1. 1 background A large body of facts associates’ financial sector development leads to boost the Ethiopian economic growth or vice versa, yet the channels through which inflation affects this relationship are not as much of systematically explored. The effect of inflation occurs through a wide variety of direct and indirect channels. Inflation increases transactions and information costs which directly inhibit economic development. For example, economic agents will find planning difficult when inflation makes nominal values uncertain.

Firms and individuals will be reluctant to enter contracts when inflation is imperfectly predicted and judgments about absolute and relative prices are uncertain. The reluctance to enter contracts over time will inhibit investment and entrepreneurship, which will affect resource allocation and economic growth. The extent that high inflation disrupts the smooth operation of a nation’s financial markets and institutions, it also discourages their integration with the rest of the world.

Since high inflation is often variable inflation as well, there will be considerable uncertainty about future prices, interest rates, and exchange rates, which in turn increases the costs of hedging financial risks among potential trade partners. If inflation also increases a currency’s vulnerability to speculative attack, hedging instruments will become even more expensive and difficult to price. All of this will discourage trade and inflows of foreign capital. (Bruno and Easterly, 1998) A few studies examine the inflation-finance-growth nexus. Haslag and Koo (1999) and Boyd et al. 2001) show that inflation is associated with financial repression. Rousseau and Wachtel (2002) identify an inflation threshold for the finance-growth relationship, finding that finance affects growth positively only when annual inflation can be held below a threshold that lies between 13 and 25 percent for the world, depending on the measure of financial depth that is chosen. They also find that disinflations are related to strong positive effects of finance on growth. 1. 2 Historical trend overview of inflation, economic growth and financial sector development In Ethiopia

In this study year 2006 is used as a reference period (i. e. December 2006=100). Similarly Central Statistics Agency (CSA) uses this year for CPI computation and we also follow the same year for other economic variables like for GDP, private credit, population, net export and capital formation. This is not without reasons it is after 2006 that many macro variables changes in different ways example inflation soared relative to the years before and inflation is one of the main indicator of macroeconomic stability of the economy.

And inflation is one of the main indicators of macroeconomic stability Therefore it is convincing to use this period as references period and adjust other variable by the same year and all are converted into index number. Figure 1. 1 below illustrate the fact that from the year 1993 – 2003 inflation is in the range of 5% and 10%, hence GDP and CPI goes together that is they are related positively this might be due to the fact at low level of income inflation positively affects growth or else according to Keynesians school inflation may relate positively if inflation emerges as a byproduct of increasing aggregate demand.

In this Keynesian framework, it is not the case that inflation is itself a positive engine of growth, certainly not a primary growth-inducing force. The point is rather that, if rising aggregate demand is leading to increased growth, then some inflationary pressures are likely to emerge in this scenario as a relatively benign byproduct (Robert Pollin et al 2005). And a structural break happens before the eve of Ethio-ertrian war in the year 1997 and 1998 due to good harvest because of good rain or supply shocks.

Government has no policy at that time to avoid price fall such as price flour credit etc to farmers. farmers have to sell their product and they cannot postpone because there were no credit availability to farmers to borrow and sell their product when price are good, thus price of agricultural items experience a major freefall . But after year 2006 inflation goes above 10% and GDP relatively grow stable . Inflation is very variable and it seems it is difficult to tell their relation with GDP if we cannot say they do not have any relation.

Figure [ 1 ]. 1 trend of annual gross domestic product versus consumer price index Even though it is difficult to tell the relationship between private credit and inflation from 1993 -2004 when inflation remains below 10% but their relation seems clear after inflation goes beyond 10% and they correlated positively. Therefore it is difficult to tell the overall relation between these variables over the whole sample period. (Fig ure1. 2) Figure1. [ 2 ] trend of annual credit to private sector vs annual average CPI . 3. Statement of the problem The investigation into the effects of inflation on the Ethiopian economy and the financial sector is an important topic. Achievement of stable inflation rate is among the fundamental objective of national bank of Ethiopia. Similarly, Central Banks worldwide have placed increasing emphasis on price stability as a macroeconomic policy strategy directed towards the achievement of higher levels of economic growth through financial stability.

Banks and financial activity within an economy contribute to the operation and development of the economy by mobilizing and pooling savings; providing payment services that facilitate the exchange of goods and services; producing and processing information about investors and investment projects in order to enable efficient allocation of funds; monitoring investments and exerting corporate governance after these funds are allocated; and helping to diversify, transform and manage risk (Demirguc-Kunt, 2008). Why the recent interest in inflation and finance? Two empirical findings are reported to be at the origin of this interest.

