15. Responding to Currency Crisis: The Case of Kazakhstan

(See handout no.18)

The Russian crisis in 1998

How can a developing or transition economy, which is small, vulnerable and newly integrated, defend itself when a large external shock occurs in the neighboring region? Is there anything the government can do to ameliorate the negative impact? Focusing on Kazakhstan, we will consider some policy issues related to crisis response.

When the Russian currency and financial crisis erupted in August 17, 1998, its adverse impact spread to the Central Asian republics. However, the transmission mechanism of the Russian crisis was different from that of the preceding East Asian crisis (1997-98). Unlike Thailand, Korea or Indonesia, the Central Asian republics did not receive large amounts of short-term private capital in the first place and, consequently, they did not suffer from overborrowing and collective capital reversal. But there were other channels of crisis contagion similar to those in East Asia, to which their macroeconomic authorities had to respond.

Within Central Asia, policy responses to the Russian crisis were distinctively different among Kyrgyzstan, Uzbekistan and Kazakhstan. Kyrgyzstan, the smallest among the three and an avid reformer, refused to impose any protective measures when the crisis occurred and remained fully open as a matter of principle. By contrast, gradualist and authoritarian Uzbekistan sustained tight trade and exchange controls, including multiple exchange rates (introduced in 1996), throughout the crisis. Finally, Kazakhstan, a country firmly committed to the market principle, deviated partially and temporarily from its market commitment in order to defend the domestic economy. These differences in policy response reflect initial economic and political conditions as well as different transition paths adopted by each country since independence.

Despite these differences, it is also important to recall the underlying similarities among the transition economies in Central Asia (and perhaps elsewhere) which make them highly vulnerable to external shocks. These include:

(1) Small economic size
(2) Large decline in output and living standards in the first half of the 1990s
(3) De-industrialization (i.e., the collapse of the existing manufacturing base)
(4) Dependence on a small number of primary commodity exports
(5) Weak and underdeveloped private sector
(6) Weak and underdeveloped financial sector

Wheat field near Astana/
View of Almaty (former capital)

Although geographically large, Kazakhstan is also a small, open and primary commodity-exporting economy. Its exports are dominated by energy and metals, plus some cereals. The total trade to GDP ratio was 62% in 1997. Russia was its major trading partner (about 50%), and its export structure was similar to (i.e., competed with) Russia's. For these reasons, Kazakhstan was particularly severely affected by the Russian crisis.

Macroeconomic management before the Russian crisis

Since its political independence in December 1991, the goal of Kazakhstan's macroeconomic policy shifted significantly over time.

Hyperinflation within the ruble zone (Dec. 1991 to Nov. 1993)

As price controls were lifted, suppressed inflation under the planned economy became open and the previous monetary overhang (excess supply of money) was quickly eliminated. This resulted in an enormous increase in price levels. But this supposedly one-time upward price adjustment turned into continuous hyperinflation for the following few years. Large fiscal imbalance was mainly responsible for this, but the existence of the ruble zone also encouraged monetization of these fiscal deficits.

Prior to the dissolution of the USSR in December 1991, the Soviet republics collectively formed a monetary union based on the Russian ruble, which was inconvertible both internally and externally. Even after the breakup, member states of the Commonwealth of Independent States (CIS) continued to operate under the ruble-based system (except the three Baltic states). The Russian ruble continued to circulate in CIS, including Kazakhstan. The newly created central banks (previously branches of the USSR Gosbank) retained the right to issue non-cash credits in rubles although all ruble notes were printed in Moscow. This meant that each government could issue ruble credits to cover its fiscal deficit while imposing most of its inflationary costs on other countries. With this perverse incentive, unrestrained monetization of fiscal deficits by CIS countries collectively led to persistent hyperinflation.

Inflation surged to triple-digit levels in the CIS countries during 1991-93. Restoration of monetary order required either imposition of strong discipline and coordination among new central banks or, alternatively, replacement of the ruble zone by separate national monies in order to terminate the adverse monetary incentive. The first option was unrealistic in the atmosphere of intensifying economic crisis and political uncertainty. One by one, CIS countries left the ruble zone to issue their own currencies. In Central Asia, the Kyrgyz som was created in May 1993, the Turkemen manat in October 1993, the Kazakh tenge in November 1993, the Uzbek sum in July 1994, and the Tajik ruble in May 1995.

Continued high inflation under a new currency (Nov. 1993 to mid 1995)

The new currency enabled Kazakhstan--at least in theory--to conduct an independent monetary policy and combat inflation. However, inflation remained high and even accelerated after the introduction of the tenge. The government and the central bank adopted a stabilization program based on fiscal adjustment and monetary aggregate targets. However, with fragile confidence, high uncertainty, and extreme volatility of the income velocity of money, disinflation was at first unsuccessful. In the mean time, the real economy suffered severely. Real GDP declined as much as 12.6 percent in 1994--the largest annual decline since independence.

