Although Egypt’ inflation is approaching industrial country rates, its past record has been uneven. Until the early 1970s, inflation was low, averaging below 5 percent (see Figure 9). The situation changed in the aftermath of the first oil shock in 1973: through the mid-1980s, inflation averaged 13 percent a year. Subsequently, the pace of inflation accelerated to average more than 20 percent annually during the 1986–92 period. In the early 1990s, the Egyptian authorities undertook a bold stabilization program. This has brought inflation down to single-digit levels in 1993/94 and to just over 6 percent in 1996/97.
Rates of Consumer Price Inflation
(In percent)
Source: IMF, International Financial StatisticsRates of Consumer Price Inflation
(In percent)
Source: IMF, International Financial StatisticsRates of Consumer Price Inflation
(In percent)
Source: IMF, International Financial StatisticsMain Variables Influencing Egypt’s Inflation Record
By definition, the price level is related to the volume of real activity, the stock of money, and the velocity of circulation of the monetary aggregate.17 Table 7 summarizes the growth rates for these variables distinguishing between four periods: the early period of 1970/71–1972/73; the high growth/moderate inflation period (1973/74–1985/86); the low growth/chronic inflation period (1986/87–1991/92); and the stabilization period (1992/93–1996/97).
Summary Growth Rates of Variables Affecting Inflation
(In percent a year)
Summary Growth Rates of Variables Affecting Inflation
(In percent a year)
| Early Period 1970/71–1972/73 | High-Growth/Moderate Inflation Period 1973/74–1985/86 | Low-Growth/Chronic Inflation Period 1986/87–1991/92 | Stabilization Period 1992/93–1996/97 | |
|---|---|---|---|---|
| Inflation | 3.0 | 12.3 | 20.2 | 8.6 |
| Money supply | 8.8 | 28.5 | 20.8 | 12.7 |
| Real GDP | 4.3 | 8.1 | 3.2 | 3.2 |
| Velocity | -0.9 | -4.8 | 3.0 | -0.5 |
Summary Growth Rates of Variables Affecting Inflation
(In percent a year)
| Early Period 1970/71–1972/73 | High-Growth/Moderate Inflation Period 1973/74–1985/86 | Low-Growth/Chronic Inflation Period 1986/87–1991/92 | Stabilization Period 1992/93–1996/97 | |
|---|---|---|---|---|
| Inflation | 3.0 | 12.3 | 20.2 | 8.6 |
| Money supply | 8.8 | 28.5 | 20.8 | 12.7 |
| Real GDP | 4.3 | 8.1 | 3.2 | 3.2 |
| Velocity | -0.9 | -4.8 | 3.0 | -0.5 |
Monetary growth. In the pre-oil shock period, money growth tended to be low. During the period following the first oil shock, moderately high inflation was accommodated by a rapid monetary expansion averaging nearly 30 percent every year. In the subsequent low-growth and chronic inflation period, growth of broad money decelerated and averaged only 20 percent a year. The stabilization period (i.e., since the early 1990s) was associated with further (and sharp) deceleration in monetary growth.
Real output. In the early 1970s, Egypt entered a high-growth period with real GDP averaging more than 8 percent annually. In the subsequent high inflation period, real GDP growth decelerated and averaged only 3 percent a year. As the country underwent stabilization, growth rates initially contracted but have more recently recovered: in 1996/97, real GDP is estimated to have grown by 5 percent.
Velocity. Velocity is inversely related to money demand. The rapid GDP growth of the moderate inflation period was associated with sharp increases in money demand implying a contraction of velocity contracted at an annual rate of nearly 5 percent a year. The subsequent period saw a clear reversal as the level of velocity rose at the annual average rate of 3 percent a year. The reversal was related to liberalization of foreign exchange deposits and of the exchange rate on such deposits. This, in the context of high inflation, created an incentive for a move away from domestic currency holding. In turn, dollarization rose from 29 percent of broad money in 1985/86 to 47 percent in 1991/92.
