Abstract
The process of currency substitution—whereby foreign-currency-denominated money balances increasingly substitute for domestic money as a store of value, unit of account, and medium of exchange— has gained importance in many countries over the past few years. This process, reflecting a shift toward foreign money in the preferences of residents that often arises from large and growing economic and financial imbalances in the domestic economy, has induced a corresponding increase in professional interest in the subject. Accordingly, several studies have been undertaken on the growth in currency substitution and its implications for fundamental macroeconomic relations as well as traditional statistical concepts. The quantitative application of the resultant framework of analysis has focused primarily on the experience of industrial countries,1 with the relatively few studies on developing economies restricted almost exclusively to countries in Latin America.2 This paper seeks to extend the quantitative analysis of currency substitution to two Middle Eastern countries, Egypt and the Yemen Arab Republic (Y.A.R.).3
The process of currency substitution—whereby foreign-currency-denominated money balances increasingly substitute for domestic money as a store of value, unit of account, and medium of exchange— has gained importance in many countries over the past few years. This process, reflecting a shift toward foreign money in the preferences of residents that often arises from large and growing economic and financial imbalances in the domestic economy, has induced a corresponding increase in professional interest in the subject. Accordingly, several studies have been undertaken on the growth in currency substitution and its implications for fundamental macroeconomic relations as well as traditional statistical concepts. The quantitative application of the resultant framework of analysis has focused primarily on the experience of industrial countries,1 with the relatively few studies on developing economies restricted almost exclusively to countries in Latin America.2 This paper seeks to extend the quantitative analysis of currency substitution to two Middle Eastern countries, Egypt and the Yemen Arab Republic (Y.A.R.).3
Egypt and the Y.A.R. have experienced an increasing level of currency substitution as measured by residents’ holdings of foreign currency deposits, both in absolute foreign currency terms and relative to their domestic currency balances. This preference of individuals to substitute foreign money for domestic money, and the prospects for continuation of this phenomenon, raises several economic issues—in particular, the implications of currency substitution for the design, execution, and effectiveness of monetary, fiscal, and exchange rate policies. To provide a satisfactory framework for further analysis of such issues, this paper attempts a preliminary empirical investigation, based on available partial data, of the magnitude and determinants of private sector currency substitution in these countries over the period 1980-86.
To this end. Section I provides background for the statistical analysis by outlining the theoretical determinants of currency substitution, with a focus on transaction, risk, and return considerations, as well as on institutional factors. On the basis of this discussion, a reduced-form model of money demand is specified that explicitly incorporates foreign-currency-denominated nominal money holdings and the associated effects on portfolio-balance decisions of asset holders. The model is estimated using quarterly data for Egypt and the Y.A.R.—two countries that are shown (in Section II) to have widespread and growing currency substitution, as measured by residents’ holdings of foreign currency deposits in domestically located banks. The results of the estimation, presented in Section III, provide an indication of the absolute and relative importance of various determinants of private sector currency substitution during the period under consideration. Some concluding remarks are offered in Section IV.
I. Determinants and Implications of Currency Substitution
Currency substitution may be defined as a phenomenon in which foreign-currency-denominated money has replaced, either wholly or in part, domestic money in serving as store of value, medium of exchange, and unit of account.4 This phenomenon, which in essence reflects individuals’ attempts to protect the value of their wealth and income, usually has taken place in the context of deteriorating economic and financial conditions that adversely affect the return on holdings of domestic money relative to those on foreign-currency-denominated money balances, as well as in the context of changes in the political and institutional environments that influence expectations regarding the absolute and relative liquidity of domestic- and foreign-currency-denominated assets.
The growth in currency substitution in several countries has been associated with increasing awareness of the consequent implications for economic and financial policy management. The numerous articles in this area have demonstrated that large and growing currency substitution is likely to be associated with changes in the effectiveness of fiscal, monetary, and exchange rate policies. In particular, it has been shown that currency substitution, among other effects, (1) contributes to a weakening of already limited fiscal instruments by widening the possibilities for tax avoidance and reducing the yield of the inflation tax;5(2) reduces the authorities’ control over domestic liquidity by enlarging the component over which the authorities have little direct influence and by reducing the stability of money demand;6 and (3) weakens the effectiveness of exchange rate policy by reducing the disabsorption effect of a given devaluation.7 In addition, the formalization of the process of currency substitution through the introduction and promotion of foreign currency deposits has been shown to require a corresponding adjustment in credit policies and regulatory practices to ensure that the financial integrity of the banking system is not adversely affected by a growth in uncovered foreign liabilities. Such policies must address the maturity and currency coverage, as well as the ability to generate sufficient foreign exchange resources in order to meet the interest obligations to depositors.8
In view of the adverse policy implications of widespread currency substitution, investigations in this area have also sought to clearly identify the specific determinants of this phenomenon. This process has been undertaken in a number of theoretical frameworks, including those based on the marginal utility theory of money demand,9 on the portfolio balance approach,10 on the theory of the precautionary demand for cash balances,11 and on the money-services production function approach (Miles (1978)). A consolidation of the various issues raised in these analyses suggests that the main determinants of currency substitution are actual or expected reductions in the value of domestic money holdings (in terms of the current or future command over goods and services) occasioned by increased domestic inflationary pressures, expectations of exchange rate depreciations, and lower interest rates on domestic currency holdings. By altering the relative yield to holdings of foreign-currency-denominated money balances, and therefore their attractiveness as a store of value, these factors induce a change in the pace of currency substitution. This process may be further affected by uncertainties about sociopolitical developments and agents’ expenditure patterns.
