DETERMINANTS OF THE CURRENT ACCOUNT BALANCE AND OUTPUT IN UKRAINE Аbstract

Shevchuk V. O. Doctor of Economics, Professor, Head of the Department of International Economic Relations Lviv University of Trade and Economics, Ukraine; e-mail: victorshevchuk@netscape.net; ORCID ID: 0000-0001-6930-1440 Synchak V. Doctor of Economics, Professor, Head of the Department of Management, Finance, Banking and Insurance Chair Khmelnytsky University of Management and Law, Ukraine; e-mail: kmf@univer.km.ua; ORCID ID: 0000-0001-8957-3654 Zaverbnyj A. S. Ph. D. in Economics, Associate Professor, of the Department of Foreign Trade and Customs, Lviv Polytechnic National University, Ukraine; e-mail: 40anzas@gmail.com ORCID ID: 0000-0001-7307-536X Baranetska O. V. Ph. D. in Economics, Associate Professor of Department of Economic Security and Financial Investigations Ternopil National Economic University, Ukraine; e-mail: o.baranetska@tneu.edu.ua; ORCID ID: 0000-0003-4352-4421

Introduction. One of the most significant features of the Ukrainian economy of 2014-2017 was the lack of an expected transition to the current account surplus after a considerable devaluation of the hryvna in nominal and real terms (Fig. 1). In fact, real (taking into account inflation) hryvna exchange rate has decreased by 30-40 % compared to the average value of 2010-2013, but the balance of export-import of goods and services remains negative, which does not allow to increase the currency reserves of the National Bank of Ukraine and creates the basis for expectations of further depreciation of hryvna. At the same time, the obvious revaluation of the hryvna could have become a factor of the deepening of the decline in production, which has worsened the payment balance due to a decrease in supply of goods for export. In this context, it is interesting to compare the effectiveness of fiscal and monetary policy as instrumental tools of economic growth (GDP) and the balance-of-payments (BOP) equilibrium in narrow sense (the current account balance), as well as to reveal the nature of the interaction between the two most important macroeconomic indicators.
The purpose of the work is to study the impact of fiscal and monetary policy, as well as changes in the real exchange rate of the current account balance and GDP growth (taking into account the mutual relationship between two indicators). This is important for determining the economic policy that can provide simultaneous internal (GDP) and external equilibriums (the current account balance).  [10][11][12], albeit at the expense of a significant loss of hryvna deposits and restrictions on bank lending [2, р. 136-146]. At the same time, there are assumptions that since the beginning of 2016 the hryvna devaluation has lost its positive impact on the volume of export, and one of the reasons may be a reduction of investment and a decrease in labor productivity [3, с. 48-54]. In general, a significant devaluation of the monetary unit may not have the expected stimulant impact on export [4], while there will be a rapid decline in imports and production [5, р. 49-64]. The excellent impact of the exchange rate on income can attributed to such factors as business activity cycle, capital movements, dollarization of debt obligations, openness to foreign trade, revaluation of the real exchange rate, economic situation in trading partner countries [6, р. 680-708].
An alternative to devaluation of the hryvna may be the transition to the budget surplus [7], as it was in 2000-2001, but such a policy may reduce aggregate demand without the expected improvement of the export-import balance. Indeed, one study does not show the favourable impact of high tax revenues on the trade balance, which denies the logic of so-called fiscal devaluation [8]. However, earlier studies show a simultaneous improvement in the current account balance and the acceleration of economic growth due to the budget surplus [9], which corresponds to the dependencies for low-income countries [10, р. 47-76] or the countries with high indebtedness [11, р. 502-510]. Less attention was paid to the dependency of the current account on the supply of money supply. In one study for Ukraine, the volume of lending increases both export and import [9]. Foreign studies mainly show an inverse relation between the money supply and the current account balance [12, р. 408-416; 13, р. 451-462; 14].
The popular explanation of the low efficiency of the domestic export sector is the predominance of primary components [15, р. 38-42] and structural problems such as backlog by the level of development of the innovation sector, trade restrictions, bureaucratic procedures, etc [16, с. 80-85]. This provides for the direct causality of «GDP»current account», which denies the standard assumption of open economy models of the inverse relationship between the two indicators. In general, the nature of the interaction between income (GDP) and the current account balance can significantly modify the impact of economic policy instruments on both indicators.