As stabilization took hold, velocity initially rose. As confidence in the domestic currency improved (as evidenced by a decline in the dollarization rate from 47 percent in 1991/92 to 24 percent in 1996/97), velocity started to decline again. In fact, during the past year (1996/97), velocity is estimated to have declined by a sharp 3 percent.
where π is inflation, M*is the proportionate change in broad money, V* is the proportionate change in velocity, Q* is the proportionate change in real output, and ∈ is a residual item.18 Thus, inflation is directly correlated with increases in the monetary aggregate, given an unchanged level of real activity and velocity. Dividing both sides by π permits a determination of each variables’ percentage contributions to total inflation (Table 8).
Contributors to Inflation
(In percent of total inflation)
Contributors to Inflation
(In percent of total inflation)
| Early Period 1970/71–1972/73 | High-Growth/Moderate Inflation Period 1973/74–1985/86 | Low-Growth/Chronic Inflation Period 1986/87–1991/92 | Stabilization Period 1992/93–1996/97 | |
|---|---|---|---|---|
| Inflation | 100 | 100 | 100 | 100 |
| Broad money | 306 | 231 | 106 | 149 |
| Velocity | -41 | -43 | 13 | -14 |
| Real output | -146 | -67 | -17 | -42 |
| Residual | -19 | -21 | -2 | 7 |
Contributors to Inflation
(In percent of total inflation)
| Early Period 1970/71–1972/73 | High-Growth/Moderate Inflation Period 1973/74–1985/86 | Low-Growth/Chronic Inflation Period 1986/87–1991/92 | Stabilization Period 1992/93–1996/97 | |
|---|---|---|---|---|
| Inflation | 100 | 100 | 100 | 100 |
| Broad money | 306 | 231 | 106 | 149 |
| Velocity | -41 | -43 | 13 | -14 |
| Real output | -146 | -67 | -17 | -42 |
| Residual | -19 | -21 | -2 | 7 |
In the post-oil shock period, Egyptian inflation was largely a monetary phenomenon: broad money growth contributed to more than twice total inflation. The monetary expansion, however, was partly accommodated by strong real GDP growth and an increase in money demand resulting in slower velocity of broad money.
Surprisingly, during the chronic inflation period that is, when inflation accelerated to more than 20 percent a year, monetary growth had less of an impact on inflation than during the high-growth and moderate inflation period. However, the deceleration of GDP meant a more limited role for real activity in offsetting the inflationary impact of monetary growth. Moreover, the turnaround in the direction of money demand (largely associated with the dollarization period) signified that the demand for additional money in the system had faltered, thus adding significantly to inflationary pressures.
During the stabilization period, the sharp deceleration in money growth was closely correlated with reduced inflation. Real GDP growth also had an important offset on inflation, and the reversal in the direction of velocity played an even stronger role in offsetting inflation. This process culminated in 1996/97 when, despite a strong growth in money supply (15 percent), inflation fell to only 6 percent a year mainly due to the continuing sharp decline in velocity.
Monetary Developments
The previous discussion demonstrated that monetary growth has historically been closely correlated with inflation in Egypt. In this section, we describe the factors affecting money supply by decomposing the growth of the counterparts to broad money into changes in the net foreign assets of the banking system and to changes in the banking system’ domestic credit (see Table 9). The latter is further decomposed into credit to the public sector (government and state-owned enterprises) and the private sector. The residual item comprises “other items net,” which includes valuation gains and losses.
Factors Affecting Money Supply
(In percent of beginning period money stock)
Factors Affecting Money Supply
(In percent of beginning period money stock)
| Early Period | High-Growth/ Moderate Inflation Period | Low-Growth/ Chronic Inflation Period | Stabilization Period | |||
|---|---|---|---|---|---|---|
| Total money | 100 | 100 | 100 | 100 | ||
| Net foreign assets | -85 | -9 | 17 | 44 | ||
| Domestic credit | 217 | 126 | 108 | 88 | ||
| To public sector (net) | 187 | 89 | 78 | 15 | ||
| To private sector | 24 | 33 | 25 | 73 | ||
| To other banks | 6 | 4 | 5 | 1 | ||
| Other items (net) | -32 | -17 | -24 | -33 | ||
Factors Affecting Money Supply
(In percent of beginning period money stock)
| Early Period | High-Growth/ Moderate Inflation Period | Low-Growth/ Chronic Inflation Period | Stabilization Period | |||
|---|---|---|---|---|---|---|
| Total money | 100 | 100 | 100 | 100 | ||
| Net foreign assets | -85 | -9 | 17 | 44 | ||
| Domestic credit | 217 | 126 | 108 | 88 | ||
| To public sector (net) | 187 | 89 | 78 | 15 | ||
| To private sector | 24 | 33 | 25 | 73 | ||
| To other banks | 6 | 4 | 5 | 1 | ||
| Other items (net) | -32 | -17 | -24 | -33 | ||
Inflation during the high-growth and moderate inflation period was largely a monetary phenomenon. In turn, money growth was explained by a sharp growth in credit to the public sector–itself a reflection of rising budget deficits. During that period, net foreign assets of the banking sector were declining, entailing a contractionary impact on the growth of broad money.
