Recently, the governments of Egypt, Morocco and Tunisia announced that they are preparing to switch to IT. It is a monetary policy whose fundamental goal is to keep the inflation rate (suitably defined) within a certain publicly announced band around a target rate. The band allows the central bank some flexibility in responding to departures of the actual inflation rate from the target rate. Although the principle of IT is simple, the success of a central bank in implementing it requires certain preconditions. Are these preconditions fulfilled or could they be fulfilled soon enough in these three countries? This report is mainly about their readiness for IT. Indeed, some of the preconditions may be achievable gradually but over a relatively short transition period while some others are not so easy to fulfil. In particular, financial soundness is a major prerequisite, and it is a fact that, although these three countries implemented many important financial reforms, their financial systems remain in various ways rather vulnerable. Fiscal deficits, high public debt and/or severe rates of non performing loans (NPLs) held by the banking system are among the main indicators of this vulnerability.
The first fundamental precondition of IT is the independence, accountability and transparency of the central bank. The central bank must be independent in order to be able to adjust freely the instruments of its monetary policy and therefore to be credible. The concept of independence is often divided into goal and instrument independence (Debelle and Fischer, 1994). In the case of IT, instrument independence is the most crucial since the inflation target is by definition the single main goal of monetary policy. Independence means that the central bank should have enough power and strength to resist any external pressure to deviate from the chosen objective, it also means that the public believes that it is free to make decisions solely in accordance with the requirements of IT, and hence it is in particular free to decide whether or not and when to finance a government budget deficit. It follows that fiscal discipline is a pre-requisite for the credibility of the central bank’s independence.
Accountability and transparency of the central bank are corollaries of its credibility and independence. An independent central bank has to be accountable for its actions to the public concerning the successes and failures of its policy. The public must have the capacity, through its elected representatives, to control and ultimately punish incompetent policymakers in order to create better incentives for them to do their jobs well (Mishkin, 2002). Monetary policy decisions must also be communicated in a clear and regular manner to the public and particularly to financial markets and private economic decision-makers. Transparency should focus on the policy analysis and the operational objectives of the central bank in order to reduce the inflation bias and uncertainties in financial markets (Faust and Svensson, 2001). A transparent monetary policy means that policy decisions (usually changes in short-term interest rates) should not contradict the expectations of the market. Thereby, transparency lowers the cost of achieving the inflation target.
The efficiency of the central bank depends on the degree of development of the country’s financial markets and its stability. It has been shown that the most serious economic contractions arise when there is financial instability (Mishkin, 2002).
Banks that are badly managed, or are under stress for any reason, are likely to lobby the government for relief through any available channels and to undermine the functioning of monetary policy. Higher short term interest rates intended to reduce inflationary pressures are then likely to put fragile banks under more stress and to hurt them more severely. The US experience, in the 1980s with the crisis of the savings and loans associations, and again in the current year (2007-2008) with the sub-prime loan crisis, shows that if a central bank decided to fight inflation, badly managed financial institutions would be put under severe stress and some of them would fail leading to uncertainty and instability.
A fragile banking system may also be unable to deal with free capital mobility and exchange rate volatility since IT cannot succeed unless a flexible exchange rate regime is put in place and only one main target of monetary policy and a single nominal anchor are adopted. In the presence of free capital mobility, stabilisation of the exchange rate subordinates monetary policy to its requirements and may lead to intolerable deviations from the inflation target, which will destroy the credibility of the central bank’s commitment to IT. Nevertheless, According to Mishkin (2002, 2004), the exchange rate flexibility condition may be weakened, and a trade-off between the inflation and exchange rate objectives may be considered within an adapted IT framework.
The existence of a well developed financial system also depends on the soundness of fiscal policy. Fiscal dominance may threaten financial stability since it means that high pressures are put on the financial system and not only on the central bank to finance the government budget deficit. Fiscal discipline is a main pillar for the credibility of the central bank’s independence and the viability of the country’s financial institutions, as it signifies that the central bank and the banks will not be forced eventually to finance government budget deficits by printing money as happened, for example, in Turkey in the 1990s and many other countries before and after.
However central bank independence and fiscal discipline do not imply that monetary and fiscal policies should be totally disconnected. In fact, the lack of harmony with regard to the relationship between the CB and the Ministry of Finance might constitute a potential difficulty. Different views between the monetary and fiscal authorities can in fact send wrong signals to the financial markets as well as to the public. These considerations need to be taken into account, not only in the design of an IT strategy, but also in any monetary strategy.
