The determinants of interest rate spreads in dollarized economies

ABSTRACT

Twenty years after dollarizing, interest rate spreads remain considerably high in El Salvador and even higher in Ecuador, despite decreased deposit rates. Using panels covering the 2005–2019 period, we explore the determinants of interest rate spreads in these two dollarized countries. We find that bank-specific characteristics, such as market share and liquidity, play a significant role at accounting for the differences in spreads. Dollarization is also expected to promote the entry of international banks due to the elimination of depreciation risk. We find that the low degree of foreign bank participation in Ecuador also contributes to explaining spread differences. These findings point out to complementary reforms to extract the full benefits of dollarization.

KEYWORDS:

• Dollarization
• interest rate spreads
• bank market power
• Ecuador
• El Salvador

JEL CLASSIFICATION:

• E40
• G21
• E42
• L10

Acknowledgment

We thank Paola Boel, Augusto de la Torre, Pablo Guerrón-Quintana, Rosa María Herrera, Yelena Tuzova, and an anonymous referee for their helpful comments

Disclosure statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Notes

¹ Philippon (2015) documents that spreads in the US have consistently averaged two percent for 130 years.

² See Demirgüç-Kunt and Huizinga (1999) for a panel of eighty countries, Brock and Rojas Suarez (2000) for seven Latin American countries, Saunders and Schumacher (2000) for six European countries plus the US or Maudos and Fernández de Guevara (2004) for five EU countries.

³ The argument is that dollarization removes the depreciation risk, eliminating the currency mismatch that afflicts banks in many developing economies. Moreover, the low inflation associated with dollarization should promote longer term investments, ameliorating maturity mismatches.

⁴ The data used to construct Figure 1 are not available for the pre-dollarization periods. However, (not entirely comparable) statistics indicate that interest rates were substantially higher than the post-dollarization values observed in both countries.

⁵ Table A1 (Appendix) shows results by bank size, which are consistent with our findings for the full sample.

⁶ In the Appendix we also show the results estimated using fixed-effects two-stage least squares instrumental variables techniques. As Entrop et al. (2015) point out, the Lerner index and non-interest income may be endogenous for reasons of reverse causality. We follow their approach and instrument those variables with their own first differences. Our benchmark results remain, for the most part, qualitatively unchanged.

⁷ Note that, on average, banks in both countries exhibit similar liquidity ratios (see Table 1). Thus, our results suggest that not only liquidity ratios matter, but also the type of liquid assets that banks hold.

⁸ That is, Net effect = ‘Market share’ coefficient $\times$ Average market share + ‘Large $\times$ Share’ coefficient $\times$ Average Large share. Thus, for Ecuador: 0.026 = $-0.419\times 0.041$ + $0.279\times 0.155$, and for El Salvador: $-0.005$ = $-0.647\times 0.078$ + $0.255\times 0.177$.

⁹ For 2005–2019, the average market shares of first-tier and non first-tier foreign banks were 11.2% and 7.3% in Ecuador, and 31.5% and 52.9% in El Salvador.