At first, a negative association between inflation and economic growth was widely developed (Barro, 1995). Then, a positive relationship between the development of the financial system and economic growth took place in the economic literature (King and Levine, 1993a,b; Pagano, 1993; Levine and Zervos, 1996). These two strands of the empirical literature, that is finance–growth and inflation–growth relationships, respectively, have lived separate lives but one obvious link is that inflation might be affecting economic growth through action on the financial sector or vice versa.

It is obvious from the above discussion that the achievement of stable inflation not only the target of national bank of Ethiopia but also almost the whole central bank in the world but the issue of inflation affect the Ethiopian economy and financial sector is still remained unresolved. No attempt has been made to investigate the effect of inflation on finance and the Ethiopian economy as far as the authors are concerned. Thus this paper provides important policy instrument useful for policy maker. 1. 4. Objective of the study

The general objective of the paper is to investigate the effect of inflation on the Ethiopian economy and the financial sector. The specific objectives of this study are: * To investigate the effect of inflation on the Ethiopian economy * To investigate the effect of inflation on the financial sector in Ethiopia * To obtain inflation threshold at which inflation is least detrimental to economic growth via financial intermediary channel. 1. 5. Hypothesis of the study The paper expects inflation to have a negative effect on the Ethiopian economy and financial sector.

This negative effect can be explained by the following hypotheses. * Higher rates of inflation are associated with a reduction of financial sector development and the Ethiopian economy. 1. 6. Scope of the study This study explores the impact of inflation on the Ethiopian economy and the financial sector in Ethiopia using annual economic data from1992-2011 and covers econometric analysis. The data sources are mainly National Bank of Ethiopia data bases. 1. 7. Significance of the study

The importance of investigation of the effect of inflation on the Ethiopian economy and the financial sector in Ethiopia lies in the fact that it can be useful in identifying inflation threshold which policy makers need to control in order to obtain the desired values of stable financial sector and growth in Ethiopian economy. It might also be helpful in developing the econometric models and designing policies 1. 8. Organization of the study This study is organized into five sections. The first being introduction, the literature is reviewed in the second section.

The third section is about methodology. Empirical result is presented in the fourth section, and the fifth section conclusion and recommendation is delivered based on the findings obtained. II. INFLATION AND THE FINANCE–GROWTH NEXUS: theory and evidence 2. 1 Theoretical literature There are many reasons why the depth of financial sector development can promote economic growth. To summarize, more intense use of financial intermediaries and increased amounts of intermediation will encourage savings and investment and improve the allocation of savings to investment projects.

This in turn encourages a higher level of capital formation and greater efficiency in the allocation of capital. However inflation will inhibit the development of the financial sector and result in financial repression. High inflation will also discourage any long term financial contracting and financial intermediaries will tend to maintain very liquid portfolios. Thus, in an inflationary environment intermediaries will be less eager to provide long-term financing for capital formation and growth; both lenders and borrowers will also be less willing to enter long-term nominal contracts.

High inflation is often associated with various forms of financial repression as governments take actions to protect certain sectors of the economy. For example, interest rate ceilings and directed credit allocations are common in high inflation environments. Such controls lead to inefficient allocations of capital that inhibit growth. The relationship between financial repression and inflation can also be bi-directional. In some instances, repression is a crude effort to protect certain sectors from inflation.

In other instances, financial repression that is introduced to assist the government in financing its own activities is a cause of both inflation and resource misallocation. Moreover, inflation will have contemporaneous effects on the finance ratios that are used to measure financial sector development. High inflation will increase the opportunity costs of holding money and lead agents to economize on money holdings. Thus, the ratio of money to GDP might decline as a direct consequence of inflation. Further, the ratios of financial assets to GDP might decline in a high inflation environment if nominal debts do not increase as rapidly as GDP.

This is particularly likely if the financial repression that is common in high inflation episodes keeps real interest rates low or even negative. . (P. L. Rousseau, P. wachtel, 2001), 2. 2 Empirical evidence Initially, scholars find out the threshold points by personal judgments, instead of statistical methods Fischer(1993 ) the pioneer of the issue, sets 15% and 40% as two check points to examine if high inflation harms growth more than low inflation does. He founds that inflation slump economic growth through the channel of decreasing investment accumulation and growth rate of productivity. esides, Bruno and easterly(1998) defines a crisis as its inflation rate in excess of 40% per year for continuous years to study growth rates report during and after high inflation crisis. They confirm that economy dramatically drops down during high inflation crisis. But rapidly revives after the crisis, even surpass the level before the crisis. However the check point method provides little information for policy making. Thus, latter researcher found accurate information threshold via various econometric techniques.