The report is organized in two parts. Part one is about preparing for IT while part two is about Monetary transmission mechanisms.
I. Preparing for IT
The first part of the report is about country cases. It includes three separate papers on Egypt, Morocco and Tunisia where the authors try to assess to what extent these countries are prepared and how well they are preparing for the transition to a formal inflation targeting regime (IT). These countries can learn from other experiences, in particular the Turkish, which is presented in the fourth paper. The titles of the papers are the following:
1. Readiness to Inflation Targeting: the Case of Egypt
2. Monetary Policy and the preparation for Inflation Targeting in Morocco
3. Readiness of Tunisia for Inflation Targeting
4. Formal Inflation Targeting Experience of Turkey: 2006-2007
In Egypt, the objective of the central bank was set within the context of a general program to move to inflation targeting (IT) once the fundamental machinery needed for its implementation is installed (CBE, 2005). During the transition period, the monetary authority intends to meet its inflation stabilization objective through the management of the short-term interest rates and the control of other factors and shocks that might affect the inflation rate (CBE, 2005). Many reforms and a lot of actions have been taking by the Egyptian authorities to prepare for IT. However, Central bank independence in Egypt has been rather hindered by fiscal dominance and the pressure due to high budget deficits and public debt. Although serious steps have been taken recently under a new law committing the country to the achievement of price stability as the primary objective, prohibiting credit extension to the government, and strengthening central bank independence, Egypt does not yet satisfy some essential IT requirements, including the central banks independence requirement. The existence of a rather fixed exchange rate regime currently constitutes another important obstacle.
The Egyptian monetary policy had been for along time and until recently without a clear anchor for inflation expectations. Even in 2008, the monetary policy strategy carried out by the Egyptian monetary authority is still far from being ccoherent and the CBR does not seem to be the main guide for public’s inflation expectations. Inflation reached a record level of 23 percent in August 2008, a level which has not been reached for at least the last two decades. Despite the reforms undertaken especially since 2003, a lot remains to be done in order to build up the central bank’s credibility.
The level of public debt and budget deficit remain too high. As long as they remain so high the country should not engage in IT. The ongoing rise in the public debt, which inevitably increases the domestic interest rates and may lead to a slower economic activity, could force the central bank to lower the real interest rate and hence to raise the level of inflation.
The absence of a clear cohesive policy framework and the weakness in terms of CB independence reflect a lack of political willingness. Is the Egyptian government really inclined to carry out a policy such as IT? The recent small increases of the interest rate decided by the CBE (during June, July and August 2008) in order to combat inflation seem to indicate that CBE is still reluctant about immediate policy choices and perhaps more about strategic choices and namely switching to IT.
The situation is somehow similar in Tunisia. Recently, the government of Tunisia announced that it is also preparing to move to IT after an indeterminate transition period, and to move to full liberalisation of capital mobility and full convertibility of the Tunisian Dinar (TND).
Although many financial reforms have been undertaken in Tunisia, many of the preconditions for IT are not yet fulfilled. In particular, the central bank of Tunisia (BCT) lacks independence, and the country lacks a strong and transparent financial system. The banks, which dominate this system, remain fragile, and the severity of their non-performing loans (NPLs) problem undermines their effectiveness in performing their role as the most important channel of financial intermediation and transmission of monetary policy signals. The current situation suggests that Tunisia must undertake major reforms before it adopts a full fledged IT policy. There is serious doubt about the transparency, accountability and independence of its central bank (BCT). The lack of a clear performance indicator, the lack of independence from the executive branch of the government and the composition of its board compromise its independence and its credibility. The governor of the BCT is appointed by government decree and is responsible for the management of all the affairs of the bank. He is assisted by a board whose members are appointed by the government. Many reforms have been implemented and laws passed to change and modernize the Tunisian financial system over the last decades. For instance, prior to 1995, the BCT and the ministry of finance in Tunisia controlled on behalf of the government almost every aspect of bank credit and management, and the banks were required to allocate a large portion of their wealth to buy government debt and to finance government projects. This has changed quite radically In July 2001 when the Tunisian government introduced a new banking law, (Law 2001-65) in order to strengthen bank supervision and to help banks deal with the new economic and regulatory environment. However, not all Basle II principles are yet satisfied and, based on the statistics on non-performing loans (NPL), both for private and public banks have high NPL rates. Not surprisingly, public banks have a higher ratio (19.7% in 2006) than private banks (18.7% in 2006). Obviously, the central bank is not using all its legal powers and instruments to rigorously implement the existing regulation and avoid such a low performance.
