Journal Issue
Share
Article

Panama: Selected Issues

Author(s):
International Monetary Fund. Western Hemisphere Dept.
Published Date:
August 2015
Share
  • ShareShare
Show Summary Details

Interest Rates in Panama: U.S. Pass-Through and ITS Effects on Local Economic Activity1

A. Executive Summary

1. Interest rates in Panama track closely U.S. interest rates, but with a pass-through lower than expected for a dollarized economy. Furthermore, lending rates exhibit lower sensitivity than deposit rates, making intermediation margins respond negatively to movements in U.S. rates.

2. The effect of U.S. interest-rate normalization is likely to be cushioned through lower intermediation margins, without significantly affecting banks’ overall profitability or economic growth. The expected increase in U.S. interest rates is likely to compress intermediation margins in Panama. However, this would not affect banks’ profitability, as the returns on liquid assets would increase. The limited real effects of U.S. interest-rate shocks are consistent with the results from an econometric exercise that incorporates real economic activity.

B. U.S. Interest-Rate Pass-Through

3. Short-term deposits represent the bulk of the deposit base, while lending is mainly concentrated in the commercial sector. Almost three-quarters of the deposit base of Panamanian banks correspond to deposits of up to one month maturity (Chart 1). In terms of retail credit exposures, about half of recently issued loans correspond to the commercial sector. The analysis in this Annex therefore focuses on the one-month deposit rate, and the one-year commercial lending rate2.

Chart 1:Panama SBN:5 Deposit and Lending Structure

(As percent of GDP)

Source: Superintendencia de Bancos de Panama and staff calculations.

4. While Panamanian rates track closely U.S. interest rates, the pass-through is lower than expected for a dollarized economy. Regression analyses between Panamanian interest rates and U.S. rates of the same maturity show that the pass-through from U.S. rates is substantially lower than 1, typically in the range 0.5–0.8 for deposit rates and between 0.1-0.3 for lending rates (Text Table 1; see also Chart 2).3 A possible explanation for the low pass-through is that Panamanian banks need to compete for deposits in a context of high credit growth, and therefore reductions in U.S. rates are not fully reflected in domestic deposit rates.4 Furthermore, while banks may be able to get cheaper financing from foreign wholesale sources in the short run, they have traditionally leaned towards domestic deposits as a more stable way of financing. Table 1 also shows that both deposit and lending rates present a higher pass-through from the U.S. 5-year government bond yield, which may suggest that investors arbitrage Panamanian assets with U.S. assets of longer (and thus riskier) maturity.

Text Table 1:Panama Interest Rate Regression
Deposit 1MImplicit deposit rate 1\Commercial lending 1Y
LocalForeignLocalForeign
U.S. 1M deposit0.5379***0.5110***0.0726***
(0.0000)(0.0000)(0.0000)
U.S. prime lending0.1576***0.0642**
(0.0000)(0.0164)
U.S. 1M gov. yield0.4908***0.4665***
(0.0000)(0.0000)
U.S. 1Y gov. yield0.1732***0.0482
(0.0000)(0.1003)
U.S. 5Y gov. yield0.8168***0.7544***0.3003***0.0418
(0.0000)(0.0000)(0.0000)(0.4003)
VIX0.0426***0.0472***0.0426***0.0189**0.0244***0.0188**0.0020***0.0228***0.0233***0.0204***−0.0162***−0.0165***−0.0176***
(0.0000)(0.0000)(0.0000)(0.0140)(0.0000)(0.0198)(0.0056)(0.0000)(0.0000)(0.0000)(0.0008)(0.0005)(0.0002)
Observations170157174170157174133151151151151151151
Adjusted R-squared0.77230.66340.76560.77580.67140.70480.91620.51450.53310.65540.03810.02680.0197

Constructed from income statement and balance sheet of banking system. Interests paid in month t divided be deposits in month t-1. Regressions control for domestic inflation and economic activity Robust pval in parentheses.

*** p<0.01, ** p<0.05, * p<0.1

Constructed from income statement and balance sheet of banking system. Interests paid in month t divided be deposits in month t-1. Regressions control for domestic inflation and economic activity Robust pval in parentheses.

*** p<0.01, ** p<0.05, * p<0.1

Chart 2:Panama Rates versus U.S. Rates

Sources: SBP and Bloomberg.

Note: Implicit deposit rate calculated as the ratio of deposit interest income in month t to total deposits in month t-1.

