Abstract
1. This paper aims to shed light on the determinants of interest rates in the dollarized economies of Ecuador, El Salvador, and Panama, with particular focus on the former. Despite the adoption of the US dollar as legal tender, and hence, lack of independent monetary policy, there exists significant dispersion in observed interest rates among these economies. In terms of deposit rates, these three dollarized economies display evident variability, with Ecuador having the highest rate, particularly in recent years. For example, in 2020, the average deposit rate in Ecuador was 6.3 percent, while for El Salvador and Panama they were 3.8 and 1.9 percent, respectively. In addition, given the strong interest Ecuadorian authorities have shown in this subject, particularly on the reasons behind why businesses and consumers face high borrowing rates, this paper will attempt to investigate the determinants of the dispersion in deposit rates. These results should guide the design of policies that could help reduce market-determined interest rates.
Determinants of Deposit Rates in Fully Dollarized Economies: The Case of Ecuador1
A. Introduction
1. This paper aims to shed light on the determinants of interest rates in the dollarized economies of Ecuador, El Salvador, and Panama, with particular focus on the former. Despite the adoption of the US dollar as legal tender, and hence, lack of independent monetary policy, there exists significant dispersion in observed interest rates among these economies. In terms of deposit rates, these three dollarized economies display evident variability, with Ecuador having the highest rate, particularly in recent years. For example, in 2020, the average deposit rate in Ecuador was 6.3 percent, while for El Salvador and Panama they were 3.8 and 1.9 percent, respectively. In addition, given the strong interest Ecuadorian authorities have shown in this subject, particularly on the reasons behind why businesses and consumers face high borrowing rates, this paper will attempt to investigate the determinants of the dispersion in deposit rates. These results should guide the design of policies that could help reduce market-determined interest rates.
2. Regression-based results show that the EMBI yield and capital outflow tax explain the high deposit rates in Ecuador. Moreover, the key factors identified in economic theory and in previous empirical literature -namely the federal funds rate, bank reserves, and inflation- also affect deposit rates. Finally, in the case of Ecuador, there seem to exist additional country-specific structural factors that could have also contributed to high deposit rates.
Nominal Deposit Rates
In percent
Citation: IMF Staff Country Reports 2021, 229; 10.5089/9781513599274.002.A003
Source Haver1/ From November 2011-April 2014 the Board of Directors of the Central Bank of Ecuador decided not to publish the reference rate.Nominal Deposit Rates
In percent
Citation: IMF Staff Country Reports 2021, 229; 10.5089/9781513599274.002.A003
Source Haver1/ From November 2011-April 2014 the Board of Directors of the Central Bank of Ecuador decided not to publish the reference rate.Nominal Deposit Rates
In percent
Citation: IMF Staff Country Reports 2021, 229; 10.5089/9781513599274.002.A003
Source Haver1/ From November 2011-April 2014 the Board of Directors of the Central Bank of Ecuador decided not to publish the reference rate.3. Market competition for funds between lower segments of banks and upper segments of cooperatives seems to drive high deposit rates in Ecuador. Cooperatives, some of which are bigger than medium-sized banks, have been paying an average deposit rate of about 2.6 percentage points higher than private banks since 20162, while also gaining market share in terms of deposits, from about 24 percent of the market in January 2016 to about 34 percent of market in February 20213. There is no straightforward answer as to why cooperatives have been able to pay higher interest rates: this could be related to different business models, cooperativesβ status as not-for profit institutions, or the fact that cooperatives and banks face different regulatory frameworks. According to market participants, the main competition across the financial sector for depositors is between small- and medium-sized banks and larger cooperatives. In this competitive market, cooperativesβ ability to consistently pay higher interest rates could put pressure on banks to increase deposit rates to protect their source of funding.