Targeting broad money growth, in addition to pursuing a highly managed exchange rate regime, has long been the core of the BCT’s monetary policy. Since the early 1990s Tunisia has followed a constant real effective exchange rate rule in an effort to index its nominal exchange rate to the domestic price level in order to protect the competitiveness of Tunisian producers.
It is remarkable that the procedure for competitive tenders actually followed since the mid 1990s has generally been such that the BCT maintains the previously applied interest rate and asks banks to specify only their demand for liquidity. This amounts to a de facto control of short term interest rates. Given that banks are quite dependent on its resources, the BCT usually satisfies their liquidity requests in order to keep interest rates stable. BCT’s policy seems to be driven also by the fear of financial system instability and by the government’s economic development policy and its fiscal deficit than by the rules of its monetary policy. As a result of various contingencies and the government need to meet the financial exigencies of its liabilities, the BCT had to meet requests for liquidity put forward indirectly by the government through the commercial banks, and hence to deviate from its money target. Banks had to supply credits to meet government needs and other demands backed by the government authorities. BCT does not supply funds directly to the government.
It may be argued that the fiscal reforms implemented during the 1990s established new rules and institutions that have led to more fiscal discipline and contributed to the lowering of the inflation rate. The government has also adopted relatively more transparent and more market-based debt management instruments. The government is now required to issue and sell new bonds on the open market to finance its deficit. These reforms led to the development of the government debt market and improved the incentive for the government to master its deficit and to practice fiscal discipline. However, these improvements do not mean that monetary policy had been completely freed from fiscal dominance.
Morocco and its central Bank, Bank El-Maghrib (BAM) seem be the candidate most fit for IT in the region and the most likely to switch to this regime in the near future. BAM has since 2005 made significant efforts and significant progress towards meeting the pre-requirement of IT. It has radically changed its chart, developed its technical and forecasting capacity regarding the key macroeconomic variables. It started producing and publishing inflation forecasts. The fiscal deficit as well as the public debt observed in Morocco have been under control since 2006 consistently with the requirements of IT. The BAM monetary policies as well as the changes in monetary instruments are announced through regular public information services (annual report, BAM website…). The BAM issues statements to the public on the progress towards meeting monetary policy objectives. The BAM has enhanced significantly its transparency and communication strategy. It is transparent about its objectives (thanks to the 2005 new statutes) as well as about its technical procedures (the model) it utilizes to reach these goals. BAM started establishing systematic and relatively accurate forecasts of inflation and a estimate of how inflation is likely to be affected by changes in the monetary policy instrument. Yet, before 2005, its financial system and policies and its fiscal and inflation indicators had not been less fragile than Tunisia’s. This means that progress is possible.
The Turkish experience is even more meaningful with this respect. It shows that it is possible to move from a highly unstable macroeconomic situation, characterized by high fiscal deficit and public debt and high inflation, to a much more stable one and to fulfill most of the major IT conditions in a reasonable transition period. Nevertheless, Turkey is still facing major challenges and its experience with IT, short for the moment, is to be observed and examined more closely.
II. MONETARY TRANSMISSION MECHANISMS AND THE EFFECTIVENESS of THE MONETARY POLICY in Egypt, Morocco and Tunisia
Understanding and measuring the time lag between a monetary policy decision and its impact on the main nominal and real macroeconomic variables is what the analysis of MTMs is all about. How does a change in the policy interest rate decided by the central bank affect key macroeconomic variables, including the exchange rate, the volume of loans and above all the inflation and output? The purpose is not only to identify these impacts but also to measure their intensity. This is of course essential for successfully conducting monetary policy in general, and more so in the case of inflation targeting. With this respect, a lot remains to be done in the region; at this stage, we still wonder which channels and mechanisms are really operating and which channel is the most important? And then, what makes this channel more prevailing?
Two main approaches are possible for identifying and analyzing MTMs. The first approach adopted by Boughzala relies on micro data mainly on banks and when possible on firms and other investors, and adopts a panel data set up. The second approach, adopted by Boughrara, relies on macro time series data and applies various types of VAR models; the latter is the most widely used.
