents think that the inflation in the future will be

higher than the target. This implies an increase in inflation expectations, which will lead to higher inflation. For

this reason, it is really essential to build credibility of the central bank (Fraga et al., 2003, p. 14).

Mishkin (2004, p. 11) makes an important remark when he states that these reforms do not necessarily need tobe implemented before adopting inflation targeting.

In fact, it cannot be said that based on satisfying the preconditions or not, an emerging market economy will

respectively succeed or fail in inflation targeting. For example, it is possible that there is fiscal dominance on

the central bank without inflation targeting to fail (Masson et al., 1998, p. 36).

Furthermore, it is also important that the central bank tries to avoid a depreciation of the currency, as it is clear

that a depreciation has negative consequences. However, the central bank should make sure that it does not

stress the exchange rate too much. Otherwise the exchange rate will become the nominal anchor instead of

the inflation target.

Another important issue is that the central bank needs to be aware of the nature of the shock that causes thecurrency depreciation, because the central bank needs to react appropriate. Two examples will make this clear.

If the domestic currency depreciation is caused by a portfolio shock, inflation will rise. Thus, inflation needs to

be lowered in order to avoid a target miss. This disinflation can be realized by increasing the interest rate. If

however a negative terms-of-trade effect causes the domestic currency to depreciate, the interest rates need

to be lowered. A negative terms of trade effect means that the goods exported by the domestic country are

less demanded by the rest of the world, which leads to a decline in the aggregate demand. This means that

prices and output will decline. To avoid a recession, the consumption needs to be stimulated again. Therefore,

the interest rates need to be lowered, which will raise the aggregate demand again (Mishkin, 2004, pp. 24-25).


20/27

20 | P a g e

4.7.Inflation targeting during the financial crisis of2008As the financial crisis was the first major shock since inflation targeting has become commonly adopted, it can

be seen as the first real test. Therefore, it seemed interesting to include a chapter on this topic in the master

thesis. First of all, a survey on the performance of inflation targeting countries will be discussed. Secondly, as

the master thesis will focus on inflation targeting in emerging market economies, an attempt will be made to

perform own empirical research on the performance during the financial crisis of inflation targeting emergingmarket economies compared to the performance of advanced economies that have adopted the framework of

inflation targeting. This research will be executed based on the method as described in Carvalho de Filho

(2011).

4.7.1. PerformanceIn a recent study performed by de Carvalho Filho (2011), the performance of inflation targeting countries

during the financial crisis of 2008 is compared to the performance of non-inflation targeting countries during

the same period. This survey demonstrates that the inflation targeting countries performed better than the

non-inflation targeters and confirms the statement of Svensson (2010a, p. 1) that inflation targeting was the

most optimal framework to be executed before, during and after the financial crisis.

The GDP growth of inflation targeting countries recovered more rapidly compared to non-inflation targeting

countries (de Carvalho Filho, 2011, pp. 5-6). Furthermore, in the period prior to the crisis the nominal and real

interest rates were higher in these countries compared to their peers. These higher interest rates made it

easier for inflation targeting countries to execute larger cuts in the interest rates without having to fear

deflation. In this way, it was possible for the inflation targeting countries to cut the interest rates about 1.5 %

more than the non-inflation targeting countries in the sample (o.c., p. 12).

Further on, inflation rates decreased less in the inflation targeting countries, as the inflation targeting countries

had higher interest rates, allowing for higher interest rate cuts, as already explained. As a consequence these

countries did never face deflation during the crisis, while their peers in the sample did (o.c., p. 13).

During the crisis, the real exchange rate of inflation targeting countries depreciated more sharply. However this

depreciation was not accompanied by an increase in the risk premiums. Therefore, this depreciation was

favorable, as this depreciation made domestic goods and exports more attracting (o.c., pp. 14-15).

In this survey it is made clear that inflation targeting performed better than other monetary policy frameworks

during the crisis. A good conclusion to this section cannot be formulated better than by this citation of the

Governor of the Bank of Canada Mark Carney: Just as inflation targeting has proven its ability to prevent the

entrenchment of high and volatile inflation, it also has the power to prevent the onset of persistent deflation.1

4.7.2. Empirical researchAs already stated, own empirical research will be executed wherein the performance of inflation targeting

emerging market economies will be compared to the performance of inflation targeting advanced economies.

