Dating Financial Crises in Turkey, 1987-2013
Ali ARI, Kirklareli University, Turkey
Raif CERGIBOZAN, Kirklareli University, Turkey
1. Introduction
After relative stability in the post-World War II period, the world economy again became familiar withrecurrent crisis episodes following the collapse of the BrettonWoods system in 1971. Afterthe Latin American economies in the early 1980s, financial crises hitsome European countries in 1992-1993,Mexicoin 1994,South-East Asiain 1997-1998, Russia in 1998,Brazil in 1999, Argentina in 2001. The lastglobal financial crisis has also been affecting the world economy since late 2008.Theglobal economic and financial instabilityof the 1990s and 2000s also touchedthe Turkish economy,which underwentsevere financialcrises in 1994, 2001and 2008.
Theserepeated crisis episodes that caused high economic and social costs for the public sector as well as for private investors stimulated a large discussion on the theoretical specification of crisis models and the empirical analysesthataimat identifyingcrisesdeterminantsin order to predict future crisis episodes. In this regard, these studies are frequently called “early warning systems” (EWS), as they are likely to informpolicymakers and investors about the occurrence of a crisis in the near future.
There are three main components ofan EWS: methodology, explanatory variables and crisis index. Regardless of the method adopted, the empirical models construct first a crisis index as the dependent variable in order to identify crisis episodes.
Constructing currency crisis indexes is more frequent than building a banking or debt crisis index. Some empirical papers describe currency crises () as large depreciation or devaluation episodes (Frankel and Rose, 1996; Kumar et al., 2003 inter alia), while others define currency crises as cases where a currency comes under severe speculative pressure (Eichengreen et al., 1994, 1996; Kaminsky et al., 1998, inter alia). This second currency crisis definition takes into account both the situations where speculative attacks lead to devaluation and where the authorities successfully defend the currency by selling foreign reserves and/or rising domestic interest rates. The authors that adopt the second definition construct then an index of speculative pressure or exchange market pressure () as a weighted average of (real or nominal) exchange rate changes , foreign reserves changes and interest rates movements .
(1)
(2)
The weights of the components of the crisis index are often chosen to equalize their volatility and thus avoid the possibility of one of the components dominating the index (Aziz et al., 2000). Note that specifically the weights are frequently the inverse of the standard deviation of the corresponding component. The “successful” attack approach may be criticized for its limited crisis definition given that every speculative attack leads to economic cost for the government (reserves losses or interest rate rises) while the speculative pressure approach is mostly criticized because of that arbitrary weighting procedure of the components.
Once components of the crisis index and their weights are determined, one specifies an arbitrary threshold. Any month, quarter or year is classified as a crisis episode if the index value exceeds a specified threshold. The crisis index then becomes a binary crisis variable which takes the value of 1 if a crisis occurs and 0 otherwise.
(3)
The threshold level is generally set to a multiple of the standard deviation of the index plus the mean of the index . Values of the thresholds used in the literature have been ranged from to above the indexmean. However, different thresholds levels may generate identification of the different crisis dates as obviously shown in some studies (Kamin et al., 2001; Lestano and Jacobs, 2007). Note also that crisis dates may be different from one study to another according to the index components (whether interest rates and/or reserves are included into the index), and to the nominal or real character of the index components. This is why empirical studies should use different crisis definitions and different values of thresholds in order to assess the robustness of their crisis dating schemes.