5. The decline in deposit rates in the face of more steady lending rates has opened a relatively large intermediation spread. As U.S. interest rates declined from their pre-financial crisis levels, the lower pass-through to Panamanian banks’ lending rates vis-à-vis deposit rates has opened up a large bank spread of about 6 percent.

6. However, banks’ returns have actually been compressing, as larger returns on intermediation only partially offset lower returns on liquid assets. Indeed, net income on average assets of the banking sector has fallen from 2.3 percent in 2004–08 to 1.5 percent in 2009-14 (see also Chart 3). In the absence of a lender of last resort, Panamanian banks maintain ample levels of liquidity. This implies that since the onset of the financial crisis banks’ interest income has been significantly affected by the low returns on its liquid positions. Hence, the widened intermediation spreads partially compensate for the negative impact of low U.S. rates on banks’ profitability (Chart 4). Note that, despite the large interest-rate spreads, there are no clear signs of lack of competition in the banking sector.

Chart 3:Returns on Assets (4-quarter MA)

Sources: SBP and staff calculations.

Chart 4:Returns on Deposits and on Intermediation

Sources: SBP and staff calculations.

7. The effect of U.S. interest-rate normalization is likely to be cushioned through lower spreads, without affecting banks’ overall profitability. In light of the previous results, it is not surprising that the bank lending-deposit spreads in Panama are negatively related to U.S. interest rates. Estimates indicate that a 100 basis points increase in U.S. interest rates causes a compression of spreads of about 25-40 basis points (see Text Table 2). Hence, as interest rates in the U.S. normalize, the response of lending rates is expected to be more subdued than the one for deposit rates. The effect on economic activity would be softened by a compression in spreads. The lower spreads would not be at the expense of banks’ profits, as the returns on liquid assets should increase at the same time.

Text Table 2:Panama Intermediation Spreads
Spread: 1Y commercial lending rate - 1M deposit rate
LocalForeign
U.S. 1M deposit−0.2515***−0.3549***
(0.0000)(0.0000)
U.S. prime lending−0.2779***−0.3665***
(0.0000)(0.0000)
U.S. 1M gov. yield−0.2959***−0.3873***
(0.0000)(0.0000)
U.S. 1Y gov. yield−0.2632***−0.3853***
(0.0000)(0.0000)
U.S. 5Y gov. yield−0.3087***−0.5099***
(0.0000)(0.0000)
VIX−0.0084−0.0121**−0.0146***−0.0120*−0.0068−0.0297***−0.0344***−0.0376***−0.0352***−0.0279***
(0.1877)(0.0352)(0.0095)(0.0505)(0.2608)(0.0001)(0.0000)(0.0000)(0.0000)(0.0001)
Observations151151151151151151151151151151
Adjusted R-squared0.41450.45360.45190.37790.29600.42010.40220.39490.41250.4119
Regressions control for domestic inflation and economic activity Robust pval in parentheses.*** p<0.01, ** p<0.05, * p<0.1
Regressions control for domestic inflation and economic activity Robust pval in parentheses.*** p<0.01, ** p<0.05, * p<0.1

C. Effects on the Real Economy

8. The limited real effects of U.S. interest-rate shocks are consistent with the results from an econometric exercise. To investigate the spillovers from U.S. interest rates, we estimate a VAR including U.S. industrial production growth, the U.S. 5-year government bond yield, a Panamanian interest rate (local 1-month deposit or 1-year commercial lending), growth of credit to the private sector in Panama, and growth of the monthly economic activity indicator of the Panamanian economy (IMAE).6 The data are monthly, and identification through the above (Cholesky) ordering, with U.S. variables being an exogenous block.7 U.S. interest rate shocks have a significant and persistent effect on Panamanian interest rates, which peaks between one and two years after the shock (Chart 5). Despite this increase in rates, domestic economic activity is not affected. Furthermore, note that while the U.S. 5Y rate peaks at about 0.26 percentage points, the response of Panamanian rates is of at most 0.12 percent for deposits and 0.04 for lending, which is broadly consistent with the results in Text Table 1.

Chart 5:Response to U.S. Interest Rate Shocks

Notes: Model on the left column includes 1-month deposit; model on right column includes 1-year commercial lending. Each row displays the response of U.S. 5-year yield, Panamanian interest rate, and IMAE, respectively, to one-standard deviation shocks in the U.S. 5-year yield.

D. Concluding Remarks

9. Interest rates in Panama track closely U.S. interest rates, but with a pass-through lower than expected for a dollarized economy. Furthermore, lending rates exhibit lower sensitivity than deposit rates, making intermediation margins respond negatively to movements in U.S. rates.

