Vladimir Sokolov

MIEF, SU-HSE

The Impact of Exchange Rate Policy on Interest Rates and Sovereign CDS Spreads Dynamics: Lessons from the Russian Experience

Introduction

After the late 90th Asian and Russian financial crises and collapse of the crawling peg exchange rate regimes the "hollowing out of the middle ground" view became popular among economists and policy makers. For example, Fischer (2001) argued that soft pegs are crisis-prone and not viable over long periods of time and only so-called “corner solution” like hard pegs or free floating exchange rate regimes are viable and desirable.

However, in the last decade many countries like Mexico and Brazil that de jure switched to independent floating regimes still de facto managed their exchange rates. Calvo and Reinhart (2002) call this phenomenon a "fear of floating". On the other pole a number of soft dollar peggers switched to a managed float with currency basket targeting. For example, on February 2, 2005 Russian Central bank announced a move from dollar pegging to a managed float against a basket of currencies. In July 2005 and May 2007 Malaysia and Kuwait[1], which for an extended period of time pursued soft dollar pegging also announced movements to a basket currency targeting. These examples demonstrate popularity of intermediate exchange rate regimes among small open economy central banks that try to pursue independent monetary policy by allowing more exchange rate flexibility but still are not willing to give up management of exchange rates.

The purpose of this project is twofold: 1) analyze how domestic money market in Russia reacted to exchange rate policy shift to basket targeting and 2) to evaluate how resilient was the basket pegging exchange rate regime since the beginning of the 2007 World financial crisis. Addressing the first issue allows to evaluate how successful was the policy change in terms of disconnecting the domestic interest rates from their foreign counterparts. The second issue deals with the idea that intermediate exchange rate regimes are crisis-prone and not viable as was argued after of the late 90th international financial crises.

Russian Exchange Rate Policy

Until February 2005 the Central Bank of Russia (CBR) used US dollar as an operating target and intervened on the foreign exchange market in order to smooth out the ruble-dollar exchange rate fluctuations. Starting from 2004 the Annual Monetary Report of the Russian Central Bank announced that in the medium-term the bank would adopt the inflation targeting (IT) monetary framework and independent floating exchange rate policy. One of the steps for achieving this objective was switching to a basket targeting exchange rate policy. The basket is composed of Euro and the USD and since February 2007 their proportions are 45% and 55% respectively[2]. This policy move has increased ruble flexibility against individual currencies since by the design the RUR/USD exchange rate became dependent on the Euro/USD rate.

Empirical part

Under IT policy the ability to control domestic interest rates is crucial for the central bank. Cross-country studies by Frankel et al. (2004), Shambaugh (2004), Obstfeld et al. (2005) confirm the existence of "trilemma" in monetary policy. First, I test how the relationship of the domestic interest rate with the foreign interest rates has changed after switch to basket pegging and if the interest rates and the RUR/USD exchange rate move in the direction suggested by Uncovered Interest Parity (UIP).

Secondly, I study how the exchange rate policy shift affected the RUR/USD exchange rate relationship with foreign cost of borrowing and the country's risk premium. This relationship measures the dependence of the ruble exchange rate on external factors across different periods of the exchange rate policy.

I utilize daily exchange rates, interest rates and sovereign CDS spread series and divide sample in four parts. The first July 2001- January 2005 covers the episode of USD targeting by the CBR. The February 2005 – July 2007 covers period of basket targeting prior to the beginning of the World financial crisis. The third August 2007 – August 2008 is a so-called “decoupling” period when Russia was insulated form the world crisis. The fourth is the Russian crisis period spanning September 2008 – November 2009.

Figures 1-3 illustrate dynamics of all series. From figure 1 we can see that prior to September 2008 the CBR maintained the bi-currency basket level quite well. Figure 2 demonstrates that introduction of the basket targeting in February 2005 reduced MosIBOR volatility and in the beginning of the world liquidity crisis in September 2007 the Russian domestic money market was unaffected.

Dynamics of the Ruble/dollar exchange rate

and Ruble/bi-currency basket

Fig.1

Dynamics of 1 month MosIBOR and LIBOR interest rates

Fig.2

From Figure 3 we can see that the risk premium on Russian debt moved together with other emerging economies risk premiums, suggesting that it is largely determined by external factors. The CDS spread significantly widened in September 2008 when investors realized that the emerging economies are facing a “sudden stop” of capital flows after collapse of the Lehman Brothers.

