Blog Name

Saving in US dollars doesn't yield highest returns in inflation targeting countries

Date

Author's note: I thank Joseph E. Gagnon for comments.

Body

For more than 30 years, dozens of countries around the world have experienced painful episodes of rampaging inflation. Many have adopted a monetary policy known as inflation targeting (IT) requiring them to ramp up interest rates that have decelerated inflation at the cost of sometimes drastic economic slowdowns. Separately, policymakers have generally accompanied inflation targeting with often large fluctuations in exchange rates as they have had to choose between stabilizing inflation or exchange rates. This volatility has compelled savers in some IT countries to invest their savings in US dollars or other major currencies to avoid losing value in case their home currency weakens.

But a question remains: Have savers in countries with inflation targeting and floating exchange rates done better or worse investing in their home currencies rather than in US dollars? This analysis finds evidence that savings in national currencies have fetched roughly the same or even better returns than investments in US dollars. Japan is the only exception owing to negative interest rates there.

According to the International Monetary Fund (IMF), 41 countries around the world currently practice inflation targeting.[1] Under this framework, central banks use interest rates to stabilize inflation around a publicly announced goal—an inflation target. If inflation is too high, central banks tighten monetary conditions by raising interest rates; if it is too low, they ease them by lowering interest rates. Previously, many of these countries targeted a fixed exchange rate. Although a fixed exchange rate provides some restraint on inflation, it limits a central bank's capacity to react to economic shocks that affect employment and inflation. Plus, fixed exchange rates are often subject to speculative attacks.

The increased volatility and uncertainty surrounding the exchange rate is a new development in some IT countries. Given this and lingering expectations of incremental or sudden future depreciations of the national currency, savers may be tempted to put their savings into US dollars because the United States is the world’s largest economy and the dollar is the most widely used currency. This phenomenon, called “deposit dollarization,” has been observed in many developing economies.[2] It has most recently plagued Turkey, where the share of the country’s deposits held in foreign currencies increased from 30 percent in 2012 to 47 percent as of the end of last year.

The table below sheds some light on whether the above perception has borne out to be true in practice. It presents several metrics for 39 inflation targeters that demonstrate the tradeoff between saving in national currencies and the US dollar. Specifically, the table displays an average annual interest rate realized on domestic short-term treasury bills starting from the first year immediately following adoption of inflation targeting until the end of 2020.[3] It then converts this average yield into US dollars. The next column shows an average rate of return earned on US treasury bills for the same period.[4] The final column calculates the difference between the previous two numbers, which provides a measure of how much more or less savers in local currencies made compared to investors in US dollars every year. This comparison can be performed using any type of assets, but the analysis presented here focuses on treasury securities because they are safe and in most cases are readily available to investors at home and abroad.[5] All averages in the table are annual compound rates.

As can be seen from the table, local currency savers collected nearly as much or more interest income[6] as savers in US dollars in all IT economies, except for Japan.[7] Japan appears to be an outlier, as it is the only country in the table whose annual average interest rate in yen was negative for the entire duration of an IT regime. This anomaly stems from domestic deflationary pressures that have bogged Japan down for many years. It is, therefore, safe to conclude that floating exchange rates under an IT regime have not made investors in national currencies worse off compared to those who chose to save solely in US dollars.

Domestic currency investors in some IT countries enjoyed a rate of return much higher than that available to US dollar savers. For instance, local savers in Brazil, Iceland, Russia, Uganda, and Ukraine reaped the rewards of annual interest rates that were more than 5 percent higher than US dollar rates, while investors in Armenia, Kazakhstan, Moldova, and Uruguay saw their fortunes rise more than 4 percent faster than similar investments in US dollars. In fact, investors in only Canada, Chile, the Dominican Republic, Georgia, Mexico, Sweden, and the United Kingdom witnessed average domestic interest rates that were less than 1 percent higher than US dollar rates. Investors in all other IT economies realized a higher return, despite fluctuations in the value of their national currencies vis-à-vis the US dollar.

Savers in local currencies did better not only in terms of US dollars but also in terms of their overall purchasing power at home, as witnessed by average annual domestic currency interest rates[8] that have been higher than average annual consumer price inflation. The only exceptions are Japan and the Philippines, where annual inflation slightly surpassed returns made on government securities.

Overall, it looks like local currency investors in inflation targeting countries have not lost their wealth as far as US dollars are concerned. Conversely, savers who consistently invested in assets denominated in national currencies have managed to grow richer compared to those holding US dollar assets.

