Analysis of External Debt-Based Growth Hypothesis with Markov Regime Switching Models for the Turkish Economy
Yusuf Doğan, Deniz ŞükrüoğluThis study presents the effects of Turkiye’s external debt on economic growth and is examined with Markov Regime Switching Models, unlike previous studies, considering that Hamilton’s (1989) regime-switching models are a very powerful tool for modeling non-stationary time series. In this study, annual data for the period 1970–2020 were used for the Turkish economy. Countries use external borrowing to achieve economic growth and development objectives. However, the effect of external debt on economic growth is still debated in the literature. The vast majority of empirical investigations have shown that external borrowing has a negative impact on economic growth. In the analyses, the 1970–2020 period of the Turkish economy was divided into three periods, which can be named growth, crisis, and precrisis periods. Analyses show that only the coefficient of the gross external debt stock variable differs in the obtained regimes. According to the results of the analyses, it is seen that the gross external debt stock has a negative effect on Turkiye’s economic growth in all regimes, including the growth period, while the gross fixed capital formation, export, and population have a positive effect
Dış Borçlanmaya Dayalı Büyüme Hipotezinin Türkiye Ekonomisi için Markov Rejim Değişim Modelleri ile İncelenmesi
Yusuf Doğan, Deniz ŞükrüoğluÜlkeler iktisâdî büyüme ve kalkınma hedeflerine ulaşabilmek amacıyla dış borçlanmaya başvurmaktadır. Ancak literatürde dış borçlanmanın büyümeye olan etkisi hâlen tartışılmaktadır. Yapılan ampirik çalışmaların büyük çoğunluğunda dış borçlanmanın büyümeyi negatif yönde etkilediği sonucuna ulaşılmıştır. Bu çalışmada, durağan olmayan zaman serilerini modellemek için Hamilton’un (1989) rejim değişim modellerinin çok güçlü bir yaklaşım olduğu görüşünden hareketle, Türkiye’nin dış borçlarının iktisâdî büyüme üzerindeki etkisi diğer çalışmalardan farklı bir şekilde Markov Rejim Değişim Modeli ile incelenmiştir. Çalışmada Türkiye ekonomisi için 1970-2020 dönemine ait yıllık veriler kullanılmıştır. Yapılan analiz neticesinde Türkiye ekonomisinin 1970-2020 döneminin büyüme, kriz ve kriz öncesi dönem olarak isimlendirilebilecek 3 döneme ayrıldığı görülmüştür. Elde edilen rejimlerde ise sadece brüt dış borç stoku değişkeninin katsayısının farklılaştığı sonucuna ulaşılmıştır. Model tahmininden elde edilen sonuçlara göre, Türkiye’nin iktisâdî büyümesi üzerinde toplam brüt dış borç stokunun büyüme dönemi de dahil olmak üzere tüm dönemlerde negatif etkisinin olduğu, sâbit sermâye yatırımı, ihrâcat ve nüfus değişkenlerinin ise pozitif etkisinin olduğu görülmüştür.
Countries strive to achieve economic growth and development. However, underdeveloped or developing countries struggle to meet these targets due to some factors including insufficient domestic savings, lack of information and technology, the balance of payments deficit, and foreign exchange constraint. Due to the inadequacy of national savings to finance the investments that will necessitate development, countries are forced to resort to external borrowing in addition to domestic borrowing. The effect of external borrowing on economic growth continues to be studied in the economics literature. Some economists believe that the chain reaction that foreign borrowing creates with growth can negatively impact both a country and the world economy. The fact that foreign debts involve interest payments in addition to the principal at the payment stage decreases national income.
According to the alternative view, countries with insufficient domestic savings and limited capital should prefer external borrowing, which is an important public financing tool, in order to achieve their growth and development goals. At the time the foreign debts are first received, they provide an additional resource inflow to the country. The new resources entering the country allow for the realization of investments, thus, an increase in national income. External borrowing is believed to be more advantageous than domestic borrowing since there is no problem with the distribution of resources within the country, as is the case with domestic borrowing. However, as stated, the payment stage causes capital outflow from the country, thus, a decrease in domestic resources. Therefore, the revenues of the resources provided by foreign borrowing should be greater than the costs. Otherwise, the country will face an outflow of capital, and the need for resources will gradually increase. Theories and discussions on foreign debt in the economic literature in this framework are given in the article’s first and second sections.
The relationship between external debt and economic growth has generally been examined with linear time series models. Despite their usefulness in various applications, linear time series models are generally incapable of representing certain nonlinear patterns such as asymmetry, or volatility clustering. Growth rates, for example, tend to be greater and more persistent during periods of expansion, whereas they are lower and less persistent during periods of recession. It would be unreasonable to expect a linear model to capture these types of characteristics for such a time series.
Nonlinear time series models have shown rapid development over a long time. The regime-switching model has become one of the most widely used nonlinear time series models in the econometrics literature. Markov regime-switching models allow us to capture more complex patterns since they introduce timevariation in the parameters in the form of state or regime-specific values, which characterize the time series in different regimes. In Markov regime-switching models, the current state is only determined by the previous period’s state and the transition probabilities matrix. Given the current period’s state, the transition probabilities matrix shows the likelihood of each of the states being visited in the next period. The estimation of state-specific parameters as well as state occurrence probabilities for each of the sample periods allows us to better interpret parameters associated with their specific regimes. The Markov regime-switching model and its main elements have been extensively discussed in the fourth section of the article.
The results of the estimated model, which examines the relationship between external debt and economic growth for the Turkish economy with the Markov regime-switching model, are shown in detail in chapter 5 of the article. The chapter also describes variables used in the model, tests, model comparisons, model validity tests, transition probabilities, and estimated regime periods for the model. The final model, in which the parameters are statistically significant, the coefficients are economically valid, and the regime periods corresponding to the periods of expansion and recession dates of the Turkish economy have been meticulously selected among the 18492 models.
According to the results of the analysis, it can be concluded that the total gross external debt stock has a negative effect on Turkiye’s growth, while the fixed capital investment, exports, and population growth have a positive effect. These findings of the final model obtained by the Markov regime-switching show that the debt overhang hypothesis is strongly supported. The results and the suggestions are discussed in the last section of the article.