Dynamics of 5 year Russian and Mexican Sovereign CDS spreads

Fig.3

The standard ADF tests can’t reject non-stationary hypothesis for all time-series, while Johansen’s cointegration tests indicate existence of at least two cointegrating vectors. I use the Vector Error Correction (VEC) model.

(1)

where a measures the speed of adjustment to the equilibrium relationship and b is the long-run slope coefficient or the levels relationship.

I impose a Johansen normalization procedure and identify two cointegrating vectors. The first is a level relationship between MosIBOR, LIBOR and RUR/USD rates within the UIP hypothesis. The second captures relationship between RUR/USD exchange rate and external factors. I also impose restriction on the adjustment coefficients a for the row of LIBOR rates as we do not expect them to adjust to the equilibrium relationship with other Russia specific variables. The estimation results for different time periods are reported in Tables 1-4.

Table 1. Managed floating with US dollar as an operating target

Sample: 4 Sept. 2001 - 2 Feb. 2005 (764 obs.)

/ / MIBORlm / LIBORlm / RUR/USD / CDS5Y
DMIBORlm / -0.010**
(0.005) / -0.032***
(0.014) / / 1
(0.000) / -2.398**
(1.268) / -1.758***
(0.504) / 0
(0.000)
DLIBORlm / 0
(0.000) / 0
(0.000) / / 0
(0.000) / 2.045**
(0.855) / 1
(0.000) / -1.139***
(0.246)
DRUR/USD / 0.003***
(0.000) / -0.002**
(0.001)
DCDS5Y / 0.000
(0.001) / 0.007*
(0.004)

For the samples that cover basket targeting period I construct the synthetic time-series for the currency basket composed of euro and USD and synthetic interest rates, composed of LIBOR rates on dollar and euro deposits. I estimates VEC specification (1) separately for US dollar and basket series. The results are reported in upper and lower panels of Tables 2-4.

Table 2. Managed floating with Bi-currency basket as an operating target

No crisis sample: 2 Fed. 2005 - 8 Aug. 2007 (608 obs.)

/ / MIBORlm / LIBORlm / RUR/USD / CDS5Y
DMIBORlm / -0.023***
(0.009) / 0.004
(0.006) / / 1
(0.000) / -0.253
(0.424) / -0.409
(0.370) / 0
(0.000)
DLIBORlm / 0
(0.000) / 0
(0.000) / / 0
(0.000) / 2.751***
(0.675) / 1
(0.000) / -0.409
(0.370)
DRUR/USD / -0.004
(0.003) / 0.002
(0.002)
DCDS5Y / -0.003***
(0.001) / -0.004***
(0.001)
/ / MIBORlm / Synthetlm / Basket / CDS5Y
DMIBORlm / -0.001
(0.004) / 0.020
(0.018) / / 1
(0.000) / -2.979***
(0.488) / -2.337***
(0.692) / 0
(0.000)
DSynthetlm / 0
(0.000) / 0
(0.000) / / 0
(0.000) / 1.385***
(0.234) / 1
(0.000) / 4.019***
(0.646)
DBasket / 0.006***
(0.001) / -0.002
(0.005)
DCDS5Y / -0.001**
(0.000) / -0.012***
(0.002)

Table 3. Managed floating with Bi-currency basket as an operating target

Decoupling sample: 8 Aug. 2007 - 1 Sept. 2008 (254 obs.)

/ / MIBORlm / LIBORlm / RUR/USD / CDS5Y
DMIBORlm / -0.028**
(0.012) / -0.007
(0.039) / / 1
(0.000) / 1.068**
(0.480) / -3.181***
(0.766) / 0
(0.000)
DLIBORlm / 0
(0.000) / 0
(0.000) / / 0
(0.000) / -0.292***
(0.139) / 1
(0.000) / 0.877
(0.606)
DRUR/USD / -0.010
(0.006) / -0.042**
(0.020)
DCDS5Y / -0.008**
(0.003) / -0.018
(0.011)
/ / MIBORlm / Synthetlm / Basket / CDS5Y
DMIBORlm / -0.036***
(0.013) / 0.078
(0.099) / / 1
(0.000) / -0.719*
(0.395) / 0.949
(2.288) / 0
(0.000)
DSynthetlm / 0
(0.000) / 0
(0.000) / / 0
(0.000) / -0.124**
(0.061) / 1
(0.000) / -0.293
(0.186)
DBasket / -0.005
(0.004) / -0.073***
(0.027)
DCDS5Y / -0.006
(0.004) / 0.008
(0.028)

Table 4. Managed floating with Bi-currency basket as an operating target

Crisis hits Russia sample: 1 Sept. 2008 - 16 Oct. 2009 (272 obs.)