In the subsequent chronic inflation period (1986/87–1991/92), domestic credit creation remained the key to overall money growth (albeit to a lesser extent than in the previous period). Here again, credit growth to the public sector was the key to overall monetary growth. Moreover, as dollarization took place, the change in net foreign assets added to upward pressure on monetary growth.
The stabilization period witnessed an important structural transformation in factors explaining monetary growth. Most important, as the budget deficit declined, credit to the public sector became a relatively insignificant counterpart to total monetary growth. Instead, the sharp increase in credit to the private sector (which may have had important implications for real growth) emerged as the most important contributor. Surges in capital inflows and the subsequent rise in the level of official international reserves also placed sustained upward pressure on monetary aggregates.
Estimation of Money Demand
The previous sections focused on the causes of Egypt’ inflation record through the decomposition of the Fisherian quantity of money identity. In this section, the impact of inflation on the behavior of money demand is analyzed by estimating a standard money demand equation. The nominal demand for money is estimated to be given by the following equation, using annual data from 1970/71–1996/97, with the expected sign of the coefficient shown in parentheses:
M represents the nominal monetary aggregate, CPI is the consumer price index, GDP is the real gross domestic product index, In/is the inflation rate, Int is an average interest rate measured in local currency, and Int* is the equivalent offshore interest rate.19 To determine the stability of the money demand in relation to different monetary aggregates, estimation results will be presented below for currency in circulation (CUR), domestic liquidity (M2X) excluding foreign currency deposits, and total liquidity (M2).20Before proceeding with estimation of money demand, tests for the stationarity of the variables were undertaken to determine the appropriate estimation procedure. The results show that all the level variables can be regarded as nonstationary, but their first derivative with respect lo time can only be considered stationary in the case of domestic liquidity (M2X). inflation (Inf), and the off shore interest rate (Inf). For this reason, the results below are estimated using the augmented autoregressive distributed lag procedure (ARDL) described in Pesaran and Shin (1995) and Pesaran and Pesaran (1996), which is independent of the stationarity classification needed under alternative econometric procedures. The ARDL procedure has been shown to give consistent and asymptotically efficient estimates, thus allowing for inference on the long-run coefficients.
Results
To employ the ARDL procedure, the hypothesis testing the long-run joint significance of the independent variables in a standard error correction model need not be rejected.21 The statistics for the joint test for each of the equations are considered and the results show that the hypothesis cannot be rejected at the 1 percent confidence interval (see Table 10).
ARDL Long-Run Estimates of Money Demand, 1973/74–1996/97
Numbers in parentheses represent t-statistics.
Numbers in parentheses for diagnostic tests show the confidence interval.