Although the magnitude of currency substitution will be primarily influenced by shifts in relative returns to money balances at a time of domestic and external imbalances, its character will also be affected by the prevailing institutional environment. In particular, and for a given level of financial and economic disequilibria, the extent of currency substitution will be higher the narrower is the range of competitive domestic assets, the lower are the legal risk premiums attached to holdings of foreign currency balances, and the smaller are the transaction costs involved in acquiring such balances. Thus, currency substitution is likely to grow faster in countries with limited capital markets and other domestic investment outlets. In such environments, foreign currency balances increasingly become a more remunerative form of savings because they avoid the losses associated with the deterioration in domestic purchasing power occasioned by worsening economic and financial conditions. This factor will be supplemented in agents’ formulation of portfolio decisions by considerations relating to the acquisition, maintenance, and use of foreign currency balances. Clearly, the wider are the possibilities for obtaining these balances, the larger is the potential for rendering effective a given level of notional demand for foreign currency holdings. Similarly, the greater are the channels for the legal maintenance and use of such balances, the wider are the possibilities for translating agents’ desired holdings of foreign currency balances into actual holdings. Accordingly, the extent of currency substitution will be more widespread, other things being equal, in an environment in which, for example, there is an absence of total-surrender requirements for foreign exchange receipts, there exist formalized and secure channels for holdings of foreign currency (for example, through appropriately maintained foreign currency bank deposits), and there are relatively few restrictive regulations on the type of transaction that can legitimately be financed.
The discussion of the factors determining the level and character of currency substitution can be partially summarized in a simplified composite, reduced-form demand for money function. This function, which provides the basis for the subsequent quantitative analysis of currency substitution in Egypt and the Y.A.R., is derived on the basis of two separate underlying money demand equations for domestic and foreign-currency-denominated balances, each of which incorporates relevant transaction demand variables as well as those measuring the opportunity cost of holding money. Manipulation of these equations, taking account of the qualifications discussed below, yields a reduced-form function that specifies the demand for foreign balances relative to domestic balances as dependent on the differential returns to the two holdings. In the context of a rapidly deteriorating economic and financial situation in which structural and institutional rigidities preclude the satisfaction of interest rate parity conditions, this differential will be reflected primarily in expectations of exchange rate changes. In addition, the specification incorporates a partial adjustment mechanism to reflect possible lags in the adjustment of desired holdings and a term to capture major changes in the institutional and political environments.
The general function, including the a priori signs of the first-order partial derivatives, can therefore be written as
where
M = nominal holdings of money balances
i = the nominal interest rate
e = the nominal exchange rate
e*= the expected nominal exchange rate
Φ = a term capturing political and institutional changes, and the superscripts d and f refer to the domestic and foreign components, respectively.
In considering the specification and application of the model, two important qualifications should be noted. First, the specification of the model is based on several simplifying assumptions. In particular, it is assumed that the scale variables in the two underlying money demand functions are similar and therefore play no role in determining the demand for local balances relative to that for foreign balances; thus, for example, the level of agents’ wealth does not appear explicitly in the reduced-form function. In addition, it is assumed that both the transaction costs involved in maintaining and switching between local and foreign balances and the overall level of monetization remain broadly unchanged throughout the estimation period. An indication of the effects of such simplifying assumptions on the ability of the model to account for changes in currency substitution, in the period under consideration, will be provided by the value of the various statistical goodness-of-fit measures.