Total Assets by Institution, December 2020
(In US$ million)
Citation: IMF Staff Country Reports 2021, 229; 10.5089/9781513599274.002.A003
Total Assets by Institution, December 2020
(In US$ million)
Citation: IMF Staff Country Reports 2021, 229; 10.5089/9781513599274.002.A003
Total Assets by Institution, December 2020
(In US$ million)
Citation: IMF Staff Country Reports 2021, 229; 10.5089/9781513599274.002.A003
4. The rest of the paper is organized as follows: Section B discusses the empirical strategy, data, and regression results; Section C focuses on the analysis of the competition for funds between cooperatives and banks in Ecuador; and Section D concludes.
B. Fundamentals of Deposit Rates
Empirical Strategy and Data
5. We use the extended Monti-Klein model of a monopolistic banking system to derive the empirical model (Dermine, 1986). The details of the theoretical and empirical model are discussed in Druck et al. (forthcoming), Since the analysis will include cross country panel data, the model for estimation is given by (1):
where X is a vector of control variables, ΞΌi are country fixed effects (given by country dummies), and Οit is the idiosyncratic error. Equation (1) will be estimated using pooled OLS regressions. The analysis employs a two-tier approach; the first will focus on a cross-country panel regression comparing Ecuador, El Salvador, and Panama to the United States, while the second focuses on Ecuador specific time-series data.
6. The panel regression analysis uses term-deposit rate data published by the Central Bank of Ecuador (BCE). This rate is the weighted average term-deposit rate of three financial entities: private banks, cooperatives, and mutuals. The other three countries in our sample do not report a similar weighted average term-deposit rate and, due to lack of data, it was not possible to construct one. Therefore, a simple average across term-deposits is used for El Salvador, Panama, and the United States. All the data is in monthly frequency and ranges from 2001β2020, except for the Ecuador term-deposit rates from November 2011-April 2014, when deposit rates remained fixed,4 data during this period was dropped. The BCE-reported rate included two additional financial entities-cooperatives and mutuals- that were not included in the other three countries. To make a deeper analysis of deposit rates paid by banks versus those paid by cooperatives we built another database based on monthly data, starting in 2016, of the deposit rates for new deposits, on a month-by-month basis, for institutions in the private bank, cooperative, and mutual sectors. Using this data, a weighted average deposit rate was calculated for private banks, cooperatives, mutuals, and the total financial system (TFS). The description and source of independent variables is discussed in detail in Druck et al. (forthcoming).
Results of Regression
7. The EMBI yield is a significant determinant of deposit rates (Table 1). When fixed country effects are considered in a pooled regression, on average a percentage point rise in EMBI yield can lead to about 0.2β0.3 percentage points increase in domestic deposit rates, while the Ecuador-specific result is about 0.1 percentage points. In time-series regression for Ecuador, the impact of EMBI yield is even higher, at around 0.4 percent. The regression results also demonstrate that a one percentage point increase in the capital outflow tax, only relevant for Ecuador, has a similar magnitude of impact to deposits as EMBI yield from the pooled regression. Moreover, the relatively large coefficient in the dummy for Ecuador indicates substantial domestic idiosyncratic factors that explain high deposit rates. In line with the theoretical model and other relevant empirical literature, we find statistically significant impacts from the federal funds rate, inflation, and bank reserves to deposits.
8. Tightening the EMBI spread through fiscal consolidation and gradual removal of the capital outflow tax would help reduce bank funding costs in Ecuador. While fiscal consolidation will benefit the country by ensuring fiscal sustainability and regaining access to private credit markets, it also brings the additional benefit of reducing funding costs for financial intermediaries, as it will compress the EMBI spread and reduce the country risk premium. Ecuador should also consider removing the capital outflow tax, to make the country attractive to capital inflows. However, this should be done in a gradual way, by monitoring the impact of relaxation on capital inflows and outflows, as well as the negative impact on fiscal revenues.
Ecuador: OLS Regressions of Deposit Rates
Capital outflow tax only exists in Ecuador since January 2008. For other years and countries, the value is zero. p-values in parentheses * p < 0.10, ** p < 0.05, *** p < 0.01.