ARDL Long-Run Estimates of Money Demand, 1973/74–1996/97
| Monetary Aggregates | |||
|---|---|---|---|
| Currency (CUR) | Domestic liquidity (M2X) | Total liquidity (M2) | |
| Joint significance test | 1.0616 | 2.9595 | 3.0573 |
| Estimation1 | |||
| Constant | -1.0553 | -0.99063 | |
| (-3.7197) | (-19.6103) | ||
| Real GDP (GDP) | 1.0674 | 1.8680 | 1.6932 |
| (8.6650) | (6.1225) | (26.5963) | |
| Consumer price index (CPI) | 0.55573 | 0.62516 | 0.65314 |
| (6.3237) | (6.9213) | (37.1576) | |
| Inflation (Inf) | -0.22165 | -0.15793 | -0.13123 |
| (-4.0477) | (-3.1394) | (-25.102) | |
| Interest (Int) | 0.004702 | 0.010384 | 0.019414 |
| (0.79304) | (1.4085) | (15.4216) | |
| Offshore interest rate (Int*) | 0.012427 | -0.001082 | 0.003946 |
| (3.3592) | (-0.87496) | (3.1348) | |
| Dummy variable 1985/86–1991/92 | 0.0076002 | ||
| (4.6671) | |||
| Error correction model variable | -0.31561 | -1.4496 | -5.3960 |
| (-4.6538) | (-2.4086) | (-8.9574) | |
| Adjusted R2 | 0.94095 | 0.99926 | 0.99995 |
| Diagnostic tests2 | |||
| Serial correlation | 16.768 | 2.6278 | 18.3284 |
| (0.001) | (0.105) | (0.001) | |
| Functional form | 15.872 | 4.4565 | 5.1605 |
| (0.001) | (0.035) | (0.023) | |
| Heteroskedasticity | 4.7210 | 0.48045 | 2.7495 |
| (0.030) | (0.488) | (0.097) | |
Numbers in parentheses represent t-statistics.
Numbers in parentheses for diagnostic tests show the confidence interval.
ARDL Long-Run Estimates of Money Demand, 1973/74–1996/97
| Monetary Aggregates | |||
|---|---|---|---|
| Currency (CUR) | Domestic liquidity (M2X) | Total liquidity (M2) | |
| Joint significance test | 1.0616 | 2.9595 | 3.0573 |
| Estimation1 | |||
| Constant | -1.0553 | -0.99063 | |
| (-3.7197) | (-19.6103) | ||
| Real GDP (GDP) | 1.0674 | 1.8680 | 1.6932 |
| (8.6650) | (6.1225) | (26.5963) | |
| Consumer price index (CPI) | 0.55573 | 0.62516 | 0.65314 |
| (6.3237) | (6.9213) | (37.1576) | |
| Inflation (Inf) | -0.22165 | -0.15793 | -0.13123 |
| (-4.0477) | (-3.1394) | (-25.102) | |
| Interest (Int) | 0.004702 | 0.010384 | 0.019414 |
| (0.79304) | (1.4085) | (15.4216) | |
| Offshore interest rate (Int*) | 0.012427 | -0.001082 | 0.003946 |
| (3.3592) | (-0.87496) | (3.1348) | |
| Dummy variable 1985/86–1991/92 | 0.0076002 | ||
| (4.6671) | |||
| Error correction model variable | -0.31561 | -1.4496 | -5.3960 |
| (-4.6538) | (-2.4086) | (-8.9574) | |
| Adjusted R2 | 0.94095 | 0.99926 | 0.99995 |
| Diagnostic tests2 | |||
| Serial correlation | 16.768 | 2.6278 | 18.3284 |
| (0.001) | (0.105) | (0.001) | |
| Functional form | 15.872 | 4.4565 | 5.1605 |
| (0.001) | (0.035) | (0.023) | |
| Heteroskedasticity | 4.7210 | 0.48045 | 2.7495 |
| (0.030) | (0.488) | (0.097) | |
Numbers in parentheses represent t-statistics.
Numbers in parentheses for diagnostic tests show the confidence interval.
Almost all independent variables have similar coefficients in all three regressions. The elasticity on real GDP is above 1 in all regressions, a feature that is common to money demand estimates in most economies and is interpreted as an indicator of financial deepening. The coefficient on the CPI is significant but the value is below I (unity is the predicted value in the absence of money illusion). Part of the reason for the low coefficient may be due to some measurement problems in Egypt’ consumer price index. It may also be an indication of a slow response in money holdings to changes in prices.
As expected, the coefficient on inflation is negative and statistically significant. This is consistent with the notion that the private sector under high inflation will reduce its holdings of the currency to minimize the impact of the inflation tax. The interest elasticity is estimated positive in alt three regressions, but it is statistically significant only in the regression for total liquidity (Ml).