The first paper, entitled “MTM and the Imperfection of the Banking System” has a wider geographical coverage but its field of investigation is more limited. It tries only to explore how and to what extent changes in the short-term nominal interest rate decided by a central bank leads to changes in the banks lending interest rate and consequently in the volume of loans they effectively provide to investors (firms and households). The second paper, entitled “Monetary Transmission Mechanism Analysis in Egypt, Morocco and Tunisia” is more comprehensive.
The fist paper argues that monetary policy effectiveness depends on which monetary transmission mechanisms (MTM) are operational and on how efficient they are and that MTMs vary across countries and are not equally efficient across countries, in particular in the European Union (EU) countries compared to MENA countries.
MTMs operate mainly through the banking system, be it for the interest rate channel or the credit channel. What makes these channels more efficient in some countries, and what to do in order to improve their effectiveness through the improvement of the performance of the banking system? How fast do banks reflect changes in the interest rate decided by the central bank (CB) in their lending activities and/or their own rates (mainly their loan rates)? And what determines the speed of their response? This is the main concern of the first paper which restricts its interest to cases where the monetary instrument used by the CB is the interest rate and where the monetary policy objective is to reach an inflation target and/or a real output and employment target (assuming that output and employment are strongly linked) and relies on the traditional interest channels and the credit channel. Hence, by exploring the reaction of the banking system to the central bank policy actions the paper indirectly measures the impact of monetary policy on inflation and real economic variables. One of the main objective of this work is to test to what extent the banks’ response depends on the banking market structure and the type of competition (imperfection), on the type of risks banks deal with. Banks’ response depends on the structure and quality of their assets (that is the quality of their loans and how much and which type of securities they hold…) and on their liquidity (Ehrmann M., Gambacorta L., Martinez-Pagés J., Sevestre P., Worms A., 2001). Indeed, banks that can obtain liquidity at the inter-bank market or by easily selling assets on the financial market would be less dependent on the CB financing and might respond more slowly to monetary policy shocks, especially in the case of an interest rate increase, and, so they would continue to supply loans. Through this investigation, the aim is also to compare countries from the MENA region (Egypt, Morocco, Jordan, Lebanon, and Tunisia) to some OECD countries (France, Germany and the UK) whose financial market and banking systems are highly developed. Most of the selected MENA countries are undergoing important financial and monetary reforms and envisaging IT.
A specific model is constructed, focusing on the banking structure and assuming imperfect competition in the banking market. Different specifications were tested using a rich panel data. The simplest is a regression of loans on the central bank rate (CB rate) and the loan rate, the volume of deposits, the bank’s total assets, non interest income for the bank, which reflects its non lending activities in the financial market, and a country dummy, which lumps together EU countries opposed to the MENA (non EU) countries.
Given the endogeneity of the loan rate a one period lagged rate is introduced instead, but this variable does not seem significant after all! The country dummy systematically has a positive and significant sign showing the superior performance of the EU countries. The other signs are as also as expected and explained above. When a two stage estimation method or a GMM method is used, basically the same results are confirmed.
The second paper, entitled “MTM in Egypt, morocco and Tunisia”, tries to investigate the four major conventional transmission channels, namely the exchange rate channel, the interest rate channel, the lending channel, and the stock market channel. Far from describing all the transmission mechanisms, these channels are more likely to be operating in the panel countries considered. To what extent these transmission channels (e.g. interest rates, exchange rates, credit aggregates, stock market indexes…) matter will is first tested using Granger causality tests. A more formal analysis will be based on SVAR models.
The results reported on Table 1 clearly corroborate the preliminary conclusion regarding the degree of the interest rate pass-through in all countries. Overall, they indicate that the pass-through is incomplete to all retail rates (deposit and lending rates). It is worth noting however that the magnitudes of pass-through coefficients associated to deposit rates are weaker than those associated to the credit rates. For instance, the long-run pass-through effect in case of Egypt is around 0.31 for lending rate and 0.11 for deposit rate. This implies that 31 per cent of any money market rate change is passed through to the lending rate in the long –run, and only 11 per cent of the policy rate is passed through the deposit rate. The pass-through in Morocco is slightly superior to that in Egypt. Approximately 40 per cent of the change in the Moroccan policy rate is passed through the lending rate whereas 33 per cent is passed through deposit rate. Likewise, the pass-through degree in Tunisia is not significantly different from that recorded in Morocco.