For systematizing these studies, we divide them into three categories by different approach to threshold estimation except for arbitrary specification just mention. The first is spline function method. The method assumes the regression has identical error variance within the range of inflation. Then, cut the range into many pieces by tiny points lying between the ceiling and the flooring of some range and work out some point which minimizes the sum of the squared errors (SSE) of regression, or equivalently maximizes the R2 that is the threshold point.

Sarel (1996) uses the way and finds the structural break, 8% by estimation, significantly exists between inflation and growth. Then, Ghosh and Philips (1998) follow similar way for cross country data and the kinked point is 2. 5%. also, Christofferson and Doyle (2000) estimate the threshold for 22 post-USSR countries and the critical point is found to be 13%. Latter Bolton and Alexander (2001) determines the inflation threshold across countries, by the same estimation method, is 3% . rather than unique threshold point, Burdekin et. l (2001) suppose there are multiple turning points in their framework. In their finds, there are two threshold points,8% and 25% for the group of industrial countries and two threshold points,3%,5% significant for the group of developing countries. In addition to spline function estimation, some researcher adopt boot strapping method proposed by Hansen(1999) for inflation threshold estimation and use likelihood ratio test to examine significance of the threshold effect. As to spline function method used by Sarel (1996), t-test is used to examine significance of the threshold point effect.

Still the purposes of these two techniques are similar that is, to find a threshold point which minimizes the sum of squared error or maximizes the R2 of a regression. Khan and Senhadji (2001), for example, find the inflation beyond the threshold in values does significant damage to economic growth, the threshold interval for world is 9-10%,industrial countries 1-3% and developing countries 11-12%. Finally, there are some papers which use quadratic function to work out inflation threshold effect.

Gylfason and Herbertsson (2001) at the beginning use interactive process of estimation to find out a threshold interval 10-20% then, re examine the outcome with one threshold growth quadratic estimator and confirm threshold around 11%. Beside, Pollin and Zhu (2006) divide all observation into four groups, that is, whole countries, OCED ones, middle income ones, low income ones. As well, the quadratic function, similar to Gylfason and Herbertsson (2001) in employed to figure out inflation threshold for each group; that is, 15-18%, insignificant, 14-16%, and 15-23% respectively.

In their conclusion, below 15-18%, inflation benefits economic growth moderately for the whole countries. On the other hand, a few scholars apply threshold estimation to case study for specific country, rather than cross sectional or pooled data, all most all of use time series ones Singh and Kalirajan (2003) use yearly data of India for 1971 to 1998 for threshold investigation. Abiding by spline method, the author find the inflation threshold is around 5% for India. Besides, Sweidan (2004) employs autoregressive conditional hetroscedasticity mode (ARCH) covering annual ata of Jordan from 1970 to 2000 and, according to spline function techniques, estimate the threshold point 2% for the country. To cite some of specific studies for developing countries, Mubarik (2005) for Pakistan founds 9% inflation threshold, Narryan et al, (2009) for china between 3. 9 to 6. 5% Seleteng, M (2004) for Leseto founds 10% inflation threshold. If the effect of financial development, on growth structurally changes under different degree of inflation researchers have found different threshold level. For china, case study, Kong (2007) establishes a time series model with yearly data covering the year 1952-2004.

The research makes use of bootstrapping method to get two structural breaks, i. e. , 3. 9% and 6. 5%. what is interesting about the finding, is financial development significantly promotes growth as inflation beyond the two threshold points but has a little effect on growth as inflation below the two points, which is dissimilar to other finding of the inflation threshold effect. Phiri. A, (2010) founds inflation level of 8% for South Africa above and below this level real activity losses gradually begin to be magnified the further one moves from the threshold through financial intermediary channel.