Ecuador: OLS Regressions of Deposit Rates
| Dollarized panel | Ecuador time series | |||||
|---|---|---|---|---|---|---|
| (ECU, ELS, PAN, US, 2001β2020) | (2016β2019) | |||||
| Public data | Public data | Public data | Total calc. | Bank calc. | Coop. calc. | |
| -1.000 | -2.000 | -3.000 | -4.000 | -5.000 | -6.000 | |
| Economic activity growth | 0.114 -0.917 |
0.128 -0.890 |
0.370 -0.688 |
0.428 -0.688 |
0.732 -0.596 |
0.339 -0.432 |
| Fed funds rate | 0.494*** 0.000 |
0.366*** 0.000 |
0.381*** 0.000 |
0.315** -0.013 |
0.381** -0.021 |
0.107* -0.086 |
| Inflation | 0.036*** -0.002 |
0.010 -0.399 |
0.017 -0.121 |
0.350*** -0.001 |
0.490*** 0.000 |
0.233*** 0.000 |
| Bank reserves to deposits | 0.011*** 0.000 |
-0.088*** 0.000 |
-0.083*** 0.000 |
-0.138*** -0.001 |
-0.209*** 0.000 |
-0.034 -0.124 |
| Gross international reserves to imports | 0.001*** 0.000 |
0.001*** 0.000 |
0.001*** 0.000 |
0.002* -0.050 |
0.003* -0.053 |
0.001** -0.035 |
| Capital outflow tax1 | 0.281*** 0.000 |
0.251*** 0.000 |
0.242*** 0.000 |
|||
| EMBI yield | 0.335*** 0.000 |
0.195*** 0.000 |
0.272*** 0.000 |
0.458*** -0.002 |
0.617*** -0.002 |
0.238*** 0.000 |
| PAN dummy | -0.121 -0.146 |
-0.253*** -0.005 |
||||
| ELS dummy | 2.697*** 0.000 |
2.314*** 0.000 |
||||
| ECU dummy | 1.382*** 0.000 |
2.191*** 0.000 |
||||
| ECU dummy * EMBI | -0.172*** 0.000 | |||||
| EMBI square | -0.021*** -0.003 |
-0.028*** -0.003 |
-0.011*** 0.000 |
|||
| Observations | 726.000 | 726.000 | 726.000 | 48.000 | 48.000 | 48.000 |
| Adjusted R2 | 0.830 | 0.884 | 0.889 | 0.647 | 0.680 | 0.710 |
Capital outflow tax only exists in Ecuador since January 2008. For other years and countries, the value is zero. p-values in parentheses * p < 0.10, ** p < 0.05, *** p < 0.01.
Ecuador: OLS Regressions of Deposit Rates
| Dollarized panel | Ecuador time series | |||||
|---|---|---|---|---|---|---|
| (ECU, ELS, PAN, US, 2001β2020) | (2016β2019) | |||||
| Public data | Public data | Public data | Total calc. | Bank calc. | Coop. calc. | |
| -1.000 | -2.000 | -3.000 | -4.000 | -5.000 | -6.000 | |
| Economic activity growth | 0.114 -0.917 |
0.128 -0.890 |
0.370 -0.688 |
0.428 -0.688 |
0.732 -0.596 |
0.339 -0.432 |
| Fed funds rate | 0.494*** 0.000 |
0.366*** 0.000 |
0.381*** 0.000 |
0.315** -0.013 |
0.381** -0.021 |
0.107* -0.086 |
| Inflation | 0.036*** -0.002 |
0.010 -0.399 |
0.017 -0.121 |
0.350*** -0.001 |
0.490*** 0.000 |
0.233*** 0.000 |
| Bank reserves to deposits | 0.011*** 0.000 |
-0.088*** 0.000 |
-0.083*** 0.000 |
-0.138*** -0.001 |
-0.209*** 0.000 |
-0.034 -0.124 |
| Gross international reserves to imports | 0.001*** 0.000 |
0.001*** 0.000 |
0.001*** 0.000 |
0.002* -0.050 |
0.003* -0.053 |
0.001** -0.035 |
| Capital outflow tax1 | 0.281*** 0.000 |
0.251*** 0.000 |
0.242*** 0.000 |
|||
| EMBI yield | 0.335*** 0.000 |
0.195*** 0.000 |
0.272*** 0.000 |
0.458*** -0.002 |
0.617*** -0.002 |
0.238*** 0.000 |
| PAN dummy | -0.121 -0.146 |
-0.253*** -0.005 |
||||
| ELS dummy | 2.697*** 0.000 |
2.314*** 0.000 |
||||
| ECU dummy | 1.382*** 0.000 |
2.191*** 0.000 |
||||
| ECU dummy * EMBI | -0.172*** 0.000 | |||||
| EMBI square | -0.021*** -0.003 |
-0.028*** -0.003 |
-0.011*** 0.000 |
|||
| Observations | 726.000 | 726.000 | 726.000 | 48.000 | 48.000 | 48.000 |
| Adjusted R2 | 0.830 | 0.884 | 0.889 | 0.647 | 0.680 | 0.710 |