Stickiness of retail rates in the case of Egypt, Morocco and Tunisia is due to the presence of adjustment costs associated with changing retail rates to customers and this may lead to the smoothing of retail interest rate changes with respect to changes in the policy rate (which could play the role of the marginal cost of funds). It is also caused by to the lack of competitiveness among banks. Indeed, the lesser the degree of competition, the higher this spread is likely to be. In the panel countries considered in this paper, the lack of competition is due, to a great extent, to regulation (case of Egypt), to collusion on the part of financial institutions or fixed costs of entering the market (Tunisia and Morocco), and to the ‘gentleman’s agreement’ (Tunisia). Besides, the monopolistic or oligopolistic structure of countries’ banking systems explains the small pass-through degree because in such banking, retail rates won’t move significantly and quickly in response to changes in money market rates. Cette situation reflète aussi une faible crédibilité du système financier dans ces pays.
Finally, the finding of a weak interest rate pass-through could be explained in Egypt, Morocco and Tunisia by the fact that the monetary institution lacks credibility; in such case, economic agents may misperceive temporary policy rate shocks as being permanent. Overall empirical results on long-run interest rate pass-through seem to indicate that the central banks policy rates do not seem to rapidly determine the market lending and deposits rates. Yet, we would expect interest rate pass-through to have improved in recent years given the improvement in the specific features of the monetary framework, banking sector and economic conditions.
The next and main step in analyzing MTMs is to examine the impact of policy measures on the key economic variables, primarily inflation and the activity level. A natural extension of the analysis of interest rate pass-through lies in analyzing the complete interest rate channel of monetary policy, from policy rate changes to movements in the consumer price index and changes in the level of economic activity.
In line with the above explanation, we start the empirical analysis of the MTM in Egypt by considering the baseline model which include in the VAR the oil price and the US federal funds rate in order to avoid well-known empirical anomalies such as price puzzles. All the variables are considered in level and in log except for rates. All the date used are from IMF International financial statistics CD ROM (July 2008). Dates are monthly and cover the period 1997:01-2007:12. The other data we used in the case of Egypt are a measure of real activity, proxied by GDP deflated by the whole sale price index (RGDP); prices proxied by the whole sale prices (WPI); a nominal exchange rate, proxied by bilateral U.S. dollar (ER), and a stock market index proxied by the Cairo and Alexandra Stock exchange index (CASE).
The response of prices to an increase in the monetary policy measure (the Treasury bill rate) is significant and quick- prices start to decrease after about one quarter or so; the monetary policy seems effective in bringing down prices. The strong deflationary impact of a tightening on the WPI is intuitive and consistent with earlier findings (Rabanal, 2005 and Al-Mashat and Billmeier, 2007). As for the response of real activity, it stands out from the IRFs pattern that the policy rate fails to impact on real activity. While the reaction of the real activity to an exchange rate hike is not significant, it appears to be consistent with prior expectations especially after one year or so; the activity start to decline just after one year. Overall, these findings are consistent with the theory and free from any puzzle.
To shed light on the existence of the lending channel in Egypt, we include into the baseline model loans quantity (Loans) and loans price (proxied by lending-rate). The ordering of the variables is chosen on the basis of the speed with which the variables respond to shocks like in the case of the channel considered above. The focus here is on the identification puzzle, that is on the supply-versus-demand of loans puzzle. A contraction of bank loans for itself is not necessarily a consequence of leftward shift of the supply schedule. Thus, testing the supply view requires identifying the supply and demand schedule of bank loans. If the lending channel of monetary policy is dominant, a leftward shift of the supply schedule must be clearly observed following a monetary tightening. For the study of the lending channel to be meaningful, it might be also worthwhile to test the effectiveness of monetary policy. Thus, the lending view will be accepted if and only if:
H1: the quantity of bank loans LQ does not increase
H2: the price of bank loans LP rises, and
H3: real output Y decreases following a monetary tightening.
For the purpose of testing H1 to H3 statistically, the VAR is estimated to stimulate impacts of monetary policy on the economy. Simulating the dynamic responses of macroeconomic variables to monetary policy shocks is equivalent to calculating the impulse response functions (IRFs) of those variables to an innovation to the measure of the stance of the monetary policy.