In the literature of inflation threshold on growth, there is one feature we get, many of the study is a cross country, and there is no universal agreement about what the threshold inflation is. Such implies no country knows what threshold result of various papers is fit for her so it is important to find country specific threshold for policy making purpose and this paper tries to establish unique threshold point for Ethiopia through which inflation affects economic growth via financial intermediary channel. III. Methodology of the study 3. 1. Model specifications

The empirical analysis presented in this paper draws the theoretical underpinnings from the works of Huybens and Smith (1999). In the theoretical model, the transmission mechanism for the nonlinear operation of inflation on economic growth is induced through the financial markets and capital accumulation. Financial market activity is identified through banking lending activity (Cr) whereby both channels influence real economic activity. When levels of capital accumulation (K) are high bank lending activity are highly correlated through internal projected finance and this xerts a significant influence on economic activity whilst at low levels of capital accumulation, little or no financial activity transpires and the relevance of bank lending activity decreases. Inflation (? ) within the model is induced through increases in money supply which lowers the rate of return on both money and assets. Under low inflation regimes, inflation does not distort the flow of information or interfere with resource allocation and economic activity up to a certain threshold. Beyond such a level, an increase in the inflation rate aggravates credit market frictions through a distorted flow of information.

Consequentially, this causes less efficient resource allocation in the banking sector and resulting in lower capital accumulation and long-run activity. Hence, a negative relationship between inflation and real activity becomes more pronounced at higher levels of inflation above the threshold. Denoting Y as output productivity, real activity within the economy is captured in the following production technology: Yt=µ+? ? t+ ? ’sXt+? t t=1………. T……………………………………………. (1) Yt is the dependent variable that indicates real GDP for some period t. ? t is the appropriate measure of inflation.

Xt is the vector including the set of control variables. Finally, ? t is the error term. Expanding the above equation: Yt= µ + ? 1 ? t + ? 1k + ? 2cr + ? t…………………………………………………………………………….. (2) Where; K is capital formation to GDP ratio Cr is total private credit to GDP ratio in the country Based on the presented theoretical model, positive coefficients are expected to be associated with ? 1, and ? 2. The sign on the coefficient ? 1 is, however, ambiguous as inflation is a threshold variable depicted to having two opposing effects on real activity below and above a certain threshold.

The sign on the coefficient ? 1 is thus subject to scrutiny in the empirical analysis. It has become standardized practice in the literature to econometrically quantify inflation thresholds in the inflation growth nexus by making use of Sarel’s (1996) threshold econometric specification. This econometric model is informed by the theoretical inflation threshold growth model and can be best thought of as a reliable representation of the theories predictions (Barnes and Duquette, 2006).

The econometric assumption underlying inflation threshold models is that observed data of inflation, growth and other growth determinant variables can be segregated into two regimes; one regime capturing the dynamics of the data below an established inflation threshold level whilst the second regime analyzes the effects of the data above the threshold. In its base form, the Sarel’s (1996) model framework assumes the following nonlinear inflation growth regression function: Yt=µ+? 1 ? t + ? 2 [? – ? *] (D) + ? ’sXt + ? t …………………………………………… (3)

Where Xt is the vector of control (explanatory) growth variables derived from the theoretical model which are included in the threshold regression; ? * is the inflation threshold; D is a dummy variable taking the value of 0 when ??? * and 1 when ? >? *; and ? t is the iid error term. Deriving from the theoretical foundations set forth in the model of Huybens and Smith (1999) as depicted in equations (1) and (2), the study employs measures of banking lending activity, volume of trading in equity markets and investment as explanatory growth variables.

In addition, the studies of Hodge (2005) and Weeks (1999) have emphasized the importance of accounting for measures of openness in empirical inflation-growth investigations for South Africa. The study hence advocates the use of the real effective exchange rate (REER) as a measure of openness as this variable simultaneously captures both the financial institutional arrangement and openness of the economy. Rousseau and Wachtel (2002) note that exchange rate measures are deemed as one of the most crucial financial institutional arrangements in the context of stimulating economic growth in an open economy.

The expected coefficient on the exchange rate is thus expected to be positive. Including real exchange rate in the equation the empirically testable model becomes: Yt= µ + ? 1 ? t + ? 2 [? – ? *] (D) +? 1K + ? 2Cr + ? 3REER + ? ts…… (4) Where; REER is real effective exchange rate ? * is threshold inflation rate Since the inflation threshold, ? *, is unknown, equation (4) is sequentially estimated for different ascending values of ? * i. e. {? *min… ? *max} to produce a series of regression estimates.