The second qualification relates to the specification of the dependent variable in the model. Whereas the concept of currency substitution is relatively easy to define, its measurement poses considerable difficulties. Most attempts to quantify the level and changes in currency substitution, including the present one, focus on residents’ relative holdings of foreign currency deposits in domestically located banks. In using such an indicator, it is important to recognize that no account is taken of deposits in banks abroad and of nonbank transactions; accordingly, this measure will underestimate the magnitude of currency substitution if residents hold significant foreign currency deposits in banks abroad and if a significant portion of financial intermediation in foreign currency takes place outside the formal banking system. Data on residents holdings of foreign currency deposits in banks abroad, reported in International Financial Statistics (International Monetary Fund, various issues), indicate that such holdings were not insignificant in the case of Egypt and the Y.A.R.13 The effect on the estimation of the model of excluding these deposits, however, is likely to be small because, in the period under consideration, both types of deposits experienced similar variations.14 As regards nonbank transactions—particularly those relating to cash holdings of foreign exchange—casual empiricism suggests that such transactions are quite widespread, especially in the Y.A.R. where banking habits are relatively underdeveloped.15 Nevertheless, a statistical analysis of currency substitution based on recorded domestic bank holdings will continue to provide meaningful insights if it is carried out within a framework that recognizes the impact of financial and economic disequilibria. Specifically, one would expect these disequilibria to be associated both with a growth in overall currency substitution, as mentioned earlier, and with a shift in the composition of agents’ holdings away from deposits in domestically located banks. Accordingly, if the estimation exercise, incorporating the latter holdings as a dependent variable, points to a positive and significant relationship between these holdings and the explanatory variables for financial and economic disequilibria, the model would provide support for the more general hypothesis of currency substitution encompassing alternative forms of foreign exchange holdings.16
II. Currency Substitution in Egypt and the Yemen Arab Republic
The extent of private sector currency substitution in Egypt and the Y.A.R., as measured by a partial indicator based on the levels of foreign currency bank deposits, has expanded significantly in recent years and is now similar to that prevailing in a number of Latin American countries.
Foreign Currency Deposits, 1979-86
(Period averages: in millions of U.S. dollars)
Citation: IMF Staff Papers 1988, 001; 10.5089/9781451956771.024.A004
Source: International Monetary Fund. International Financial Statistics (IFS). and Fund staff estimates.Based on data for 1980.Based on data for the six quarters ended June 1986.Foreign Currency Deposits, 1979-86
(Period averages: in millions of U.S. dollars)
Citation: IMF Staff Papers 1988, 001; 10.5089/9781451956771.024.A004
Source: International Monetary Fund. International Financial Statistics (IFS). and Fund staff estimates.Based on data for 1980.Based on data for the six quarters ended June 1986.Foreign Currency Deposits, 1979-86
(Period averages: in millions of U.S. dollars)
Citation: IMF Staff Papers 1988, 001; 10.5089/9781451956771.024.A004
Source: International Monetary Fund. International Financial Statistics (IFS). and Fund staff estimates.Based on data for 1980.Based on data for the six quarters ended June 1986.The growth of these deposits took place against a background of a higher availability of private sector foreign exchange transfers in the 1970s, mainly the result of increased employment of nationals in neighboring oil producing countries. The introduction and promotion of foreign currency bank deposits provided the private sector with a legal means of holding the higher foreign currency balances within the banking system. At the same time, it appears that these deposits were viewed by the authorities as a way of increasing the supply of foreign exchange resources to their respective economies, as well as a method for directing these resources into formal banking channels, thereby providing the authorities with increased potential control over the process of financial intermediation of domestically located foreign currency balances. This section considers the magnitude and growth of this measure of currency substitution, thereby setting the stage for the statistical analysis of this process presented in the following section.
The introduction of foreign currency deposits in Egypt in 1974 was part of a broader effort to liberalize the economy. Although comprehensive data are not available for the early years, it appears that these deposits quickly became a significant component of residents’ holdings of money. By the early 1980s, foreign currency deposits were well established in the banking system and continued to grow in a deteriorating domestic economic and financial environment characterized by increasing inflationary pressures and a weakening external position. In the six-year period that ended June 1986, foreign currency deposits in domestically located banks grew at an annual rate of 22 percent in U.S. dollar terms (Figure 1).17 Although this growth was reflected both in demand and in time and savings deposits, the latter increased at a more rapid rate, registering an annual growth rate of 24 percent during this period.18 By the end of the period, residents’ holdings of foreign currency in domestically located banks, amounting to some US$7.3 billion, were far in excess of the country’s gross official reserves. If recorded foreign currency accounts in banks located abroad are included, Egyptian residents’ total foreign currency holdings were equivalent to US$10.3 billion in June 1986.
The growth in residents’ holdings of foreign currency deposits has had a significant impact on the level and composition of Egypt’s money supply. The rapid expansion of these deposits, compounded by stock adjustments arising from exchange-rate-induced valuation effects, have been reflected in the substantial growth of the recorded domestic liquidity. Moreover, because these deposits grew at a faster rate than other components of money, their relative importance increased from some 25 percent of broad money in June 1980 to around 40 percent in June 1986 (Figure 2).