Empirical results. For empirical study, quarterly tine series for the period 2000-2017 of the budget and current account balances (% of GDP) and the real effective exchange rate, monetary aggregate М2 and GDP (all three indicators are in deviations from equilibrium trend, in %) are used, as provided by the IMF and NBU databases. Visually the link between the GDP cycle and excess money supply, which can be considered an effective indicator of excessive supply of money supply, has been traced quite definitely since the mid-2000s, when the process of re-monetization of the Ukrainian economy was completed (Fig. 2). Similarly, the deterioration of the budget balance in 2008-2013 could be one of the reasons for the unfavourable dynamics of the current account balance (Fig. 1), although the reverse causality between the two indicators should not be excluded. Over the past three years the budget balance is rather unstable. As both the Augmented Dickey-Fuller (ADF) and Phillips-Perron (PP) tests indicate that our variables are stationary, it allows the use of structural vector autoregression (SVAR) model, with no need for error correction procedure. Such an approach has advantages in comparison to a canonical structural regression model, as it is difficult to specify properly complicated relationships between several interrelated variables. Compared with a general class of VAR models, specification of a SVAR model allows for imposition of specific structural restrictions upon individual functional relationships in accordance with economic theory.
Assuming the vector of moving average in the form of where X t is the 1 N  vector of endogenous variables, containing the budget balance (bd t ), excessive money supply (m2c t ), deviations of the RER from its equilibrium trend (rerс t ), the current account balance (ca t ) and the business cycle (yc t ). Matrix A 0 accounts for relationships between endogenous variables in the current period, A(L) is the polynomial covariance matrix, where L is the lag operator, C(L) is the matrix of lagged relationships between endogenous variables,  t is the vector of normally distributed uncorrelated serial stochastic shocks with dimension of 1 k  , u t is the vector (1) of normally distributed shocks with dimension of 1 N  , which are not serially correlated, but can be correlated in the current period.
For empirical estimations, the following SVAR model is chosen (in current shocks): 12 2, m c a bd u  (3) It is assumed that the budget balance is independent of other endogenous variables in the current period (equation (2)), being a factor behind money supply developments (equation (3)). It is natural to suppose that there is an inverse relationship between the budget balance and money supply ( 1 0 a  ). Deviations of the RER from its equilibrium trend are dependent on the fiscalmonetary mix and business cycle (equation (4)). As money supply developments reflect changes in both money supply and demand for money, accounting for the business cycle reflects realities of the dollarized economy, when a substantial amount of transactions is conducted with foreign currency. While the excessive money supply contributes to the RER appreciation ( 2 0 b  ) and a higher GDP is of an opposite impact ( 5 0 b  ), the effect of the budget balance upon RER is dependent on the structural features of the economy ( 1 0 b  ) (Shevchuk, 2008).
The budget surplus and RER depreciation are likely to improve the current account ( 13 ,0 cc ) (equation (5)). Effects of money supply are rather ambiguous ( 2 0 c  ). While monetary models imply worsening of the current account by an increase in the money supply, an opposite effect can be achieved if allow for the aggregate supply effects of money in an economy with financial constraints. In this case, an increase in the money supply contributes to production of goods for exports or import substitutes thus improving the current account balance.