From the series of threshold regression estimates, the optimal inflation threshold value is associated with the regression estimate which produces the highest regression explanatory power through the minimization of the sum-of-squared residuals (SSR). The t-statistic of the coefficient ? 1and ? 2 tests whether or not the structural break effect is significantly valid (Sarel, 1996). In re-arranging equation(4), Yt= µ + (? 1+ ? 2)[CPIt – CPI*](D) + ? 1K + ? 2Cr + ? 3REER + ? t and the sum of the coefficients ? 1 and ? is proved to effectively measure the effect of inflation on economic growth in the long run as a result of a unit increase experienced in the inflation rate. Latter in the estimation stage two control variables are included in the estimation they are population and net export. The use of the included variable is twofold one it makes the growth model complete, Solow (1956) and Swan (1956) who developed the first neo-classical models of growth, take the rate of growth of population as one of exogenous variables in their model to show that the faster the rate of population growth, the poorer the country.

And because the model is suffered from serial autocorrelation it helps to minimize this problem and the model can be rewritten as: Yt= µ + (? 1+ ? 2) [CPIt – CPIt*] (D) + ? 1K + ? 2Cr + ? 3REER + ? 4pop + ? 5hx + h? t …….. (5) Where: pop is population and Hx is net export the others are defined as before. 3. 2. Method of estimation and analysis The Augmented Dickey Fuller (ADF) tests were done to test for stationarity of all the variables. All variables were found to be integrated of order one I (1), meaning that they had to be differenced once in order to be rendered stationary (see computability of variables below).

Whereby the variables are computed as: Y =100*DLog Y Inf =100*DLog(P) Pop=100* DLog Pop K =100* DLog K Growth rates of other variables are computed using similar method. Whereas, the dummy variable is defined as: Dt = 1: 100*DlogPt > k =0: 100*DlogPt ? k The variables were further transformed into logarithm form due to the following advantages as suggested by Sarel (1996) and, Ghosh and Phillips (1998): * The log transformation provides the best fit. That is to say, the log transformation also, to some extent, smoothes time trend in the dataset. The log transformation can be justified by the fact that its implications are more plausible than those of a linear model. The dataset is further smoothed using Hodrick-Prescott filter. This is a smoothing method that is widely used among macroeconomists to obtain a smooth estimate of the long-term trend component of a series. Before estimating the model, Granger-Causality test is applied to measure the linear causation between inflation and economic growth. Mubarik (2005), M.

Seleteng (2004) The study employed OLS econometric analysis by using Eviews software to test the impact of inflation on Ethiopian economy and financial sector development. The descriptive approach is analyzed using figure whereas the econometrics approach is analyzed using unit root, diagnostic, and causality test. IV. ESTIMATION RESULT Table 4. 1: Pair wise Granger Causality Tests| Date: 03/12/12 Time: 14:21| Sample: 1992 2011| | Null Hypothesis:| Obs| F-Statistic| Probability| HCPI does not Granger Cause HGDP| 17| 9. 35220| 0. 00356| HGDP does ot Granger Cause HCPI| 28. 8289| 0. 000026 | Test statistics in Table 4. 1 show that the null hypothesis is rejected, which means that inflation rate Granger-Causes real GDP growth. The causality between the two variables is two-directional. The second null hypothesis of economic growth Granger-Causes inflation is also rejected, which implies that there is a two-way causality between economic growth and inflation. Granger-causality test also implies that there is a long-run relationship between the above-mentioned variables and hence the variables are co-integrated.

However, one has to be very careful in implementing the Granger-Causality test because it is very sensitive to the number of lags used in the analysis. Thus the Akaike Information Criterion (AIC) and Schwarz Information Criterion (SIC) were used to determine the lag length with the minimum preferred. The econometric analysis of inflation thresholds is conducted by use of OLS procedure. Sarel’s threshold model as depicted in equation (4) by improving in the way of Khan & Senhadji (2001) equation (5), is estimated using ordinary least squares (OLS) technique. The empirical results are reported in Tables 4. . The t-statistics for the regression coefficients estimates, as well as, their associated probability values and standard errors are also reported. Searching for an inflation threshold (? *) is conducted over the range of ? *min=1 and ? *max=15 resulting in the inspection of each employed estimation techniques. But results with even number of threshold are only depicted and the whole result is attached in the annex part. This study follows rolling regression method to get the threshold level of inflation. The method assumes the regression has identical error variance within the range of inflation.