During this period, the passive floating of the exchange rate was the only practical option in currency management. To do otherwise would have resulted in enormous overvaluation. Between Nov. 1993 and Nov. 1994, consumer prices jumped 15 times while the tenge depreciated 11 times against USD. By mid 1995, however, monthly inflation came down to below 5 percent and the speed of nominal depreciation, which already slowed down in late 1994, further decelerated.

Inflation-targeting and exchange rate-targeting (mid 1995 to Aug. 1998)

In the ensuing period, macroeconomic stabilization continued to be the main policy objective. Lower inflation and stability of the tenge was considered necessary for sustained growth. Adopting a restrictive policy stance, the National Bank of Kazakhstan (NBK) gradually ended financing of the state budget deficit as the government turned to issuing securities (which was non-inflationary). To restore confidence in monetary policy, the NBK also began to announce annual targets for inflation and the exchange rate.

To achieve this, the announced pace of the tenge's depreciation was reduced significantly. From May to December 1997, the exchange rate crawling was virtually halted as the currency was fixed at KZT 75.55= USD1. Increased exchange rate predictability and reduced incentive to speculate in the currency market led to the growth of the government securities market. 

However, the combination of tight money and a stable exchange rate gradually caused a real appreciation of the tenge and illiquidity (credit crunch) in the real sector of the economy. In addition, the Kazakh economy began to suffer from the 1997 Asian crisis, as (i) exports to and FDI from East Asia (especially Korea) declined; and (ii) global commodity price deflation had an adverse effect on Kazakhstan's export receipts.

Policy response to the Russian crisis

Initial response (Aug.1998 to Mar. 1999)

Instability in the Russian financial markets culminated in the floating of the Russian ruble and external defaults in August 1998. As Kazakhstan's tradable sector lost price competitiveness and export markets, its balance of payments worsened, downward pressure on the tenge mounted, and the Kazakh economy started to contract. The government adopted the following policies to offset the shock and avoid an uncontrollable collapse of the tenge and inflationary resurgence.

--A bit faster depreciation (but still under a crawling peg)
--Moderately tight money and budget and a mild rise in interest rates
--Temporary trade barriers against Russia, Uzbekistan and Kyrgyzstan
--Additional exchange control to curb illegal imports

Specifically, from January to June 1999, imports of certain Russian goods (21 items) including food were banned, and higher tariffs were imposed on food and light industry goods from Kyrgyzstan and Uzbekistan (although Kazakhstan's long and open border made these protective measures difficult to enforce). The minimum amount of currency that could be taken out of the country without documentation was lowered from USD10,000 equivalent to USD3,000 equivalent. Exchange control was tightened through the introduction of "contract passports" which required cross-checking by exporters and importers, banks and customs officers. This was meant to stamp out undeclared foreign trade.

Despite these measures, the negative impact of the Russian crisis turned out to be more severe and prolonged than expected. Real growth in 1998 (including seven pre-crisis months)  was -2.5 percent. Due to sustained tight macroeconomic stance, price deflation and illiquidity at small and medium enterprises intensified. During the 12 months ending in April 1999, WPI declined 8.4 percent while CPI fell 1.8 percent. The monetary authority was criticized by some as being too strict in demanding banks to comply with the BIS capital adequacy standards in the midst of the regional crisis. Export performance did not improve, and the import-substitution sector (especially food and light industry) was hit hard by Russian competition. Usable international reserves dwindled from USD 1.46 billion in July 1998 to USD 0.99 billion in March 1999.

Floating with regulatory measures (Apr. 1999)

The relatively stable KZT/USD exchange rate could no longer be supported, even at the accelerated pace of depreciation. In April 1999, Kazakhstan abandoned the managed slide of the currency and moved to a floating exchange rate system. The official statement said:

During the last months the Government and the National Bank carried out detailed analysis of all proposals on the tenge exchange rate policy for 1999 (continuing the policy of a managed float with a smooth and even depreciation, implementing a one-step devaluation, linking tenge to a hard currency basket, introduction of a fixed exchange rate, freely floating exchange rate). Advantages and disadvantages of each possible exchange regime were evaluated, taking into account the situation in the country and abroad. Based on the results of this analysis the Government and the National Bank made a decision to move to the regime of a free floating exchange rate of tenge as the most appropriate regime under the current conditions of Kazakhstan's open economy and unstable conditions in international financial and commodity markets ("Statement of the Government of the Republic of Kazakhstan and the National Bank of the Republic of Kazakhstan on further policy of tenge exchange rate," April 5, 1999).