However, building a banking crisis indicator is more difficult and less frequent in the literature. Researchers instead tend to rely on the judgments of observers with expertise about countries’ banking systems (Bell, 2000). As stated by Boyd et al. (2009) a variety of classifications of banking crises have been used since the mid-1990s by many researchers. But these classifications are generally event-based and are usually founded on the available ex post figures, which are related to banks’ losses and governments’ bailout costs (Kaminsky and Reinhart 1999; Caprio and Klingebiel, 1996, 1999; Demirguc-Kunt and Detragiache, 1998, 2005; Leaven and Valencia, 2013). As stated by Von Hagen and Ho (2007), this has several shortcomings. First, it tends to identify banking crises too late. Second, there are few objective standards for deciding whether a given policy intervention is “large.” Third, the timing of crisis periods on this basis is difficult because the exact date of policy interventions is often uncertain or unclear. Fourth, the events method identifies crises only when they are severe enough to trigger market events. Crises successfully contained by prompt corrective policies are neglected. This means that empirical work suffers from a selection bias.
Note that there are some early efforts to construct a banking crisis or banking fragility indicator. For example, Kibritcioglu (2003) builds a “banking sector fragility index” which is composed of changes in bank claims to private sector, changes in bank foreign liabilities, and changes in bank deposits. Von Hagen and Ho (2007) define “money market pressure index” as the weighted average of changes in the ratio of bank reserves to bank deposits and changes in the short-term real interest rate. Cesmeci and Onder (2008) also create a “money market pressure index” which is composed of changes in the ratio of central bank’s lending to banks over bank deposits and changes in interest rates. Ari (2012), inspired by Kibritcioglu (2003) and Ari and Dagtekin (2007, 2008), constructs a “financial fragility index” as the weighted average of changes in bank loans granted to private sector, changes in banking sector foreign liabilities, and changes in bank deposits.
Recent debt problems particularly in some Eurozone countries motivated a large number of economists to empirically identify determinants of the debt crises (Arellano and Kocherlakota, 2008; Fioramanti, 2008; Candelon and Palm, 2010; Reinhart and Rogoff, 2010; Baldacci et al., 2011a, 2011b among others). They use fiscal fragility index or debt crisis index in order to identify debt crises. Candelon and Palm (2010) employ the ratio of public debt (total debt) to GDP, while Pescatori and Sy (2007) use the refinancing rate or bond spreads (on the sovereign bond market) as a proxy of the sustainability of a country’s sovereign debt. However, there is no consensus about what index may generate robust results.
In this paper, we aim to construct differentcurrency, banking and debt crises indicators for the Turkish economy, which suffered several financial crises through the last three decades, and to assess the performanceof these indicators in identifying crisis episodes. In this sense, this paper has some common points with the study of Lestano and Jacobs (2007). To the best of our knowledge, there is not any study that employs different crisis indexes for the Turkish economy, except for Ari (2010) that compares the performance of different currency crisis indicators. However, this paper goes beyond Ari (2010) as it also assesses the performance of banking and debt crisis indicators. Furthermore, this study covers the entire post-liberalization era (1987-2013) by using monthly, quarterly and annual data gathered from the International Financial Statistics of the IMF and the Central Bank of Turkey.
The paper is organized as follows.Section 2 presents a brief history of the Turkish crises. Section 3 details the development of different financial crises indexes for the Turkish economy and assesses their performance. Section 4 concludes with some policy implications to avoid future crisis episodes in the Turkish economy.
2. A brief history of the Turkish economy (1980-2013)[1]
Following the severebalance-of-payments and debt crises in the late 1970s, Turkey reoriented its development strategy, based on import substitution on the real side and on negative real interest rates on the financial side, by adopting a radical structural adjustment program in January 1980. This program, largely supported by the IMF and the World Bank, aimed to implement a market-based mode of regulation in order to restore economic growth by improving economic and financial efficiency, increasing domestic savings and attracting foreign capitals.
The early phase of the program (1980-1984) was mostly characterized by the trade liberalization process,which consisted of export promotion and gradual import liberalization accompaniedby regulated capital movements (Boratav and Yeldan, 2001). As for the second phase (1985-1989),it was characterized by the financial liberalization process, which mainly consisted of ending interest rate controls and the liberalization of capital movements.