Then, cut the range into many pieces by tiny points lying between the ceiling and the flooring of some range and work out some point which minimizes the sum of the squared errors (SSE) of regression, or equivalently maximizes the R2 that is the threshold point. The method has been applied by many authors includes Sarel (1996), Ghosh and Philips (1998) Christofferson and Doyle (2000) Khan & Senhadji (2001), etc. Table 4. 2 Estimation results of equation 5 at k=2 to 15 % (Dependent variable real GDP growth) K INFLATION TRESHOLD| variables| coefficients| t- statistics| prob. | RSS| 2%| CPI| K2| CR| K| CREER| POP| X| C| | 0. 84248| 0. 018714| 0. 248273| 0. 594876| -0. 00255| 1. 015855| -0. 1115| 5. 851522| | 5. 05771| 0. 372409| 1. 72774| 2. 352849| -1. 45423| 4. 430398| -3. 56856| 10. 73445| | 0. 0005| 0. 7174| 0. 1147| 0. 0404| 0. 1765| 0. 0013| 0. 0051| 0| | 0. 034557| 4%| CPI| K4| CR| K| CREER| POP| X| C| | 0. 081959| 0. 002572| 0. 270759| 0. 633814| -0. 00289| 0. 986449| -0. 11605| 5. 922592| | 5. 270088| 0. 061722| 2. 064964| 2. 738649| -1. 66923| 4. 511097| -4. 01969| 11. 41681| | 0. 0004| 0. 952| 0. 0658| 0. 0209| 0. 126| 0. 0011| 0. 0024| 0| | 0. 035023| 6%| HCPI| K6| CR| K| CREER| POP| X| C| | 0. 087517| -0. 05413| 0. 238348| 0. 74488| -0. 0034| 1. 064592| -0. 10592| 5. 753934| | 6. 056018| -1. 52315| 2. 002815| 2. 729159| -2. 61318| 5. 25192| -3. 94847| 12. 03106| | 0. 0001| 0. 1587| 0. 073| 0. 0212| 0. 0259| 0. 0004| 0. 0027| 0| | 0. 028439| 8%| CPI| K8| CR| K| CREER| POP| X| C| | 0. 092461| -0. 07171| 0. 206224| 0. 522302| -0. 00383| 1. 18786| -0. 09705| 5. 454026| | 6. 388461| -1. 87799| 1. 763171| 2. 532559| -2. 95534| 5. 510556| -3. 62264| 10. 71611| | 0. 0001| 0. 0898| 0. 1083| 0. 0297| 0. 0144| 0. 0003| 0. 0047| 0| | 0. 025901| 10%| CPI| K10| CR| K| CREER| POP| X| C| | 0. 092461| -0. 07171| 0. 206224| 0. 522302| -0. 00383| 1. 18786| 0. 09705| 5. 454026| | 6. 388461| -1. 87799| 1. 763171| 2. 532559| -2. 95534| 5. 510556| -3. 62264| 10. 71611| | 0. 0001| 0. 0898| 0. 1083| 0. 0297| 0. 0144| 0. 0003| 0. 0047| 0| | 0. 025901| 12%| CPI| K12| CR| K| CREER| POP| X| C| | 0. 093072| -0. 07624| 0. 200997| 0. 518713| -0. 00347| 1. 222295| -0. 09715| 5. 371779| | 6. 109029| -1. 67761| 1. 644632| 2. 420819| -2. 71349| 5. 136908| -3. 48575| 9. 567815| | 0. 0001| 0. 1244| 0. 1311| 0. 036| 0. 0218| 0. 0004| 0. 0059| 0| | 0. 027341| 14%| CPI| K14| CR| K| CREER| POP| X| C| | 0. 090729| -0. 09575| 0. 223813| 0. 560375| -0. 00367| 1. 230488| -0. 10182| 5. 344373| 6. 105485| -1. 6135| 1. 874365| 2. 674069| -2. 76187| 5. 009839| -3. 74769| 9. 176531| | 0. 0001| 0. 1377| 0. 0904| 0. 0233| 0. 0201| 0. 0005| 0. 0038| 0| | 0. 027799| 15%| HCPI| K15| HCR| K| CREER| POP| X| C| | 0. 07636| -0. 15079| 0. 334152| 0. 759266| -0. 00277| 0. 969406| -0. 1359| 6. 051532| | 5. 621396| -1. 91011| 2. 88583| 3. 673559| -2. 29902| 5. 190931| -5. 08056| 13. 51425| | 0. 0002| 0. 0852| 0. 0162| 0. 0043| 0. 0443| 0. 0004| 0. 0005| 0| | 0. 02567| Analysis of Inflation threshold Inflation threshold level between 8% and 10% is statistically significant and at this level of inflation the RSS is minimized 0. 25901 and the net effect of inflation is positive as measured by the sum of coefficients’ ? 1+ ? 2= 0. 092461+ -0. 07171= 0. 020752 beyond this level of inflation its impact is negative like at percentage point of inflation level of 13 the sum of ? 1+ ? 2= -0. 00502. This implies nonlinear relationships between growth and inflation. This result is consistent with findings of Espinosa-vega and Yip (1999), Hung (2001). The finding of the threshold level of inflation is consistent with results of Sarel (1996) of 8% and Khan and Senhadji (2001) of 7% to 11% for developing countries.