Foreign Currency Deposits as a Proportion of Broad Money, 1979-86
(Period averages: in percent)
Citation: IMF Staff Papers 1988, 001; 10.5089/9781451956771.024.A004
Source: IFS and Fund staff estimates.Based on data for 1980.Based on data for the six quarters ended June 1986.Foreign Currency Deposits as a Proportion of Broad Money, 1979-86
(Period averages: in percent)
Citation: IMF Staff Papers 1988, 001; 10.5089/9781451956771.024.A004
Source: IFS and Fund staff estimates.Based on data for 1980.Based on data for the six quarters ended June 1986.Foreign Currency Deposits as a Proportion of Broad Money, 1979-86
(Period averages: in percent)
Citation: IMF Staff Papers 1988, 001; 10.5089/9781451956771.024.A004
Source: IFS and Fund staff estimates.Based on data for 1980.Based on data for the six quarters ended June 1986.The Y.A.R. also experienced a significant growth in currency substitution, as measured by limited data on private sector foreign currency time and savings deposits in domestically located banks. In the fourteen-year period ended June 1986, these deposits grew at an annual rate of 36 percent in U.S. dollar terms, increasing to some US$270 million by the end of this period.19 The increase in the earlier portion of this period was part of a more generalized movement toward higher monetization and formal financial intermediation in the Yemeni economy which, as noted above, occurred against a background of increased foreign exchange inflows. Foreign currency deposits grew more rapidly in the 1980s in an environment of deteriorating economic and financial conditions and changing institutional factors. For example, there was a substantial rise in the size of these deposits in 1983/84 that coincided with changes in banking regulations governing selected foreign exchange activities of commercial banks. These changes had the effect of reducing agents’ access to the banking sector for acquiring and using foreign exchange to finance imports. The resultant higher demand for “own-financed transactions” is likely to have led to an increased demand for foreign exchange holding by traders—a hypothesis that is investigated below.
In terms of broad money, the share of foreign currency deposits in the Y.A.R. rose from under 1 percent in June 1980 to over 10 percent in June 1986. Because the money supply is dominated by local currency in circulation, a better measure of the growing importance of these deposits is provided by their share in total financial sector deposits. This share increased from under 5 percent in June 1980 to around 25 percent in June 1986.
It is clear from the above that the phenomenon of currency substitution is important in both Egypt and the Y.A.R. The following section presents a preliminary statistical investigation of the determinants of this phenomenon during the period 1980-86. An attempt is also made to identify additional data and analytical limitations, the further consideration of which could constitute a basis for supplementary empirical country work in this period and for an extension of the analysis over a longer time period.20
Estimates of the Determinants of Currency Substitution in Egypt
Standard error of estimate: * denotes significance of coefficient at the 95 percent confidence level; ** denotes significance ai the 99 percent confidence level.
Estimates of the Determinants of Currency Substitution in Egypt
| Variable | Constant | Exchange Rate Expectations | Interest Rate Differentials | Political Disruptions | Stock Adjustment | DW | Standard Error of Regression | ||
|---|---|---|---|---|---|---|---|---|---|
| Equation (1) | |||||||||
| Coefficient | -0.362 | 0.069 | 0.099 | 0.103 | 0.698 | — | — | — | |
| SEEa | 0.104** | 0.028* | 0.354 | 0.032** | 0.085** | 0.94 | 2.52 | 0.031 | |
| Equation (2) | |||||||||
| Coefficient | -0.364 | 0.065 | … | 0.104 | 0.694 | — | — | — | |
| SEEa | 0.101” | 0.024** | … | 0.031** | 0.081** | 0.94 | 2.48 | 0.030 | |
| Equation (3) | |||||||||
| Coefficient | -0.233 | 0.039 | 0.296 | 0.079 | 0.722 | — | — | — | |
| SEEa | 0.082* | 0.022 | 0.310 | 0.028* | 0.099” | 0.81 | 2.22 | 0.027 | |
| Equation (4) | |||||||||
| Coefficient | -0.231 | 0.242 | … | 0.083 | 0.713 | — | — | — | |
| SEEa | 0 822* | 0.015 | … | 0.028** | 0.099** | 0.81 | 2.06 | 0.027 | |
Standard error of estimate: * denotes significance of coefficient at the 95 percent confidence level; ** denotes significance ai the 99 percent confidence level.
Estimates of the Determinants of Currency Substitution in Egypt
| Variable | Constant | Exchange Rate Expectations | Interest Rate Differentials | Political Disruptions | Stock Adjustment | DW | Standard Error of Regression | ||
|---|---|---|---|---|---|---|---|---|---|
| Equation (1) | |||||||||
| Coefficient | -0.362 | 0.069 | 0.099 | 0.103 | 0.698 | — | — | — | |
| SEEa | 0.104** | 0.028* | 0.354 | 0.032** | 0.085** | 0.94 | 2.52 | 0.031 | |
| Equation (2) | |||||||||
| Coefficient | -0.364 | 0.065 | … | 0.104 | 0.694 | — | — | — | |
| SEEa | 0.101” | 0.024** | … | 0.031** | 0.081** | 0.94 | 2.48 | 0.030 | |
| Equation (3) | |||||||||
| Coefficient | -0.233 | 0.039 | 0.296 | 0.079 | 0.722 | — | — | — | |
| SEEa | 0.082* | 0.022 | 0.310 | 0.028* | 0.099” | 0.81 | 2.22 | 0.027 | |
| Equation (4) | |||||||||
| Coefficient | -0.231 | 0.242 | … | 0.083 | 0.713 | — | — | — | |
| SEEa | 0 822* | 0.015 | … | 0.028** | 0.099** | 0.81 | 2.06 | 0.027 | |
Standard error of estimate: * denotes significance of coefficient at the 95 percent confidence level; ** denotes significance ai the 99 percent confidence level.