But unlike the floating of the Thai baht on July 2, 1997, this "free floating" of the tenge was accompanied by new but temporary trade and capital controls. The same statement also made it clear that "the National Bank will not let exchange rate fluctuations to exceed rational expectations." In addition to the possibility of smoothing currency interventions, the following measures were introduced as the tenge began to float:

--Temporary requirement to sell 50% of foreign exchange earnings to the market (not to the government)
--Attractive forward currency contracts (offered by the central bank) for those who refrain from withdrawing tenge deposits
--Regulatory rules on banks were eased to avoid credit crunch
--Public utility tariffs were temporarily frozen and new poverty measures were announced
The tenge depreciated 33%, then stabilized within a few months.

Note that this approach was very different from the IMF conditionality at the time of East Asia's 1997 crisis (lecture 11). At that time, aggressive macroeconomic tightening and structural reforms were imposed during the crisis. In Kazakhstan, no such measures were taken; instead, temporary non-market measures were introduced to soften the blow on the population, banks and enterprises.

The IMF, ADB and World Bank country representatives praised the Kazakh crisis response. I also think that this was a very pragmatic and brilliant macroeconomic management by the National Bank of Kazakhstan. (Another policy that seems reasonable in Kazakhstan is the management of oil revenue so as to channel it into productive purposes, smooth annual fluctuations, and avoid the Dutch disease.)

Contagion channels

The Kazakh tenge was current-account convertible and the country had fairly open capital-account transactions (Kazakhstan acceded to Article VIII of the IMF Agreement, which stipulates current-account convertibility, in July 1996). However, geographical remoteness and underdevelopment of its market economy (in both real and financial sectors) have been the major constraints in attracting foreign private capital. Rampant corruption and other official irregularities were also frequently reported by foreign investors. Although Kazakhstan successfully issued sovereign eurobonds, inward investment or private bank borrowing had been limited.

Since Kazakhstan's domestic financial markets were highly insulated (due mainly to the lack of investors rather than formal barriers), there was no surge of short-term private capital inflow as seen in East Asia. As a consequence, "capital-account crisis" characterized by massive capital reversal and adverse balance-sheet effects did not occur. However, in the Russian crisis there existed real-sector channels of contagion which were powerful enough to upset neighboring economies. Significant downward pressure on currencies can arise even without massive international capital movements.

Generally speaking, Central Asian economies are small and highly open. The GDP of Kazakhstan (largest in the region) is only 1/20 of Russia's. Other republics are even smaller. In Kazakhstan, the largest shock from the Russian crisis came through the trade channel, especially declining exports.

The crisis contagion from Russia occurred through the following real-sector channels:

--Loss of price competitiveness (overvaluation)
--Loss of markets (in Russia and globally)
--Import penetration of cheap Russian goods
--Global decline of commodity prices (due to the Asian crisis)
--Declining FDI from East Asia (especially from Korea)

[Statistics will be supplied in class]

Some policy questions

Based on the above discussion, we pose the following three questions:

Q1. Can we make a meaningful comparison between East Asia and Kazakhstan?

Between the two crises, currency markets evolved very differently. In both cases the monetary authorities initially resisted the downward exchange rate pressure by intervention, losing international reserves. But in East Asia, after the floating of the Thai baht the regional currencies continued to depreciate out of control for several months, way below any reasonable estimates of purchasing power parities. By contrast, Kazakhstan maintained the downward crawl of the tenge for seven months after the initial shock. When the tenge was finally floated, it immediately lost 22 percent of its value against USD but stabilized shortly after four days. In two months, the KZT/USD rate became very stable at 33 percent below the pre-float level--as if the dollar peg was restored. After the floating began, the NBK refrained from intervening in the foreign exchange market.

It seems that in Kazakhstan the currency market stabilized much sooner than in East Asia. Does this reflect difference in the nature of the crisis, or better policy response?

The nature of the crisis was certainly different. First, East Asia's crisis was a "capital-account crisis" in which private financial institutions played a key role, whereas crisis contagion in Central Asia occurred primarily through real channels. Second, currency depreciation in East Asia was more simultaneous across countries with relatively similar economic size, in comparison with Central Asia where Russia was the dominant external source of instability and regional currencies behaved more independently from each other. Third, economic structure was fundamentally different: East Asian market economies were more developed, diverse and industrialized than the newly-born Central Asian republics.

But at the same time, it cannot be denied that policy also made an additional difference. Kazakhstan's response was significantly different from the East Asian experience in 1997:

--Fiscal and monetary policies were greatly tightened in East Asia, but not in Kazakhstan.
--HIgh interest rate policy was adopted in East Asia but tenge interest rates were raised only moderately and for a short time.
--Kazakhstan implemented temporary trade and exchange controls to protect the national economy, but no such measures were taken in East Asia (except Malaysia, belatedly).
--To avoid bank runs and deposit withdrawals, the NBK offered artificial forward exchange contracts and eased bank regulations. In East Asia, bank closure and prudential requirements were enforced.
--Bold financial and corporate reforms were initiated during the Asian crisis but no acceleration of structural reform was observed in Kazakhstan.
--Temporary poverty alleviation measures were prepared by Kazakhstan prior to floating.