This large structural reform program obtained an initial success by reducing the triple-digit inflation rates to an average of30%, increasing export earnings at an annual rate of 10% and ensuring an average economic growth rate of 5.5% in the 1982-1989 period. With an out-looking economy and a liberalized financial system in the early 1990s, the Turkish economy was an example of a “success story” for other developing countries.
However, this early success was shadowed by the occurrence of two deep financial crises in April 1994 and February 2001. These crises led to severe economic consequences in terms of increasing interest rates (overnight ratesfrom 75% in December 1993 to 700% in March 1994, from 40% in November 2000 to 4000% in February 2001), large reserves losses, high currency depreciations of about 100% on the year basis, excessive output losses (5% and 7.5% of GDP respectively) and transfer of more than twenty national banks to the Savings Deposit Insurance Fund (SDIF) from 1994 to 2002.
As underlined by many scholars, the 1990s had been a lost decade for Turkey. A continuous deterioration of macroeconomic fundamentals, a highly vulnerable banking sector and a very unstable and fractional political environment due to successive coalition governments, early general elections and military operations against the PKK created all the ingredients of a crisis-prone economic structure (see Figure A1 in the Appendix for selected indicators of the Turkish economy,1990-2010).
Fiscal discipline in the Turkish economy was never achieved over the 1990s, except for some short-lived improvements in 1995 and 1998. These large budget deficits were caused by many factors: an expansionary fiscal policy mainly related to infrastructure investments, large subventions granted to exporting firms, an inefficient fiscal system, off-budget expenditures of the central government, and increasing interest payments on the public debt. The persistent deficits (on average 8% of GDP from 1993 to 2002) led in parallel to an increase to unprecedented levels in the public sector borrowing requirement (%12 of GDP in 2001) and in public debt stock. Moreover, inflation rates followed a steady path around 65% over the period. This unstable context of increasing uncertainties consequently caused high real domestic interest rates, low private investments (crowding-out effect), and low economic growth rates (about 3% of GDP over the 1994-2001 period).
The liberalization of capital movements in 1989 was another factor of vulnerability for the domestic economy, since excessive short-term capital inflows due to high real interest rates led to an overvaluation of the domestic currency that lowered the competitiveness of exporting firms. This was a big problem for an economy whose growth strategy was based on export revenues. This situation naturally deteriorated the current account balance (more than 4% of GDP before the occurrence of both crises). In order to offset these current deficits, the country needed more capital inflows that increased its short-term foreign debt stock (the ratio short-term debt over central bank international reserves was generally superior to100%).
Moreover, capital account liberalization also amplified the banking sector’s weaknesses by providing banks with the opportunity to borrow in international capital markets. As the unstable economic environmentincreased the credit risk for banks, they preferred to finance public deficits instead of granting credits to the private sector (the ratio bank loans to central government over banks’ total loans was around 40 % between 1993 and 2002). In this context, banks borrowed in foreign currency in international markets to invest in public sector securities in domestic currency, which generated a strong growth of theiropen positions (foreign assets over foreign liabilities about 75% preceding the 2001 crisis).
High foreign liabilities and increasing credit risk were not the only source of banking system weaknesses. The great part of state-owned banks in the sector created negative effects on the entire system. Interventions of the political authorities in their management and the use of their assets to finance public deficits deteriorated their balance sheets. Their massive borrowing requirementsthus increased the instability of the banking system.
Furthermore, the full deposit insurance implemented after the 1994 crisis in order to restore confidence in the system created moral hazard problems. This structure encouraged banks and their depositors to take excessive risks in order to get higher profits. Poor supervision and regulation of the system accompanied by the close connections between bank conglomerates and political authorities unquestionably fostered this excessive risk-taking setting.