As to the impact of inflation on economic growth. After establishing the threshold level of inflation the next question should be addressed is; what is the marginal impact of inflation on economic growth and financial sector development above and below the established threshold level of inflation. * Below the threshold level of 8-10%, according to the result of the regression inflation and economic growth are related positively and statistically significant. Therefore any inflation below the threshold level is very restrictive for the Ethiopian economy to grow.

For instances a marginal increase of inflation from 4% to 5% percent expands the economy by 0. 25percent. * And what happen as inflation increases beyond or above the threshold level? As one goes above 8-10% the detrimental impact of inflation intensifies. For instances for a marginal increase on inflation from 14 to 15% percent economic growth is impeded by -6. 9401 percent keeping other thing as they are. The implication of the above result is that economists now a day’s convinced that a low but positive inflation is one better environment for economic growth.

If so, this implies central banks shall keep inflation under their threshold level As to the impact of inflation on the financial sector development * Similar analysis with impact of inflation on economic growth, inflation below the threshold level of 8-10% has positive impact on the development of the financial sector. For instances for a marginal increase on inflation from 6 % to 7% increases financial sector development by 0. 5 percent. * But what surprising result of this finding is beyond some level i. e. from the threshold level of inflation, increase in inflation level have no impact on financial sector development. And above threshold level of inflation marginal increases of inflation impedes financial sector development. But what is distinct result with the above analysis of inflation on economic growth is that, further increase of inflation above the threshold level of inflation ceases to be detrimental to the financial sector. Specifically after inflation level of 12% marginal increase in annual average of inflation is no more impedes the development of the sector.

It can be argued that inflation has done its damage already, for long period since 2007 annual average inflation remains above 12% except for the year 2010 which is 2. 8% due to the price cap. First we need to have a persistent for permanent inflation rate at least for long period of time over the said range of inflation (10-12%) to pinpoint the damage inflation has done on the financial sector. According to the data at hand, we observe no persistent inflation rate between the ranges of 10-12% except short lived and point example and this makes the analysis complicated.

What we are sure about is this range if inflation financial sector ceases to be productive or have no significant contribution to the country’s economic growth. But since 2007 inflation is over 12% and financial sector positively contributing to the country’s growth and this implies above inflation level of 12% inflation have no damaging impact on the development. This can be substantiated by the expansion of credit due to new opening of banks and huge expansion of branches despite the higher inflation.

In some cases, once the rate of inflation exceeds this critical level, perfect foresight dynamics do not allow an economy to converge to a steady state displaying either an active financial system or a high level of real activity. When this occurs, further increases in inflation have no additional detrimental effects on the financial system. Thus, in effect, these models imply that once the rate of inflation reaches a certain critical threshold, “all of the damage to the financial system has already been done. ” Further increases in inflation will have no additional consequences for financial sector performance or economic growth.

Azariadis and Smith (1996) or Boyd, Choi, and Smith (1997), [Boyd and Smith 1998; Huybens and Smith 1998, 1999]. Population and investment is significant at all levels of inflation in explaining economic growth regardless of inflation threshold level of 10% and both population and investment growth affects economic growth positively. But as to exchange rate, it is insignificant but shows negative sign at very low level of inflation. But as inflation level increases specifically above 6% exchange rate begins to be significant in explaining economic growth.

Population and investment is significant due to higher population growth increase the stock of human capital which is one part of investment and will thus contribute to economic growth. When inflation increase price of home country also increases and exporters chooses to sell in domestic market rather than to foreign market. Despite the fact that both export and import of a country in recent years shows increasing trend the rate at which import increases greater than the rate at which export increases. The overall result shows trade deficit or negative net export which reduces economic growth.