The coefficient on the offshore interest rate is positive and statistically significant, except for the regression for domestic liquidity (M2X), where the exclusion of foreign currency deposits weakens the statistical relationship. The positive coefficient on the offshore interest rate (int*) is puzzling, since interest arbitrage would suggest that higher offshore interest rates net of currency depreciation may lead to capital outflow. In this case, capital account restrictions in Egypt up to 1991 may have attenuated this link and strengthened the effects on domestic foreign currency deposits. The dummy variable encompassing the peak of the dollarization experience is only significant for the regression on total liquidity as expected, as this is the only monetary variable involving dollar-denominated deposits.
Standard diagnostic tests were also run for all three regressions and are presented in the bottom part of Table 10, with the confidence interval shown in parentheses below the statistics. The only significant problems relate to the regression for domestic liquidity (M2X), where the hypotheses of serial correlation and heteroskedasticity cannot be rejected at the 10 percent confidence interval. In all other cases, the diagnostic tests can be rejected at the 1 percent confidence interval.
Overall, the results in Table 10 suggest that the nominal demand for money in Egypt is closely associated with the price level, real output, and inflation and that dollarization in the form of foreign currency deposits played a significant role in monetary developments during the second half of the 1980s. Moreover, of the three measures of money analyzed here, narrow (CUR) and broad money (M2) demand have been the most robust over the period, tending to confirm the appropriateness of targeting money demand in Egypt as a guide to monetary policy, provided that out-of-sample projections of independent variables can be confidently projected.
Monetary Policy Objectives with Moderate Inflation
Previous sections have shown the extent of Egypt’ successful reduction of inflationary pressures in the last six years. The main source of the success lies ultimately in the improved fiscal stance, which has over the years alleviated the pressure on domestic liquidity growth coming from the financing of the government deficit. In addition, the successful management of capital inflows in 1991–93, and again in 1996–97, has ensured that inflation has been kept under control in spite of the large increases in net foreign assets of the banking system (see Section V for a discussion of policy responses to capital inflows). Looking ahead, both the ongoing trade liberalization and the structural reforms should increase competition, thus reducing pressure on prices from monopolistic practices. This structural element is likely to further contribute to disinflation.
At this juncture in the stabilization process, with annual inflation in the region of 5 percent, two important issues in Egypt’ monetary policy require close scrutiny. First, what should be Egypt’ medium-term inflation target given its growth objectives? And second, what policies and institutional aspects can best achieve this target?
What Rate of Inflation?
The optimal rate of inflation a country should target has been extensively discussed in the academic literature, with somewhat inconclusive results. While there is a broad consensus on the negative effects on investment and growth of inflation rates above 10 percent, there is little agreement on the relative benefits of price stability (0–2 percent inflation) versus moderate inflation (2–10 percent). Arguments in favor of somewhat higher inflation targets have also been supported for developing and transitional economies with significant nominal rigidities, on the basis that liberalization and structural reforms induce substantial changes in relative prices and thus put upward pressure on inflation.
From a theoretical perspective, an assessment of the optimal inflation target can be divided into three aspects: (1) short-term costs of disinflation, (2) medium-and long-term benefits of low inflation on growth, and (3) the credibility effect of low inflationary policies.22
The short-term costs of disinflation are well-known and are associated with the extensive literature on the Phillips curve. The policy implication is that, to bring down inflation, aggregate demand needs to be reduced so as to alleviate pressure on prices and the cost structure of production. In particular, the slope of the Phillips curve, the so-called sacrifice ratio, gives a quantitative measure of the short-run output cost involved in disinflation policies. In countries where the Phillips curve is estimated to be nonlinear, the sacrifice ratio will be higher at lower rates of inflation.
The theoretical and empirical evidence on the impact of moderate inflation on investment and growth is mixed.23 On the theoretical side, it is argued that macroeconomic uncertainty associated with inflation reduces the incentives for domestic and foreign investment and thus leads to a lower growth path. It is unclear though what rate of inflation will bring about this negative impact on investment. On the empirical side, there is some evidence that moderate rates of inflation compared with price stability may actually foster investment and growth, as they allow enough flexibility for relative prices to adjust to structural changes brought about by growth. The benefits of moderate inflation or price stability are therefore not easily quantified, and this leads to the contention between the arguments in favor, or against, strict price stability vis-à-vis moderate inflation.