Capital outflow tax only exists in Ecuador since January 2008. For other years and countries, the value is zero. p-values in parentheses * p < 0.10, ** p < 0.05, *** p < 0.01.
C. Deposit Interest Rates and Market Share
9. The Ecuador financial sector is comprised mainly of private banks and cooperatives. As of July 2021, measured by deposits, banks represented about 72 percent of the financial system, while the cooperative and mutuals sectors represented about 28 percent. Ecuador has two regulatory institutions for the financial system, the Superintendency of Banks (SB) that oversees private and public banks and the Superintendency of Popular and Solidarity Economy (SEPS) that oversees cooperatives and mutuals. In this section we will refer to these two sectors as SB and SEPS.5
Cooperative and Private Bank Term Deposit Rates
In percent
Citation: IMF Staff Country Reports 2021, 229; 10.5089/9781513599274.002.A003
Source; BCECooperative and Private Bank Term Deposit Rates
In percent
Citation: IMF Staff Country Reports 2021, 229; 10.5089/9781513599274.002.A003
Source; BCECooperative and Private Bank Term Deposit Rates
In percent
Citation: IMF Staff Country Reports 2021, 229; 10.5089/9781513599274.002.A003
Source; BCE10. Deposit rates in cooperatives have been higher than deposit rates for banks. Based on data availability (Jan 2016-Feb 2021), cooperatives have been paying higher deposit rates than banks, an average of about 2.6 percent higher during the period. Also, the difference in deposit rates between these two groups of financial institutions widened during 2017β18 and has been widening again since late 2020. These two episodes coincide with increases in market liquidity; during 2017β18 the central bank injected liquidity into the economy while in the latest episode the accumulated liquidity is a side effect of the pandemic. The different behavior between these institutions suggest that banks and cooperatives react differently to liquidity shocks (text chart). Also, data on monthly average deposit rates paid by each financial institution shows that most institutions overseen by SEPS have been paying consistently higher interest rates than banks (text chart).
Term Deposit Rate by Institution
In percent
Citation: IMF Staff Country Reports 2021, 229; 10.5089/9781513599274.002.A003
Source BCETerm Deposit Rate by Institution
In percent
Citation: IMF Staff Country Reports 2021, 229; 10.5089/9781513599274.002.A003
Source BCETerm Deposit Rate by Institution
In percent
Citation: IMF Staff Country Reports 2021, 229; 10.5089/9781513599274.002.A003
Source BCE
Share of Term Deposits by System: All Term Deposits
In percent
Citation: IMF Staff Country Reports 2021, 229; 10.5089/9781513599274.002.A003
Source:BCEShare of Term Deposits by System: All Term Deposits
In percent
Citation: IMF Staff Country Reports 2021, 229; 10.5089/9781513599274.002.A003
Source:BCEShare of Term Deposits by System: All Term Deposits
In percent
Citation: IMF Staff Country Reports 2021, 229; 10.5089/9781513599274.002.A003
Source:BCE11. Market share. Measured from the deposit side, and using data only on new deposits done in the respective month, the market share of cooperative has increased to about 34 percent in February 2021, up from about 24 percent in January 2016 This is an average increase of about 1.6 percent of the market share per year.