10. The effect of U.S. interest-rate normalization is likely to be cushioned through lower intermediation margins, without affecting banks’ overall profitability. The expected increase in U.S. interest rates is likely to compress intermediation margins in Panama. However, this would not affect banks’ profitability, as the returns on liquid assets would increase. A VAR analysis that incorporates real economic activity suggests that movements in U.S. rates do not have large effects on the Panamanian economy.

Reference
Appendix I. Asymmetric Response to U.S. Rates
Table 1:Separating Positive and Negative U.S. Rate Changes
Deposit 1MImplicit deposit rate 1\Commercial lending 1Y
LocalForeignLocalForeign
U.S. 1M deposit (+)0.4482***0.4681***0.0675***
(0.0375)(0.0345)(0.0025)
U.S. 1M deposit (−)0.6343***0.5570***0.0813***
(0.0343)(0.0244)(0.0030)
U.S. prime lending (+)0.1301***0.0652**
(0.0192)(0.0285)
U.S. prime lending (-)0.2061***0.0625
(0.0235)(0.0392)
U.S. 1M gov. yield (+)0.4222***0.4464***
(0.0332)(0.0255)
U.S. 1M gov. yield (-)0.6044***0.4999***
(0.0439)(0.0286)
U.S. 1Y gov. yield (+)0.1580***0.0189
(0.0217)(0.0327)
U.S. 1Y gov. yield (-)0.1980***0.0960**
(0.0322)(0.0408)
U.S. 5Y gov. yield (+)0.7551***0.7058***0.2857***0.0131
(0.0467)(0.0430)(0.0292)(0.0528)
U.S. 5Y gov. yield (-)0.8477***0.7788***0.3125***0.0657
(0.0379)(0.0413)(0.0321)(0.0551)
VIX0.0375***0.0429***0.0385***0.0165**0.0231***0.0156*0.0018***0.0180***0.0219***0.0193***−0.0160***−0.0194***−0.0198***
(0.0081)(0.0069)(0.0100)(0.0072)(0.0054)(0.0082)(0.0006)(0.0033)(0.0035)(0.0032)(0.0053)(0.0050)(0.0049)
Observations170157174170157174133151151151151151151
Adjusted R-squared0.80460.69290.77450.78580.67340.71230.93050.54550.53470.65690.03150.03670.0266

Constructed from income statement and balance sheet of banking system. Interests paid in month t divided be deposits in month t-1. Regressions control for domestic inflation and economic activity Robust pval in parentheses.

*** p<0.01, ** p<0.05, * p<0.1

Constructed from income statement and balance sheet of banking system. Interests paid in month t divided be deposits in month t-1. Regressions control for domestic inflation and economic activity Robust pval in parentheses.

*** p<0.01, ** p<0.05, * p<0.1

Prepared by D. Cerdeiro and W. Shi.

There are no data available on the maturity profile of loans by sector. The interest rates reported by the Superintendency of Banks (SBP) correspond to new deposits and loans (i.e. they are marginal rates), and are the simple (i.e. unweighted) average across banks and customers.

The relatively low pass-through from U.S. rates has already been pointed out by Swiston (2011), who also presents estimates for other dollarized economies in the region (Ecuador and El Salvador). Table 3 in the Appendix presents regression results that allow for different responses to U.S. interest rate increases and decreases. The pass-through is typically higher for U.S. rate increases, but the largest pass-through (at 0.85) is only slightly larger than the largest one found the regressions of Table 1.

A closer inspection of Figure 2 also reveals that the relation between Panamanian and U.S. rates used to be stronger in the past. Since the handover of the Canal in 1999 the Panamanian economy has become generally less dependent (and thus also less synchronized) with the U.S. cycle. The pairwise correlation in quarter-on-quarter real GDP growth rates fell from 0.4 during 1996–2001, to 0.2 during 2002–2013.

SBN stands for the national bank system of Panama which excludes the offshore banks, i.e. banks that hold an international license and are prohibited by law to involve in domestic business. In terms of total deposits, the size of the offshore banks is approximately one-fifth of the onshore banks.

The U.S. 5-year yield is used since it showed the highest explanatory power for Panama’s deposit and lending rates in our previous regression analyses. Refer to Table 1.

The Cholesky ordering within the Panamanian block is justified under the following identifying assumptions. Bank officials decide on a rate without taking into account total credit being given and economic activity within the month, but responding to foreign variables. Credit is affected by interest rates but not by economic activity in that same month, as firms need to plan at least one month ahead when they want to take a loan.

Other Resources Citing This Publication