III. Quantitative Analysis of Currency Substitution in Egypt and the Yemen Arab Republic
To obtain an indication of the absolute and relative importance of the determinants of currency substitution in Egypt and the Y.A.R., the reduced-form function presented in Section I was estimated using quarterly data for the period 1980-86. The dependent variable (relative holdings of foreign currency deposits) was specified in two alternative ways that capture narrow substitution between foreign currency bank deposits and domestic currency bank accounts and broader substitution between foreign currency bank deposits and holdings of domestic balances encompassing both local currency bank deposits and cash holdings. The results of the various estimations are presented in Tables 1 and 2. The discussion in the text focuses on the specifications incorporating the concept of broader monetary substitution (that is, equations (1) and (2) in Table 1 and (5) and (6) in Table 2).
The results of the estimations suggest that agents’ preferences for increased relative holdings of foreign currency balances, as measured by the growth in foreign currency time and savings bank deposits relative to total money holdings, reflected anticipation of higher yields on foreign currency balances attributable to expectations of exchange-rate-induced capital gains. In addition, increased political uncertainties and changes in banking regulations provided an impetus to preferences for foreign money in lieu of domestic money. Finally, the results indicate that, given a change in these causal factors, actual holdings of foreign currency balances adjusted with a time lag to their unconstrained desired levels. The detailed country results are summarized in the following subsections.
Egypt
All estimated coefficients for Egypt (Table 1) had the expected signs and, with the exception of the term relating to interest rate differentials,21 were significant at the 95 percent confidence level; three of the coefficients were significant at the 99 percent confidence level (equation (1) in Table 1). In addition, changes in the explanatory variables were found to account for the bulk of the variations in the extent of currency substitution.2
The results of the estimation indicate that increases in the importance of currency substitution in Egypt were associated, among other things, with higher expectations of exchange rate depreciations and with greater political uncertainty.23 The proxy for exchange rate expectations—an unobservable variable—was specified in terms of the level of the parallel market rate, which is known to be highly flexible and the most responsive of the various exchange rates maintained in Egypt to changes in underlying demand and supply conditions, relative to the current spot formal sector rate.24 On the basis of this specification, the elasticity of currency substitution with respect to exchange rate expectations during the period under consideration amounted to 0.07—that is, a 10 percent increase in the expected command of foreign currency over local goods and services resulted, other things being equal, in a 0.7 percent increase in the share of foreign currency in broad money.25 The relative high value of the partial adjustment coefficient suggests that agents did not immediately and fully adjust their holdings to variations in relative yields of foreign balances. As a result, actual holdings at any one time significantly affected behavior in the next period.
Although changes in the proxy for interest rate differentials were not found to be statistically significant in explaining variations in the extent of currency substitution, this result should be interpreted with considerable caution. In specifying the causal relationships, difficulties were encountered in deriving an adequate proxy for the interest rate differential. In particular, because of the repressed state of domestic financial markets, the formal sector (administered) domestic interest rate variable used in the estimation is a poor indicator of available alternative returns to holdings of domestic balances; moreover, no account is taken of the return on other domestic assets, including instruments associated with the growth of Islamic investment companies in Egypt. In view of these factors, and because of the associated rigidities in the formal sector of the economy that precluded the attainment of interest rate parity in the period under consideration, a relative demand for money function excluding an interest differential term was estimated (equation (2) in Table 1).26 The signs of all estimated coefficients in this second specification met the a priori expectations; in addition, all coefficients were found to be significant at the 99 percent confidence level, and their values were similar to those of the earlier estimation. These results provide further support for the findings cited above, as well as for the hypothesis regarding the limitations of the interest rate proxy.
Yemen Arab Republic
In the case of the Y.A.R. (Table 2), all estimated coefficients also met the a priori expectations regarding their signs and, with the exception of those for the relative interest rate and constant terms, were significant at the 95 percent confidence level (equation (5) in Table 2), In addition, the coefficient of determination indicated that most of the variations in foreign currency deposits could be “explained” by changes in the explanatory variables.27 Estimation of the money demand function excluding the relative interest rate term, undertaken for reasons similar to those cited for the case of Egypt, also yielded highly significant results and satisfactory goodness-of-fit measures (equation (6) in Table 2).
Compared with Egypt, changes in currency substitution in the Y.A.R. appear to have been more responsive to anticipations of increased yields on foreign currency arising from exchange rate depreciations. In particular, the elasticity with respect to exchange rate expectations28 was estimated to amount to around 0.7 for the period under consideration.29 The changes in the exchange and trade system in 1983/84 appeared to have resulted, as anticipated, in an upward shift in the level of currency substitution by the private sector in the Y.A.R., thereby providing sup-port for the hypothesis that the limitation of access to foreign exchange resources through the banking system led traders to accumulate their own stock for transaction purposes. As regards the speed of portfolio adjustment, the value of the estimate of the partial adjustment coefficient was relatively large, indicating sluggishness in agents’ adaptation of actual balances to their desired levels.