/ / MIBORlm / LIBORlm / RUR/USD / CDS5Y
DMIBORlm / -0.028**
(0.014) / -0.118***
(0.045) / / 1
(0.000) / -3.393**
(1.404) / -3.020***
(0.273) / 0
(0.000)
DLIBORlm / 0
(0.000) / 0
(0.000) / / 0
(0.000) / 1.611***
(0.409) / 1
(0.000) / -0.759***
(0.078)
DRUR/USD / 0.023***
(0.007) / 0.044**
(0.021)
DCDS5Y / 0.049***
(0.017) / 0.174***
(0.053)
/ / MIBORlm / Synthetlm / Basket / CDS5Y
DMIBORlm / -0.027**
(0.014) / -0.112***
(0.046) / / 1
(0.000) / -5.869**
(1.155) / -3.226***
(0.302) / 0
(0.000)
DSynthetlm / 0
(0.000) / 0
(0.000) / / 0
(0.000) / 2.142***
(0.312) / 1
(0.000) / -0.724***
(0.079)
DBasket / 0.023***
(0.006) / 0.044**
(0.019)
DCDS5Y / 0.042***
(0.017) / 0.172***
(0.056)

Conclusions

For USD operating target sample the domestic MosIBOR rate was positively related to LIBOR and RUR/USD exchange rate, which confirms the "impossible trinity" and UIP hypothesis for managed float currencies as documented in Frankel and Poonawala (2006). The signs of the adjustment coefficients suggest the equilibrium correction of MosIBOR and RUR/USD towards the equilibrium.

As can be seen form upper panel of Table 2 after the switch to bi-currency basket targeting MosIBOR rate got disconnected from both LIBOR and RUR/USD rates, however, the lower panel of the table indicates that it became positively related to bi-currency basket and the synthetic combination of US LIBOR and Euro LIBOR rates. At the same time the CDS premium is negatively related to the RUR/USD exchange rate suggesting that increase in a country's risk premium is correlated with ruble strengthening.

During the decoupling episode MosIBOR is negatively related to LIBOR rate and positively to the synthetic rate, suggesting that Russian domestic money market became more in-sync with the euro area monetary policy and didn't follow steep quantitative easyning of the US Fed. At the same time, MosIBOR rate is no longer related to the bi-currency basket but is positively related to RUR/USD rate.

As can be seen from both panels of table 4 for the crisis sample there is no distinction between dollar variables and synthetic variables with relation to MosIBOR and RUR/USD. Since ruble depreciated both against USD and basket this findings suggest that market participants didn't consider bi-currency basket as a viable operating target of the CBR. The estimated long-term coefficients are in line with those reported in Table 1 for the USD targeting sample. However, the adjustment coefficient a for the cointegrating relation is 4 times bigger, suggesting a much faster equilibrium correction during the crisis.

References

1.  Calvo, G.A., and Reinhart, C. (2002). “Fear of Floating”, Quarterly Journal of Economics, CXVII(117), pp. 379-408.

2.  Fischer, S., (2001). “Exchange Rate Regimes: Is the Bipolar View Correct?” IMF
Distinguished Lecture

3.  Frankel, J., Schmukler, S., and Serven, L., (2004). "Global Transmission of Interest Rates: Monetary Independence and the Currency Regime," Journal of International Money and Fiance, vol.3, pp. 701-733.

4.  Frankel, J., and Poonawala, J., (2006). "The Forward Market in Emerging Currencies: Less biased than in Major Currencies," NBER Working Paper 12496.

5.  Obstfeld, M., Shambaugh, J., and Taylor, A., (2005). "The Trillema in History: Tradeoffs Among Exchange Rates, monetary Policies, and Capital Mobility," The Review of Economics and Statistics, 87(3), pp. 423-438.

6.  Shambaugh, J., (2004). "The Effect of fixed Exchange Rates on Monetary Policy," Quarterly Journal of Economics, February, pp.301-352.

9

[1] Other notable countries that pursue basket targeting exchange rate policy are China and Singapore. Also during the period of dollar weakening against euro before the 2007 World financial crises Saudi Arabian policy makers were debating about abandoning the dollar peg regime and following the Kuwait example.

[2] It should be pointed out that other countries that pursue basket currency targeting do not disclose currency weights and use this policy tool for promoting the country's competitiveness against its main trading partners. In case of Russia the basket composition is publicly known and the policy is oriented at "gaining more exchange rate flexibility".