The research will be strongly based on the survey performed by de Carvalho Filho (2011). The goal of thissurvey is to find out whether the performance during the crisis was different between the 2 samples.

As the total sample includes only 23 countries, if the criteria set up by de Carvalho Filho (2011) are strictly

applied, it is possible that the sample is for statistical reasons not large enough (o.c., p. 27). However, at the

beginning of 2011 27 countries had adopted inflation targeting. Therefore, it does not seem appropriate to

loosen the criterion of de Carvalho Filho (2011, pp. 4-5) that the nominal GDP in 2002 of the countries needs to

11 CARNEY, M. 2009. Governor Carney Remarks, Remarks to the Halifax Chamber of Commerce, Halifax, Novia Scotia, 27 January 2009 ascited in (de Carvalho Filho, 2011, p. 2)


21/27

21 | P a g e

be higher than USD 10 billion. The other criterion, which defines that 3 out of 4 variables, namely

unemployment rate, industrial production, policy interest rates and the credit default swap spread, should be

available, will be adopted in the survey as well.

The period covered in the survey will be 2000 to 2010. Differences in real GDP, industrial production, nominal

and real interest rates, 12-month inflation, deflation scares, real effective exchange rate, the EMBI spread and

CDS spread between the emerging market and advanced economies will be investigated. The latter twoparameters are parameters that measure the risk perception of the market. The figures for these parameters

will be obtained by consulting the IMF Database and Eurostat. Based on the results, a conclusion will be drawn

on whether there was a difference in performance of both groups of countries.

4.8.Case study: ChileIn this chapter, the Chilean experience with inflation targeting will be described. The reason why Chile was

chosen cannot be better expressed by Mishkin (2004, p. 15): Chile is the poster child for inflation targeting

regimes in emerging market countries because it has had great success in lowering inflation to levels found in

advanced countries, while at the same time experiencing very rapid economic growth. This shows that Chile

was successful in adopting inflation targeting and therefore it seems a good example of inflation targeting best

practices. This will facilitate to draw conclusions on the reasons for which the framework had less success in

Turkey. This will be done by comparing the Turkish and Chilean experience with inflation targeting.

Chile adopted the inflation targeting framework in 1990. In previous years, Chile was an exchange rate targeter

(Ptursson, 2005). In 1989, Chile endorsed a new central bank legislation, which gave the Chilean central bank

independence and made price stability the primary objective. The target is set by the central bank itself, so the

government does not interact in the process (Hammond, 2011, p. 8). In addition, Chile had a budget surplus of

somewhat less than 1 % from 1991 to 2002 and had a strong financial system, which was regulated and

supervised in a very efficient way. These factors were responsible for the success of inflation targeting in Chile

(Mishkin, 2000, pp. 107-108; Mishkin, 2004, pp. 15-16).

Chile evolved towards inflation targeting gradually, which proves the concern of the central bank for both

output stabilization and inflation (Loayza et al., 2002, pp. 12-13). In the years before adopting the inflation

targeting framework, the country realized some disinflation. In 1990, the Chilean central bank adopted inflation

targeting. The first inflation target was announced in 1991 and required the inflation to be in the range of [15%;

20%] (Kadiolu et al., 2000, p. 24; Fraga et al., 2003, p. 7). However, in these first years the targets were more

official inflation projects than targets. After some years, in 1994, the central bank announced real targets for

which they were accountable (Mishkin, 2004, p. 16). Initially the country used wide target ranges, as the

inflation at the moment of adoption amounted to 27.31 % (Fraga et al., 2003, p. 7). Choosing a too narrow

range would cause frequent target misses, as it is more difficult to control inflation, the higher the inflation rate

is. From the announcement of hard targets onwards (and the abandoning of official inflation projects), the

central bank used point targets (Mishkin, 2000, pp. 107-108). From 1999 onwards, the Chilean central bank

announced multi-year targets (Kadiolu et al., 2000, p. 42). However, Chile did not want the exchange rate to

float because of the negative consequences of a currency depreciation, as explained in paragraph 4.3.