Figure 4. 1: The value of k versus the residual sum of squares Table 4. 2 and figure 4. 1 illustrates the level of inflation, which is conducive for economic growth, and this is found to be 10 per cent and this is in line with the findings by Khan and Senhadji (2001). These authors found out that for developing countries, the optimal level of inflation ranges between 7 and 11 per cent and that of Sarel 8%. Diagnostic tests were done for all estimated equations as depicted in table 4. 3; however, only diagnostic results for the optimal level of inflation are depicted below.

Table 4. 3: Diagnostic tests for optimal level of inflation Equation| Test for| Test Statistic| Conclusion| K=8,9,10| 1. Normality(JB test)| P = 0. 572217| Residuals Normallydistributed| | 2. Serial Correlation(LM test)| P = 0. 416779| No serial correlation| | 3. Heteroscedasticity | P = 0. 285784| Noheteroscedasticity| | 4. StabilityCusum| Within the band| stable| The diagnostic tests carried out for all twelve equations were all satisfied. The residuals for all the estimated equations were found to be normally distributed and stable.

No serial correlation and hetroscedasticity were observed in all the equations, implying that the estimates are reliable and therefore, can be relied upon. V Conclusion and policy implication Conclusion This paper assesses the impact of the rate of inflation on the Ethiopian economy and financial sector performance. Boyd et al. (2001) show that there is a significant, and economically important, negative relationship between inflation and both stock market development and banking sector activity. Further, the relationship displayed is not linear. We extend the work of

Sarel (1996) of 8% and Khan and Senhadji (2001) to the Ethiopian economy using annual data with the Hansen’s (1999) methodology in order to estimate threshold levels. For Ethiopia, we find that inflation has a positive and significant incidence on Ethiopian economy financial sector development, below a threshold level of inflation between 8% and 10%. In other words, we show that a marginal increase of inflation is harmful to economic performance and financial sector development above threshold inflation. But for marginal increase of inflation far away from the threshold level have no impact on financial sector.

In other way saying once inflation has impact the sector its negative impacts will be null this might be due to agent’s adjustment to the situation. Attainment of a low and stable inflation rate is the fundamental objective of monetary policy worldwide, with Ethiopia bearing no exception to this rule. During the past decade or so, Central Banks worldwide have placed increasing emphasis on price stability as a macroeconomic policy strategy directed towards the achievement of higher levels of economic growth through financial stability.

Banks and financial activity within an economy contribute to the operation and development of the economy by mobilizing and pooling savings; providing payment services that facilitate the exchange of goods and services; producing and processing information about investors and investment projects in order to enable efficient allocation of funds; monitoring investments and exerting corporate governance after these funds are allocated; and helping to diversify, transform and manage risk (Demirguc-Kunt, 2008).

It is well known that central bank of Ethiopia has put its target i. e. inflation targeting in its documents and that is single digit core inflation targeting. But the questions that should be addressed is how much is single digit? 1% inflation is single digit and 9% inflation is also single digit. But which have very huge and different macro economic implications National bank of Ethiopia did not put explicitly the inflation rate it exactly targeted to and inflation level that is optimal for healthy economic growth and financial development.

Policy implication In relevance to policy conduct, according to the finding of the study inflation level between 8% and 10% gives optimal level of inflation which is conducive for economic growth via financial intermediary’s channel. Since the data is made smoothed i. e. volatility is avoided general inflation targeting in the long run can be possible in the range specified above. And this level of inflation is optimal for the financial sector development in Ethiopia. But beyond this level of inflation is detrimental to economic growth.

Considering food price is prone to weather condition the authors have adjusted the data for fluctuations. By using different data smoothing method discussed under methodology part like log, differencing and Hordic Perscot filtering method (Mubarik (2005), Monaheng Seleteng (2004). But considering core inflation only for policy targeting really under estimate the picture and most of the time we cannot target shocks once they happen we take measure if possible and they lived for short period of time.

Inflation in Ethiopia is thought to be both the result of demand and supply shocks therefore even if it is not the objective of the study at hand and the analysis do not answer such question we generally regarded appropriate prudent monetary and fiscal policy shall be exercised to maintain the suggested threshold level of inflation. Indeed, this calls for further research on the topics what exactly determine or cause inflation in Ethiopia and what measure could be taken. ——————————————– [ 2 ]. Annual average inflation is used for analysis throughout the document