Q2. Were the timing, size, and manner of the tenge's depreciation appropriate?

This was the issue I discussed with the Kazakh fiscal and monetary authorities during my visit to Almaty and Astana in 1999.

Timing of floating: The tenge was floated eight months after the outbreak of the Russian crisis. Was this too late or appropriate? Views differed widely even among top officials. Some argued that the tenge should have floated immediately, and blamed the delayed decision for the loss of international reserves, disrupted trade, and negative growth. Others contended that the initial defense of the currency was justified to avoid exchange rate overshooting, excessive inflation and balance-sheet damage.

Size of depreciation: from end March to early June 1999, the nominal depreciation of the tenge was 33 percent against USD. Was this too much or too little? Exporters wanted a larger depreciation, but the authorities were happy with the level at which the tenge settled (around KZT 132=USD1). CPI rose 12 percent in the first four months of depreciation, implying the pass-through coefficient of 0.36. Even with this partial offset, our REER calculation shows that the tenge's depreciation largely succeeded in restoring competitiveness to the pre-crisis level.

Astana, Kazakhstan's new capital

Macroeconomic policy stance: since the crisis started, the government raised taxes and cut expenditures only moderately (including the slowing of the construction of Astana, the new capital city). Relatively tight monetary policy was also continued. The government adjusted the 1999 budget to be more expansionary, but the IMF remained more cautious about fiscal expansion. The general question here, with an important implication for the Asian crisis, is whether macroeconomic policies should be tightened, eased, or remain neutral as the currency crisis begins, or when the home currency is floated in the aftermath. In this connection, fiscal stance should be assessed by cyclically-adjusted rather than actual deficits. During a severe recession, deficits will naturally increase but that should not be interpreted as fiscal expansion.

Soundness of the banking system: Kazakhstan's banking reform is one of the most advanced within CIS. At the time of the Russian crisis, the NBK felt that it had a good supervision over the balance sheets of domestic banks. It believed that banks faced no exchange risks because their positions were fully covered. But large domestic enterprises which borrowed from these banks in foreign currency in turn were exposed to such risk (unless they were also covered). Enterprises' foreign-currency liabilities additionally included external loans and FDI-related credits. If these enterprises go bankrupt due to exchange losses, banks will face associated default risk (I raised this question but the NBK people said there was no problem). But even in that case, systemic impact would be smaller than in East Asia due to limited credit expansion in the economy.

Q3. More generally, how should a transition economy manage its exchange rate?

Saule Zhakparova, my former student, at her desk in NBK (1999).

Here, we are talking about countries which are not yet "emerging economies" due to limited international financial integration. For a large number of developing and transition economies, there should be a wide range of possible exchange rate systems from fairly flexible to highly rigid, which must be selected pragmatically and case by case (Frankel 1999). Under the current instability of the international monetary system, the choice of exchange rate regime is a problem too complex to be reduced to a simple formula.

Despite the hypothesis of the "Vanishing Intermediate Regime" that says that only free float or rigid fix is sustainable (Eichengreen 1994, Obstfeld and Rogoff 1995), these corner solutions are precisely the ones to be avoided by economies with limited financial integration. This is because they are faced with the delicate task of balancing the two conflicting goals of exchange rate stability and flexibility under diverse and shifting conditions.

In this lecture, Kazakhstan's shifting macroeconomic problems and external vulnerability, before and after the Russian crisis, have been described. Initially, the challenge was stopping the runaway inflation by establishing macroeconomic independence and discipline. By mid 1997, this was more or less achieved through the combined use of inflation- and exchange rate-targeting. However, the subsequent Russian crisis seriously upset the Kazakh economy. The government and the central bank responded by adjusting macroeconomic and exchange rate policies as well as introducing temporary measures to soften the blow. While the details may be criticized, the orientation of Kazakhstan's policymakers towards attentive pragmatism should be highly evaluated.

@


<References>

Eichengreen, Barry, International Monetary Arrangements for the 21st Century, Brookings Institution, Washington DC, 1994.

Frankel, Jeffrey A., "No Single Currency Regime is Right for All Countries or at All Times," NBER Working Paper 7338, September 1999.

Obstfeld, Maurice, and Kenneth Rogoff, "The Mirage of Fixed Exchange Rates," NBER Working Paper 5191, 1995.

Ohno, Kenichi, and Saule Zhakparova, "Responding to Regional Currency Crisis: Kazakhstan in the Aftermath of the 1998 Russian Crisis," December 1999 (ADB Institute, unpublished).

Zhakparova, Saule, "Exchange Rate Policy in the Republic of Kazakhstan," policy paper, Graduate School of Policy Science, Saitama University, August 1997.