Estimates of the Determinants of Currency Substitution in the Yemen Arab Republic
Standard error of estimate;
denotes significance of coefficient at the 95 percent confidence level;
denotes significance at the 99 percent confidence level.
Estimates of the Determinants of Currency Substitution in the Yemen Arab Republic
| Exchange | Interest | Standard | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Rate” | Rate | Institutional | Stock | Error of | |||||
| Variable | Constant | Expectations | Differential | Changes | Adjustment | DW | Regression | ||
| Equation (5) | |||||||||
| Coefficient | -0.370 | 0.683 | 1.194 | 1.146 | 0.905 | — | — | — | |
| SEEa | 0.332 | 0.618* | 1.935 | 0.209** | 0.076** | 0.95 | 2.71 | 0.196 | |
| Equation (6) | |||||||||
| Coefficient | -0.347 | 0.758 | … | 1.163 | 0.920 | — | — | — | |
| SEEa | 0.325 | 0.596* | … | 0.204** | 0.071** | 0.95 | 2.66 | 0.193 | |
| Equation (7) | |||||||||
| Coefficient | -0.248 | 0.541 | 1.084 | 1.152 | 0.910 | — | — | — | |
| SEEa | 0.259 | 0.598 | 1.896 | 0.210** | 0.083** | 0.94 | 2.65 | 0.195 | |
| Equation (8) | |||||||||
| Coefficient | -0.247 | 0.625 | … | 1.168 | 0.923 | — | — | — | |
| SEEa | 0.255 | 0.570 | … | 0.204** | 0.078** | 0.94 | 2.61 | 0.191 | |
Standard error of estimate;
denotes significance of coefficient at the 95 percent confidence level;
denotes significance at the 99 percent confidence level.
Estimates of the Determinants of Currency Substitution in the Yemen Arab Republic
| Exchange | Interest | Standard | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Rate” | Rate | Institutional | Stock | Error of | |||||
| Variable | Constant | Expectations | Differential | Changes | Adjustment | DW | Regression | ||
| Equation (5) | |||||||||
| Coefficient | -0.370 | 0.683 | 1.194 | 1.146 | 0.905 | — | — | — | |
| SEEa | 0.332 | 0.618* | 1.935 | 0.209** | 0.076** | 0.95 | 2.71 | 0.196 | |
| Equation (6) | |||||||||
| Coefficient | -0.347 | 0.758 | … | 1.163 | 0.920 | — | — | — | |
| SEEa | 0.325 | 0.596* | … | 0.204** | 0.071** | 0.95 | 2.66 | 0.193 | |
| Equation (7) | |||||||||
| Coefficient | -0.248 | 0.541 | 1.084 | 1.152 | 0.910 | — | — | — | |
| SEEa | 0.259 | 0.598 | 1.896 | 0.210** | 0.083** | 0.94 | 2.65 | 0.195 | |
| Equation (8) | |||||||||
| Coefficient | -0.247 | 0.625 | … | 1.168 | 0.923 | — | — | — | |
| SEEa | 0.255 | 0.570 | … | 0.204** | 0.078** | 0.94 | 2.61 | 0.191 | |
Standard error of estimate;
denotes significance of coefficient at the 95 percent confidence level;
denotes significance at the 99 percent confidence level.
IV. Concluding Remarks
Holdings of foreign-currency-denominated bank balances by residents in Egypt and the Y.A.R. have grown substantially in recent years and now constitute an important component of domestic liquidity in these countries. The growth of these holdings, made possible by the increased access of the private sector to foreign exchange (in large part attributable to additional employment opportunities in neighboring oil producing economies), has recently reflected attempts by agents to protect and increase the value of their wealth and income stream in an environment of deteriorating financial and economic conditions. In particular, the results of quantitative analysis, based on partial indicators of currency substitution in the 1980s, suggest that the growth in agents’ foreign currency portfolio preferences was strongly influenced by expectations of increased financial yields on holdings of foreign-currency-denominated balances relative to those on domestic balances; in addition, the portfolio preferences during this period were found to be significantly affected by political and institutional changes.
The analysis suggests that, for a given political and institutional environment, an intensification of inflationary pressures and a further deterioration in the external accounts, to the extent that these factors adversely affect exchange rate expectations, will be reflected in an increased tendency by residents to substitute foreign balances for local balances in meeting their demands for a store of value, unit of account, and medium of exchange. At the same time, given the importance of the stock-adjustment mechanism, the implementation of measures designed to reduce the pace of currency substitution would need to be sustained over a significant period if a meaningful reduction in individuals’ preferences for foreign money holdings is to be achieved. Finally, as regards economic and financial policies, it is clear from the results of the analysis that the extent of currency substitution in Egypt and the Y.A.R. is such that the existence of the phenomenon needs to be explicitly recognized in the formulation and implementation of such policies, particularly those in the monetary, fiscal, financial intermediation, and exchange rate areas.