Therefore, Chile used an exchange rate band around a crawling peg. In 2000, the Chilean central bank finallyadopted a full-fledged inflation targeting regime, with a floating exchange rate (Mishkin, 2004, p. 17).

In the beginning, Chile had some difficulties that can be placed within the rules-discretion framework. Chile

realized that the more they applied strict monetary policy rules to build credibility, the less flexible their

monetary policy was. This trade-off was elegantly solved. In the beginning, Chile faced high inflation. Thus,

Chile realized that it had first to build credibility and therefore they used very strict monetary policy rules. Later

on, when disinflation had been realized, the country chose to operate more flexibly, in such a way that it could


22/27


23/27

23 | P a g e

central bank and set out price stability as the primary goal. Moreover, the central bank became accountable for

missing the target and the central bank needed to communicate towards the public on the monetary policy

performed by the central bank. The target is set by the government and the central bank. On top, this Central

Bank Law tried to restrict the fiscal dominance by prohibiting the central bank to grant advances and to give

credit to the Treasury. Furthermore, the Central Bank of the Republic of Turkey was not allowed to buy

instruments issued by the Treasury (Kara, 2006, p. 6).

Kara (2006, p. 11) states clearly that the implicit inflation targeting regime was successful as disinflation was

realized and a target was never missed. In the period in which the implicit inflation targeting regime was

adopted, namely from 2002 to 2005, the central bank developed better forecasting method, which makes it is

easier to control inflation. Secondly, until 2004, the central bank had no fixed timings for announcing its

monetary policy decisions. However, in 2005, the central bank started to announce its monetary policy on fixed

dates and published periodical reports (Kara, 2006, p. 10).

However, Ersel et al. (2007, p. 19) state that Turkey could already adopt formal inflation targeting in the

beginning of 2005. However, two problems impeded this evolution. The first reason was that the institute of

statistics published from 2002 onwards a new consumer price index series. This new system made it difficult

for the central bank to forecast the inflation. Missing the target was thus highly probable. On top, this new

system asked for communication towards the public as the public is unfamiliar with this new system. Thesecond reason to postpone the adoption of the formal inflation targeting regime was that the central bank

introduced a new currency in 2005 as the Central Bank of Turkey dropped 6 zeros from the lira.

From 2006 onwards, Turkey switched to full-fledged inflation targeting as a new stand-by agreement with the

IMF was signed for the period from 2005 to 2007. This agreements goal was to restructure the social security

and financial services and to realize a tax reform. It was expected that based on these 3 reforms, a positive

trend would be recorded in multiple macro-economic variables and would make it easier to control the

inflation (Ersel et al., 2007, pp. 19-23).

Instead of announcing the target one year in advance, the central bank announces a 3-year target and

publishes an Inflation Report. If the target is missed, the central bank needs to justify the reasons for this

deviation and needs to formulate which actions it is going to take (Kara, 2006, pp. 13-14). The point target for2011 amounts to 5.5 % with a tolerance band of +/- 2 %. Inflation is measured by headline CPI. The target is a

multiyear target and more precisely a 3-year target. Thus, the target horizon is 3 years.

In the second quarter of 2006, the first problem was encountered. Due to political problems in Turkey and the

instability on the international financial markets, a depreciation of the lira and a deterioration of the

creditworthiness of Turkey were noted. In addition, the increase in petroleum prices increased the inflation and

inflation expectations in Turkey. This caused the central bank to raise its target (Ersel et al., 2007, p. 22). This

incident shows clearly that Turkey has been unable to reach its target and even needed to adjust its target,

which decreases the credibility of the central bank in 2006. Even in 2007, 2008 and 2009 the central bank

needed to adjust its target during the year (Erdem et al., 2011, p. 6). The governor of the central bank Durmus

Yilmaz, elected in April 2006, has be

voorlopige versie

Documents

voorlopige versie