Tradeoff between saving in national currencies and the US dollar in inflation targeting (IT) countries, from IT adoption year through 2020
Inflation targeter IT adoption year Annual averages (compounded)
Inflation Local interest rate Local interest rate converted into USD US interest rate Difference, last two columns
Albania 2009 1.83 2.85 2.37 0.55 1.82
Armenia 2006 3.58 8.39 5.62 0.86 4.76
Australia 1993 2.42 4.22 4.72 2.30 2.42
Brazil 1999 6.02 12.72 7.14 1.57 5.57
Canada 1991 1.75 2.76 2.39 2.36 0.02
Chile 1999 3.16 3.94 2.38 1.57 0.81
Colombia 1999 4.69 6.02 2.99 1.57 1.42
Czech Republic 1997 2.47 2.70 4.88 1.84 3.03
Dominican Republic 2012 2.81 6.04 1.28 0.72 0.56
Georgia 2009 3.74 7.70 1.39 0.55 0.84
Ghana 2007 11.88 17.21 2.12 0.59 1.53
Guatemala 2005 4.44 5.20 5.03 1.11 3.92
Hungary 2001 3.71 4.91 4.56 1.25 3.30
Iceland 2001 4.30 7.45 6.26 1.25 5.01
India 2015 4.13 5.80 3.78 1.13 2.65
Indonesia 2005 4.92 7.49 4.93 1.11 3.82
Israel 1997 1.66 3.72 4.15 1.84 2.31
Japan 2013 0.59 -0.13 0.10 0.82 -0.71
Kazakhstan 2015 6.75 10.24 5.64 1.13 4.51
Korea 2001 2.22 2.74 3.72 1.25 2.47
Mexico 2001 4.18 5.83 1.57 1.25 0.32
Moldova 2013 5.17 9.25 5.02 0.82 4.21
New Zealand 1990 1.94 4.95 5.68 2.46 3.22
Norway 2001 2.00 2.60 2.90 1.25 1.64
Paraguay 2011 3.61 5.46 0.52 0.65 -0.13
Peru 2002 2.76 3.51 3.34 1.23 2.11
Philippines 2002 3.98 3.34 3.92 1.23 2.69
Poland 1998 2.73 5.26 4.92 1.71 3.21
Romania 2005 3.49 5.68 3.98 1.11 2.87
Russia 2015 3.79 7.75 7.45 1.13 6.33
Serbia 2006 5.01 8.09 4.55 0.86 3.69
South Africa 2000 5.48 7.41 3.91 1.36 2.55
Sweden 1993 1.52 2.35 2.41 2.30 0.11
Thailand 2000 1.94 2.09 3.97 1.36 2.61
Turkey 2006 9.73 17.69 4.52 0.86 3.66
Uganda 2011 4.43 11.75 7.11 0.65 6.46
Ukraine 2016 8.10 13.07 11.97 1.33 10.64
United Kingdom 1992 1.96 3.09 2.65 2.32 0.33
Uruguay 2007 8.09 10.76 5.14 0.59 4.55

Notes: Average annual (compound) interest rates are calculated based on yields on short-term treasury bills. Where such data are not available, money market rates or deposit rates are used as substitutes. Specifically, money market rates are used for Australia, Chile, Colombia, Czech Republic, Dominican Republic, India, Korea, Norway, Paraguay, Peru, Poland, Russia, Serbia, and Ukraine. Deposit rates are used for Guatemala, Indonesia, and Turkey. For several other countries, these rates are applied for specific years. The TONIA (Tenge OverNight Index Average) indicator is employed in the case of Kazakhstan. End of period exchange rates are applied to convert average local interest rates into US dollars. Average annual inflation rate calculations use the International Monetary Fund’s (IMF) World Economic Outlook projections for the index of consumer prices at the end of 2020. The table lists 38 inflation targeters as identified in this article by Sarwat Jahan at the IMF. Ukraine is included based on IMF’s Annual Report on Exchange Arrangements and Exchange Restrictions 2017, where it was first classified as an IT country. Thus, 2016 is shown as the starting year of Ukraine’s IT regime. Countries not listed are Costa Rica and Jamaica, which have been classified as inflation targeters only recently. More time needs to pass to compute meaningful metrics for these countries.

Sources: The primary source of data for treasury bill rates, money market rates, deposit rates and end of period exchange rates are the IMF's International Financial Statistics database (accessed on February 19, 2021). Missing data, particularly for 2020, are taken from Macrobond or Bank for International Settlements and national sources (central banks, finance ministries, security exchanges and other organizations, accessed either via Macrobond or directly). End of period indexes of consumer prices are obtained from the IMF's World Economic Outlook database, October 2020.

Notes

1. This number is valid as of April 30, 2019, and might change in future publications of IMF’s Annual Report on Exchange Arrangements and Exchange Restrictions.

2. This IMF working paper notes that deposit dollarization is particularly common in Latin American countries and transition economies in Europe. Please also see this analysis for more examples.

3. Where treasury bill rates are not available, money market rates or deposit rates are used. Please see notes in the table for more information.

4. 13-week US treasury bills (for more details, please see IFS World and Country Notes).

5. US treasury bills and government securities of countries covered in this analysis are not exactly comparable, as they carry different country default risks. The aim of this analysis, however, is to show interest rates generally available to investors who face the decision of whether to save in national currencies or US dollars.

6. Local interest income or interest rates refer to income or rates converted into US dollars, unless otherwise specified.

7. Savers in Paraguay also suffered a small loss.

8. Not converted into US dollars.

Data Disclosure

The data underlying this analysis are available here [zip].

More From

Related Topics