REFERENCES
Batten, Dallas S., and R.W. Hafer, “Currency Substitution: A Test of Its Importance,” Review, Federal Reserve Bank of St. Louis (St. Louis), Vol. 66 (August-September 1984), pp. 5–11.
Batten, Dallas S., and R.W. Hafer, “Money, Income, and Currency Substitution: Evidence from Three Countries,” Review, Federal Reserve Bank of St. Louis (St. Louis), Vol. 67 (May 1985), pp. 27–35.
Batten, Dallas S., and R.W. Hafer, “Impact of International Factors on U.S. Inflation: An Empirical Test of the Currency Substitution Hypothesis,” Southern Economic Journal (Chapel Hill, North Carolina), Vol. 53 (October 1986), pp. 400–12.
Bordo, Michael D., and Ehsan U. Choudhuri, “Currency Substitution and the Demand for Money: Some Evidence for Canada,” Journal of Money, Credit and Banking (Columbus, Ohio), Vol. 14 (February 1982), pp. 48–57.
Brillembourg, Arturo, and Susan M. Schadler, “A Model of Currency Substitution in Exchange Rate Determination, 1973-78,” Staff Papers, International Monetary Fund (Washington), Vol. 26 (September 1979), pp. 513–42.
Canto, Victor A., “Monetary Policy, ‘Dollarization,’ and Parallel Market Exchange Rates: The Case of the Dominican Republic,” Journal of International Money and Finance (Guildford, England), Vol. 4 (December 1985), pp. 507–21.
Daniel, B.C., and H.O. Fried, “Currency Substitution, Postal Strikes, and Canadian Money Demand,” Canadian Journal of Economics (Toronto), Vol. 16 (November 1983), pp. 612–24.
Dodsworth, J., Mohamed A. El-Erian, and D. Hammann, “Foreign Currency Deposits in Developing Countries—Origins and Economic Implications,” IMF Working Paper WP/87/12 (Washington, March 1987).
El-Erian, Mohamed A., “Foreign Currency Deposits in LDCs,” Finance & Development (Washington), Vol. 24 (December 1987), pp. 38–40.
Fasano-Filho, Ugo, Currency Substitution and Liberalization: The Case of Argentina (Aldershot, England: Gower, 1986).
Fischer, Stanley, “Seigniorage and the Case for a National Money,” Journal of Political Economy (Chicago), Vol. 90 (April 1982), pp. 295–313.
Girton, Lance, and D. Roper, “Theory and Implications of Currency Substitution,” Journal of Money, Credit and Banking (Columbus, Ohio), Vol. 13 (February 1981), pp. 12–30.
Gruben, William C., and Patrick J. Lawler, “Currency Substitution: The Use of Dollar Coin and Currency in the Texas Border Areas of Mexico,” Economic Review, Federal Reserve Bank of Dallas (Dallas), July 1983, pp. 10–20.
Husted, Steven Leslie, “The Theory and Empirical Estimation of Currency Substitution” (Ph.D. dissertation; East Lansing: Michigan State University, 1980).
International Monetary Fund, International Financial Statistics (Washington, various issues).
Khan, Mohsin S., and C.L. Ramirez-Rojas, “Currency Substitution and Government Revenue from Inflation,” (unpublished; Washington: International Monetary Fund, September 1984).
Leite, Sergio Pereira, Cyrus Sassanpour, and Harry Snoek, “Yemen Arab Republic: Issues in the Transition from a Traditional to a Modern Financial System” (unpublished; Washington: International Monetary Fund, November 1986).
Marquez, J. (1985a), “Currency Substitution and Economic Monetary Aggregates: The U.S. Case,” Economics Letters (Amsterdam), Vol. 19 (No. 4), pp. 363–67.
Marquez, J. (1985b), “Money Demand in Open Economies: A Currency Substitution Model for Venezuela,” International Finance Discussion Paper 265 (Washington: Board of Governors of the Federal Reserve System, October 1985).
Melvin, M., “Currency Substitution and Western European Monetary Unification,” Economica (London), Vol. 52 (February 1985), pp. 79–91.
Miles, M.A., “Currency Substitution, Flexible Exchange Rates, and Monetary Independence,” American Economic Review (Nashville, Tennessee), Vol. 68 (June 1978), pp. 428–36.
Ortiz, Guillermo, “Currency Substitution in Mexico: The Dollarization Problem,” Journal of Money, Credit and Banking (Columbus, Ohio), Vol. 15 (May 1983), pp. 174–85.
Poloz, Stephen S., “Currency Substitution and the Precautionary Demand for Money,” Journal of International Money and Finance (Guildford, England), Vol. 5 (March 1986), pp. 115–24.
Ramirez-Rojas, C.L., “Currency Substitution in Argentina, Mexico, and Uruguay,” Staff Papers, International Monetary Fund (Washington), Vol. 32 (December 1985), pp. 629–67.
Tanzi, Vito, and Mario I. Blejer, “Inflation, Interest Rate Policy, and Currency Substitution in Developing Economies: A Discussion of Some Major Issues,” World Development (Oxford), Vol. 10 (September 1982), pp. 781–89.