It stands out from the empirical results that the existence of the lending channel in Egypt is very weak. While hypothesis H1 (loans do not decrease) can be supported by data, hypothesis H2 (loans price as measured by the Lending_Rate variable increases) cannot. In addition, the real activity does not seem to depress as suggested by the theory leading thereby to rejecting hypothesis H3. The response of real activity does seem to be amplified significantly when accounting for the lending channel. Hence, the lending channel in Egypt can hardly be supported by data. Bank loans do not seem to react subsequent to an interest rate hike. Some of the reasons that could explain such a finding are the high volume of non-performing loans, and the under-developed state of the financial market.
Nevertheless, the CBE has come a long way in developing its set of monetary instruments, but these instruments still lack effectiveness as the diverse channels of monetary transmission are not operating effectively and properly. Improving the performance of these channels will be paramount for a successful transition to a full-fledged inflation targeting monetary policy framework in Egypt. Uncertainty in the channels of interest rate policy may be caused by a strong and fast-working exchange rate channel. Indeed, the exchange rate channel continues to play an important role in the transmission of the monetary stance, as it magnifies the impact of policy shocks drastically.
The same methodology is applied to Morocco and Tunisia. The Moroccan monetary authorities have pegged the Dirham to a basket of currencies where the (nominal) exchange rate fluctuates within a narrow band. Under this exchange rate regime, the strength of the exchange channel is expected to be negligible if not absent. To verify this, to the basic VAR model was added the nominal effective exchange rate (NEER) as a potential transmission variable.
The VAR simulations related to the exchange rate channel (see figure x) indicate that a restrictive monetary policy (as measured by one standard deviation hike) is insufficient to impact significantly on the effective nominal exchange rate (see figures 1.a and 1.b). On the other hand, the depreciation of the currency seems to have significant effects on real activity in Morocco. The depreciation of the currency depresses real output. A depreciation has an asymmetric impact on prices. As to the reaction of output it does not seem to vary when including the NEER variable in the baseline model. This would indicate that the exchange rate channel is not operative. Overall, these findings corroborate to some extent the a priori belief about the strength of the exchange channel in Morocco. At best, the contribution of the exchange channel in propagating the impulses of the monetary policy in Morocco negligible. The absence of role of asset prices in transmitting monetary shocks is not surprising either given that share ownership is low, and that firms’ reliance on equity financing has not been so significant compared to bank credit. The role of asset prices in the propagating the effects of the monetary policy might increase in the future in line with the continued developments in capital markets.
In the case of Morocco, the monetary policy appears to be relatively more effective in impacting on real activity when compared with case of Egypt. Two channels are operative, namely the traditional interest rate channel and the bank lending channel. These two channels co-exit. However it remains to be seen which of them is the most dominant. The finding relative to the existence of the bank lending channel in Morocco does not surprise seeing that Moroccan system is more structured and healthy than that of Egypt or Tunisia and the level of NPLs is more manageable.
In Tunisia, the response of price loans depicts a clear tendency to increase subsequent to a tight monetary policy. This reaction is immediate and statistically significant at 5 per cent level. In other words, the response of price loan is such that ∂E(LPt+i|Λt)/∂u>0 for i=1,2,…,16 implying that H2 cannot be rejected. Furthermore, Rt bank loans IRFs show that loans quantity does not increase. Rather, it exhibits a clear downward tendency, which thought non statistically significant for the eight first quarters becomes significant at the latest: ∂E(LQt+i|Λt)/∂uR=0 for i=1,2,…,8 and ∂E(LQtt+i|Λt)/∂u<0 for i=9,10,…,16. Therefore, H3 cannot be rejected. Seeing that the hypotheses H1, H2 and H3 cannot be rejected, one may conclude that the lending channel is operative in Tunisia. When considered together the patterns of price loan and bank loans it appears that the supply schedule for bank loans shifts left following a monetary tightening. Thus, data do support the lending (supply) view in the case of Tunisia. It is worth noting that this finding does not imply the rejection of the money view. We can even state is that the lending channel is not only operative but also dominant. The empirical findings relative to the case of Tunisia are not really so different from those of Egypt and Morocco. Neither the exchange rate channel nor the asset price one seems to be operative in Tunisia. These findings are in line with previous studies.