As shown above, the 1994 and 2001 crises occurred in a very similar highly deteriorated economic and financial environment. However, their triggering factors were different. The change of the financing mechanism of public deficits from domestic debt to short-term advances from the central bank in late 1993 caused sharp rises in domestic credit thatincreased inflationary expectations and raised concerns about the viability of the economic situation (Celasun, 1998;Ozatay, 1999). A sharp downgrading in Turkey’s credit rating in the beginning of 1994 then played a triggering role in the occurrence of a severe currency crisis. In the following days, the currency crisis spread to the banking sector as the SDIF had to take control over three small-scale banks. In that regard, the 1994 twin crisis seems to have characteristics of the first generation crisis models.
The crisis was controlled with the announcement of a stabilization program supported by a stand-by agreement with the IMF on April 5, 1994. In the framework of this orthodox program, tightening monetary and fiscal policies were carried out to correct the fiscal fundamentals. A full coverage of insurance scheme for bank deposits was also put into effect after launching the stabilization program (Ertugrul and Selcuk, 2001).
This program obtained an initial success in the 1995-1997 period, but the generalized deficiency of financial markets towards emerging economiesfollowing the 1997-1998 Asian crisis, the 1998 Russian and the 1999 Brazilian crises again caused a deterioration of economic and financial fundamentals of the domestic economy. Eight other banks were transferred to the SDIF from late 1998 to the end of 1999. Therefore, another IMF-supported disinflation program was launched in January2000.However, the program that aimed at decreasing the inflation rate to a single digit by the end of 2002 collapsed in November 2000 due to a severe banking crisis. Even if the Supplementary Reserve Facility from the IMF of US$7.5 billion in December calmed down the markets, the public disclosure of a political disagreement between the Prime Minister and the President of the Republic on February 19, 2001 was perceived as a sign of political instability and thus caused speculative attacks (Yeldan, 2001). The monetary authorities defended the fixed exchange rate by mobilizing reserve stocks (US$5 billion in three days) and increasing the short-term interest rates to 4000%. However, following the investors’ generalized distrust, the authorities were forced to let the currency float on February 22nd. In that regard, the 2001 twin crisis seems to have characteristics of the third generation crisis models.
Another stabilization program, namely “Transition to the Strong Economy”, was announced on May 15 by the new Minister of Economy, Kemal Dervis, (former Vice President of the World Bank), again under the guidance of the IMF. The program, backed by US$19 billion IMF stand-by credits, restored relative economic stability by restructuring the banking sector (which cost over US$50 billion) and fulfilling many structural reforms. This reform process wasquite fruitful as remarkable improvements were recorded in terms of inflation (from over 50% in 2002 to 10% in 2010), budget deficit (from 12% of GDP in 2001 to 4% in 2010) and economic growth (on average 6% of GDP from 2002 to 2010 except in 2008-09).
Tight fiscal policies accompanied by an independent central bank focusing only on price stability were key issues behind the successful inflation targeting regime implemented since mid-2002. High and steady growth rates were also achieved by ensuring fiscal discipline with high primary budget surpluses (4% of GDP in the period from 2002 to 2008). The restructuring of the banking sector and large capital inflowsalso played an important role as the banks’ claims onthe private sector doubled after 2002. Moreover, political stability was also a key determinant of economic stability since the AKP came into single-party-government in November 2002.
However, these achievements cannot hide the vulnerabilities that the economy encounters. Sharp increases in current account deficits (more than 6% of GDP from 2002-2008), very high unemployment rates, a highly indebted private sector (short-term external debt over US$70 billion and overall foreign debt superior to US$190 billion), and difficulties to attract direct investments from abroad are the highly notable ones. The economic growth strategy, which is closely related to export earnings and short-term capital inflows, seems indeed to be the main reason behind the vulnerability of the country to external shocks, as the May 2006 and October 2008 episodes illustrated. Contrary to the 2006 episode, which only led to a large fluctuation in the financial markets, the 2008one heavily affected the real economyas the country recorded an economic recession of nearly 5% of GDP, slumping export revenues (more than US$30 billion) and rising unemployment rates to 14% in 2009.