12. Loans. While banks serve all types of customers, cooperatives and mutuals serve mostly to microcredit and consumer loans (panel chart).
13. Regulatory framework. The regulatory frameworks are different between the SB and the SEPS. As the following table shows, the regulatory framework for the SPES is more lenient that the regulatory framework for the SB.
Ecuador: Credit by Economic Segment 1/ 2/
Citation: IMF Staff Country Reports 2021, 229; 10.5089/9781513599274.002.A003
1/ BP: Private Banks; CO: Cooperatives; MU: Mutuals; TFS: Total Financial System.2/ Data for November-December 2017 is not available.Ecuador: Credit by Economic Segment 1/ 2/
Citation: IMF Staff Country Reports 2021, 229; 10.5089/9781513599274.002.A003
1/ BP: Private Banks; CO: Cooperatives; MU: Mutuals; TFS: Total Financial System.2/ Data for November-December 2017 is not available.Ecuador: Credit by Economic Segment 1/ 2/
Citation: IMF Staff Country Reports 2021, 229; 10.5089/9781513599274.002.A003
1/ BP: Private Banks; CO: Cooperatives; MU: Mutuals; TFS: Total Financial System.2/ Data for November-December 2017 is not available.14. While more research is needed to understand the reasons behind these developments, closing regulatory gaps for all deposit takers would ensure resilience and a level playing field. The regression-based model suggests that an additional factor explaining high deposit rates could be some structural idiosyncratic features. Several hypotheses could be in play, such as (i) different regulatory frameworks set up by the SB and the SEPS; or (ii) since cooperatives are not-for-profit maximizers, maybe they can afford to pay higher deposit rates. Nevertheless, the cooperative sector has gained substantial market share in recent years and has been growing rapidly, which some market participants attribute to them offering higher deposit rates compared to commercial banks. The higher interest rates offered by this sector and its rapid growth require the authorities to closely monitor these developments to ensure that these institutions have adequate risk management practices. As cooperatives gain systemic importance in the Ecuadorian financial sector, closing regulatory gaps for all deposit takers is necessary.
Ecuador: Comparison of Regulations
Using simple average of accounts included in the regulations.
During the period of the study. Currently, a measure to deal with Pandemic was for the SB and SEPS to extend to 60 days the reclassification to NPL.
Ecuador: Comparison of Regulations
| SB | SEPS | ||
|---|---|---|---|
| Liquidity requirements | Computation for liquidity requirement 1/ | Sight deposit at 25 percent, term deposits at 9 5 percent | Sight deposit at 3.75 percent, term deposits at 6.6 percent |
| Invest at least 60 percent domestically | Imposes obligations to buy domestic financial assets. | Mo obligation to invest in domestically financial assets. With deposit in central bank fulfill this obligation | |
| Reserve Requirements | 5% for banks with more than USS 1 billion in assets; and 2% for banks with less than US$1 billion in assets. | No regulations for SEPS on reserve requirements. | |
| Liquidity Fund | 5% of deposits (before pandemic 8%). As of June 2021 liquidity fund assets are US$2379 million. | 2% of deposits (2.5% before pandemic). As of May 2021 liquidity fund assets are US$354 million. | |
| Deposit insurance | 0 6%, plus a risk premium between 0.01β0.05%. | 0.65% for cooperatives segments 1 and 2; 0.25% segment 3; 0.12% segment 4: and 0.05% -βor segment 5. | |
| Non-Performing Loans (NPL) 2/ | 16 days after missed payments | 35 days after missed payments | |
| Provisions | Lower provisioning for SEPS |  |
|
| Capital to risk-weighted assets ratio | 9%, for a number of assets weights for the calculation of this ratio are higher for SB than for SEPS. | 9% for segments 1 and 2, and lower for the rest. | |
Using simple average of accounts included in the regulations.