The credibility of monetary policy may be of particular relevance to the Egyptian case. From Barro and Gordon (1983), it is clear that the systematic pursuit of low inflation over the medium term has the added benefit that it increases the credibility of the central bank in the eyes of the private sector and thus reduces the uncertainty associated with the “inflation bias” of monetary policy. The benefit of increased credibility is cumulative inasmuch as central banks are recognized to have either a strong or moderate stance against inflation from their past performance. The issue of credibility is particularly important for Egypt, as the monetary authorities since 1991 have signaled their commitment to low inflation through the maintenance of a narrow band between the Egyptian pound and the U.S. dollar. The exchange rate policy has been successful in reducing inflationary expectations, as shown by the steady decline of real interest rates, thus adding credibility to monetary policy.
Further efforts will be needed to increase credibility in Egypt’ monetary policy in the years ahead, as indicated by the premium associated with short-term domestic interest rates–typically about 3 percentage points over comparable U.S. rates. The pegging of the Egyptian pound to the U.S. dollar in the presence of a significant, if falling, inflation differential has generated real exchange rate appreciation. In Section VI, it is shown how such real appreciation is approximately consistent with underlying internal and external equilibrium, in particular higher productivity growth rates in Egypt and with the prolonged confidence effects of the Paris Club debt reduction. Should economic fundamentals (terms of trade, productivity growth, and so forth) point to a more moderate appreciation of the equilibrium rate in the future, the maintenance of the exchange rate band may require additional disinflation so as to bring the Egyptian inflation rate closer in line with that of the United States. Unless this is accomplished, and unless there are substantial improvements in Egypt’ average productivity, the erosion in competitiveness brought about by the inflation differential may reduce the attractiveness of Egyptian investments and put pressure on the balance of payment in the medium term. The cost of the additional disinflation effort will consist of further moderation of aggregate demand in the short run. The benefit will be the achievement of price stability, which could possibly induce further domestic and foreign investment in the medium term and added credibility to monetary policy. If the current rate of inflation is deemed appropriate based on the overall assessment of costs and benefits, a more flexible exchange rate policy would likely be required.
How to Achieve the Inflation Target?
In addition to the question of what Egypt’ optimal inflation target should be, there is also the parallel issue of how best to achieve the target from a policy and an institutional perspective. The implementation of the inflation target involves the issue of which monetary framework to use for the conduct of monetary policy and, more generally, how to manage short-term liquidity to the banking system and improve the institutional arrangements to achieve this goal.
The choice of an effective framework for conducting monetary policy is a vibrant topic in the theoretical literature. In many countries, the discussion in recent years has focused on the merits of targeting monetary aggregates or future inflation.24 The former involves the targeting of either narrow or broad money to a growth rate that is consistent with that of the real economy plus the desired inflation objective. The latter involves the announcement of an official inflation target for the medium term, and the formulation and publication of an inflation forecasting model that links the set of instruments used for liquidity control, usually short-term interest rates, to the inflation forecast for the medium term (12 months to 24 months ahead).
The adoption of direct inflation targeting in a number of industrial countries and its success in reducing inflation in recent years25 suggest that there may be advantages to targeting inflation directly. One of the main motivations for switching to inflation targeting is that the demand for money tends to become relatively unstable as financial innovation loosens the relationship between narrow money and domestic credit. Also, a direct inflation-targeting framework prescribes greater transparency in the formulation and operation of monetary policy. This gives the private sector a way to measure directly the performance of the central bank; hence, it reduces the uncertainty surrounding monetary policy.
While the case for inflation targeting in Egypt is not as strong as in other economies, it still reveals important institutional aspects that need further enhancement. Many of these enhancements are already under way. Supported by technical assistance from the IMF and the U.S. Agency for International Development (USAID), the Egyptian authorities are improving the statistical database on real sector indicators, which should aid in the conduct of monetary policy. In addition, the set of monetary instruments is being widened by the introduction of sale and repurchase (repo) operations and reverse repo operations and by lengthening the maturity structure of government debt. The repo market will allow for improved short-term liquidity management and foster the secondary market for short-term government debt. The lengthening of the maturity structure will provide additional liquidity to the long end of the market and thus deepen the financial system. All these actions should, in the medium term, improve the monetary policy framework and facilitate the effective implementation of monetary policy.
