Financial dollarization in Russia: causes and consequences

Abstract

We review some aspects of financial dollarization in Russia, applying the main relevant theories to analyse the dynamics of several dollarization indicators. An econometric model of the short-run dynamics of deposit and loan dollarization is estimated for the last decade. We find that ruble appreciation was the main driver of the de-dollarization that occurred then and of the later episode of renewed dollarization. We estimate the overall (and sectoral) currency mismatches of the Russian economy. Evidence is presented for the significant currency risk vulnerability of the non-banking private sector.

Keywords:

• financial dollarization
• currency mismatch
• emerging markets

JEL classification:

• E44
• F34
• G32

Acknowledgements

We are grateful to an anonymous referee, Laura Solanko and to the seminar participants at the Bank of Russia and at the 8th ESCB Emerging Markets Workshop for their helpful comments. All views expressed in this article are those of the authors and do not necessarily represent the position of the Bank of Russia.

Notes

1. Traditionally the term ‘dollarization’ implies replacement of domestic currency by US dollars as the medium of exchange, store of value and unit of account. After the formation of the European Monetary System and the introduction of the single currency (euro), the new term ‘euroization’ (completely analogous to dollarization) came into use. Euroization was quite typical for a number of transition economies, including Russia. We use the term ‘dollarization’ to denote the replacement of the national currency by any foreign currency.

2. Financial institutions include investment funds, trust, leasing and factoring companies, commodity and stock exchanges, brokerage firms operating in the stock market, insurance companies, non-governmental pension funds, the state's Deposit Insurance Agency and the open join-stock company RUSNANO, established via reorganization of the state corporation Russian Corporation of Nanotechnologies.

3. The formal econometric analysis of foreign debt dollarization is hampered by the shortness of time series available only at quarterly frequency while the data for domestic dollarization are available monthly.

4. Naturally we only consider those variables that are suitable for country-specific analysis and therefore omit such indicators as the measures of financial integration and institutional quality even though they were found to be important in the cross-section analysis.

5. The changes in exchange rate were transformed so as to be proportional to the mechanical re-evaluation effect at time t (see Honohan 2007 for details). That is, for deposit dollarization erf _t = (1-DD_t−1)*DD_t−1*(e_t/e_t−1–1), where e is the ruble/foreign currency exchange rate. The erf variable in the loan dollarization equation was similarly constructed.

6. The volume of foreign cash conversion operations by Russian banks may be regarded as one of these indicators. As reported by the Bank of Russia, the proportion of volumes of sales of USD and euro cash to households amounted to 0.65 and 0.35 accordingly on average in 2005–11.

7. Similarly to the exchange rate factor variable changes in monetary base were transformed so as to be proportional to the mechanical effect coming from presumed changes the nominal value of national currency deposits as in Honohan (2007). That is, for deposit dollarization mbf _t = (1-DD _t−1)*DD _t−1*(m _t/m _t−1–1), where m is the broad monetary base. The mbf variable in the loan dollarization equation was similarly constructed.

8. We have included as many lagged variables as it was needed to prevent the autocorrelation of residuals. In both cases that amounted to two lags.

9. We used the Quadratic Spectral kernel that was shown to be optimal in Andrews (1991). The bandwidth selection is also based on Andrews (1991). The instruments include 3 lags of explanatory variables in deposits dollarization equation and 3 lags of exchange rate factor, monetary base factor and changes in foreign liabilities to total liabilities ratio in loan dollarization equation. In addition we use 3 lags of oil price changes as instruments in both equations.