During the period of the study. Currently, a measure to deal with Pandemic was for the SB and SEPS to extend to 60 days the reclassification to NPL.
Ecuador: Comparison of Regulations
| SB | SEPS | ||
|---|---|---|---|
| Liquidity requirements | Computation for liquidity requirement 1/ | Sight deposit at 25 percent, term deposits at 9 5 percent | Sight deposit at 3.75 percent, term deposits at 6.6 percent |
| Invest at least 60 percent domestically | Imposes obligations to buy domestic financial assets. | Mo obligation to invest in domestically financial assets. With deposit in central bank fulfill this obligation | |
| Reserve Requirements | 5% for banks with more than USS 1 billion in assets; and 2% for banks with less than US$1 billion in assets. | No regulations for SEPS on reserve requirements. | |
| Liquidity Fund | 5% of deposits (before pandemic 8%). As of June 2021 liquidity fund assets are US$2379 million. | 2% of deposits (2.5% before pandemic). As of May 2021 liquidity fund assets are US$354 million. | |
| Deposit insurance | 0 6%, plus a risk premium between 0.01β0.05%. | 0.65% for cooperatives segments 1 and 2; 0.25% segment 3; 0.12% segment 4: and 0.05% -βor segment 5. | |
| Non-Performing Loans (NPL) 2/ | 16 days after missed payments | 35 days after missed payments | |
| Provisions | Lower provisioning for SEPS | |
|
| Capital to risk-weighted assets ratio | 9%, for a number of assets weights for the calculation of this ratio are higher for SB than for SEPS. | 9% for segments 1 and 2, and lower for the rest. | |
Using simple average of accounts included in the regulations.
During the period of the study. Currently, a measure to deal with Pandemic was for the SB and SEPS to extend to 60 days the reclassification to NPL.
D. Conclusion
15. The assessment conducted to shed light on the determinants of interest rates suggest some policy measures that could help to reduce interest rates: As mentioned earlier, the interest rate for Ecuador government bonds and the tax on transfer abroad were found to be determinant factors for deposit interest rates. Hence, a stronger fiscal position that reduces the cost of funding for the government could help reduce deposit rates, as well a reduction in the tax on transfers abroad. Regarding the structural issues, it seems that the competition for funds between cooperatives and banks could be pushing up deposit rates. A comprehensive assessment of this market is needed to understand the reason behind this. Nevertheless, as cooperatives gain systemic importance in the Ecuadorian financial sector, closing regulatory gaps for all deposit takers would ensure resilience and a level playing field.
References
Dermine, Jean, 1986, Deposit rates, credit rates, and bank capital: the Klein-Monti model revisited Journal of Banking and Finance, 10 (1986), pp. 99β114, 10.1016/0378-4266(86)90022-1
Druck, P., Baltabaev, B., and Burgara, I., forthcoming, Determinants of deposit rates in fully dollarized economies: the case of Ecuador, IMF Working Paper.
Prepared by Pablo Druck (MCM), Botir Baltabaev, Juan Pablo Erraez, and Ivan Burgara (all WHD).
For this paper Banco Pacifico in included as a private bank.
Using data for new deposits of private banks, as defined by the Superintendency of Banks, and Cooperatives-which primarily includes Segments 1 and 2.
During this period the Board of Directors of the Central Bank of Ecuador decided not to publish the reference rate.
This section uses only data classified by SB as private banks and Cooperatives of segments 1 and 2, and for market share measures only consider the information incorporated in the sample as described in the section of data.