According to the assumptions of the dependent economy model with tradable and nontradable goods, there is a possibility of asymmetrical money supply effects upon the current account and output (GDP) due to specific features of the real money supply and RER effects upon production on domestic and external markets (Trachanas, 2013). Generally, it is not ruled out that depreciation of the RER brings about a decrease in output ( 3 0 d  ). Although an improvement in the current account is likely to be expansionary in most of cases, it is possible that a crowding out of financial and other resources from inward-oriented branches of the economy in favor of the exporting sector can be of an opposite restrictionary effect. For obtaining of the impulse response functions (Fig. 3) and forecast error variance decomposition (FEVD) of residuals (Table), a SVAR model with two lags is chosen, as suggested by the Akaike and Schwartz criteria. Analysis of the estimates argues in favour of the budget balance surplus for the Ukraine's economy, as it contributes to a simultaneous improvement in the current account and GDP. At the same time, the budget surplus is associated with a decrease in the excessive money supply and strengthening of the exchange rate (in real terms) in the short run. However, the fraction of bd t in the FEVD of business cycle is rather low (no more than 7%), gradually increasing from 12 to 20% in the FEVD of m2c t . Such a result suggests about a significant monetization of budget deficits in Ukraine over the period from 2000 to 2017. Changes in the budget balance account for 19 % of FEVD for the RER and current account. On the other side, the budget balance improves as a result of improvements in both current account and GDP (to less extent), but the fraction of both factors in FEVD of bdt is small enough.  The pattern of money supply effects does not contradict assumptions of the dependent economy model under conditions of crowding out of inward-oriented branches of the economy by export-oriented activities. There are no grounds for frequent claims that «insufficient» money supply is an obstacle for economic growth in Ukraine. A depreciation on the RER on impact is another favourable outcome of a decrease in the excessive money supply. Money supply does not affect the budget balance, as suggested by both impulse response function and FEVD. Compared with the impact of budget balance, the fraction of m2c t in changes of the current account is smaller (no more than 5%), but gradually increases in FEVD of GDP to 17-18%. As the National Bank of Ukraine reacts to higher output by an increase in the money supply above equilibrium trend, it rejects the argument of the so-called monetary «famine» in the Ukraine's economy (fraction of y t in FEVD of m2c t approaches 13 %). Also, there is support for sterilization of the balance-of-payment monetary effects (the fraction of сa t in FEVD of m2c t gradually increases to 15 %) and restrictionary monetary stance in response to exchange rate depreciation (however, the fraction of rerс t in FEVD of m2c t does not exceed 9 %).

191
A depreciation of the RER results in a short-lived improvement of the current account, but with a simultaneous significant decrease in GDP. In the former case the fraction of rerс t in FEVD of ca t is small (up to 6%), but it is different in the latter case. Changes in the RER determine 50% of FEVD of yс t on the 4 quarters horizon, remaining significant even after 16 quarters (30%). It means that a steep depreciation of the hryvna can be considered as the single most important factor behind a deep fall in output in 2009 and again in 2014-2016. The impact of RER upon the budget balance and money supply is marginal.
One of unusual results is that an improvement in the current account is associated with depreciation of the RER (the fraction of ca t in FEVD of rerс t increases to 13 %). This kind of outcome can be explained in part by sterilizati on of the balance-of-payments monetary effects, but it is either convincing that it is the result of deliberate efforts to avoid an appreciation of the hryvna following the improvement in the current account. Depreciation of the RER resulting from an increase in output is consistent with the consequences of higher demand for money, with a high fraction of y t in FEVD of rerc t (22 %) further supporting the importance of this causal link. As the combined share of changes in fiscal and monetary policies does not exceed 16% of FEVD of rerc t , it does not contradict the logic of de facto fixed exchange rate practiced by monetary authorities since 2000 till the beginning of 2014. If account for changes in output and current account, the share of endogenous factors in FEVD of rerc t increases, but even in this case changes in the RER are determined by its own dynamics up to 50 %.
All said above looks like an increase in GDP is an important factor behind improvements in the current account in the long run, while a favourable reverse causality «ca t y t » is of the long run nature either (on the contrary, the current account surplus is worsening the output in the short run), However, both «ca t y t » and «y t ca t » causal relationships are rather weak according to FEVD. It is clear that changes in the current account are influenced mainly by external factors, and those of GDP being dependent predominantly on domestic factors, exchange rate and money supply in the first place (fiscal policy effects are less important). .

Conclusions.
The obtained results confirm the importance of fiscal discipline as a factor of improving the current account balance and GDP growth in Ukraine (the reverse favourable causality is traced at a statistically significant level only for the current account). At the same time, there are preconditions for strengthening the hryvna, which seems an important factor of increasing GDP, at least in the short term. Concomitant short-term deterioration of the current account balance does not seem to be a significant problem in terms of the direct dependence of this indicator on GDP. There are no signs of stimulating GDP by growing supply of the money supply, as many domestic experts say. The revealed direct relationship between the excess money supply and the current account can be the consequence of concentration of financial resources in the export sector, which has a negative impact on the rest of fields. The National Bank of Ukraine responds to the increase in GDP by the increase in the supply of the money supply, while we have an opposite reaction to improving the current account balance. Both dependencies seem to be standard and consistent from the view of the macroeconomic equilibrium, but some concern is monetization of the budget deficit.