Mr. El-Erian, an economist in the Middle Eastern Department, is a graduate of Cambridge University and received his doctorate from Oxford University.
Industrial country studies include Batten and Hafer (1984, 1985. and 1986), Bordo and Choudhuri (1982), Brillembourg and Schadler (1979), Daniel and Fried (1983), Marquez (1985a), and Melvin (1985).
Analyses of the experience of developing countries can be found in Canto (1985), Gruben and Lawler (1983), Fasano-Filho (1986), Marquez (1985b), Ortiz (1983), and Ramirez-Rojas (1985).
The paper draws on background work undertaken in preparing an earlier paper on foreign currency deposits in developing countries (Dodsworth, El-Erian, and Hammann (1987))
The type of currency substitution under consideration has been referred to in the literature as “nonsymmetrical” because residents’ demand for a “stronger” currency is not accompanied by a demand for the “weaker” currency by agents abroad.
See Khan and Ramirez-Rojas (1984) and Tanzi and Blejer (1982). The analysis in Fischer (1982) indicates that the seignorage rate in the Y.A.R. was among the highest in the world.
Analysis of the various aspects of this problem can be found in Batten and Hafer (1985), Brillembourg and Schadler (1979), and Fasano-Filho (1986).
This, and related issues, are discussed in Girton and Roper (1981), Husted (1980), and Miles (1978).
A discussion of these issues is contained in Dodsworth, El-Erian, and Hammann (1987) and in El-Erian (1987).
In this approach, both domestic and foreign currency balances enter the agents’ utility functions. Such an approach is used in Bordo and Choudhuri (1982).
See Brillembourg and Schadler (1979). where the choice of assets in agents’ portfolio decisions depends on the mean and variance of the returns.
In this approach, explicit account is taken of expenditure uncertainty as a motive for holding money. See Poloz (1986).
The a priori sign on Φ depends on the exact specification of institutional and political factors. For example, a positive relationship would be expected if the variable is specified in such a way that an increase denotes greater political uncertainty.
As of June 1986, these holdings amounted to US$3.0 billion for Egypt and US$360 million for the Y.A.R.
More specifically, the coefficients of correlation between developments in these two types of deposits were estimated at 0.91 and 0.90 for Egypt and the Y.A.R., respectively.
Around three quarters of the recorded money stock in the Y.A.R., for example, is held outside the banking system in the form of currency in the hands of the nonbanking public.
This argument is also valid for the other explanatory variables—that is, those capturing changes in institutional and political factors.
In local currency terms, the deposits grew at an annual rate of 42 percent during this period.
The two types of deposits grew by annual rates of 30 percent and 44 percent, respectively, when measured in local currency.
The growth rate of these deposits, in local currency terms, amounted to 42 percent annually.
Such an extension would require, among other things, a shift of emphasis toward longer-term institutional factors (for example, the pace oi monetization and spread of banking habits) as well as toward more general issues relating to the level and structure of overall economic and financial development (for example, changes in the “openness” of the economy).
Defined as the three-month LIBOR (London interbank offered rate) U.S. dollar rate relative to the domestic rate of interest on three-month local currency bank deposits. Movements in these rates were found to correspond closely to those in rates on other bank maturities.
Based on the estimated coefficient of determination, adjusted for degrees of freedom, reported in the table.
As proxied by a dummy variable specified to take into account the sudden increase in political uncertainty associated with the assassination of President Sadat in October 1981.
This specification differs in an important way from that adopted in a number of previous quantitative studies in this area. Previous studies have tended to take account only of the absolute level of the expected exchange rate, whereas the present study relates changes in currency substitution to the deviation of the expected rate from the relevant spot rate.
The use of the more limited ex change-rate-expectations proxy based on the absolute level of the rate, as was done in the study by Ramirez-Rojas (1985), yields an “elasticity” for Egypt similar to that estimated for Uruguay. The estimated elasticity for Egypt is also similar to that found by Ramirez-Rojas for Mexico in the specification in which the level of the future price for Mexican pesos was used as the proxy for exchange rate expectations.
Several other country studies appear to have omitted, a priori, the interest rate term in estimating the determinants of currency substitution.
In the case of the Y.A.R., the specification of the “institutional term” is such as to take account of the major banking and exchange system changes of 1983/84. As noted earlier, these changes effectively reduced the foreign ex-change resources for import purposes available through banking channels, there-by increasing the attractiveness of “own-resource finance” for transaction purposes. Details of the changes are contained in Leite, Sassanpour, and Snoek (1986).
In view of the relatively more rigid nature of the exchange system in the Y.A.R. in the period under consideration, expected exchange rate develop-ments were derived on the basis of the lagged value of the future formal rate relative to the spot formal rate.
Specifications based on the more limited proxy of exchange rate expectations yielded an elasticity of 2.0, which is similar to those estimated by Ramirez-Rojas (1985) for Argentina (which were in the range 1.5-2.1).
