Does mutual causality between direct and indirect

---

Anais da Academia Brasileira de Ciencias (2017) 89 (3 Suppl.): 2544-2557 (Annals of the Brazilian Academy of Sciences) Printed version ISSN 0001-3765 / Online version ISSN 1678-2690 http://scielo.br.com/en/scielo.php/script=sci_serial&pid=0001-65&nrm=iso www.scielo.br/aabc

Does mutual causality between direct and indirect finance and economic growth exist in Russian Federation? MAYA KATENOVA1, SANG HOON LEE2 1Doctor

in Business Administration, Assistant Professor, Department of Finance and Accounting, Dr. Bang College of Business, KIMEP University, Kazakhstan, Almaty, 050010, Abai avenue, 2 2Professor, PhD, Dean of Dr. Bang College of Business, Department of Finance and Accounting, Dr. Bang College of Business, KIMEP University, Kazakhstan, Almaty, 050010, Abai avenue, 2

ABSTRACT The mutual causality between stock market development and the economy is empirically investigated in Russia. The methodology employs Vector Autoregression technique to test the mutual causality between Russian Industrial Production Index and the stock market development. Russian Trading System Index is employed as an indicator of stock market development in Russia. Money Supply is used as an additional macroeconomic variable besides Industrial Production Index. The credit spread is employed as an additional variable because of its anticyclical nature. It grows during recessions and decreases during an economic boom. Besides the aggregate data, two sectors’ data is employed. Sectors employed are oil and gas sector and services sector. The limitation of the study is that only ten-year period between 2006 and 2016 is employed. Another limitation is that only four variables are employed. The aggregate results and sector-specific results show that there is no any mutual causality between the Industrial Production Index and the stock market development in Russian Federation. Keywords: mutual causality, Vector Autoregression, stock market, economic growth, macroeconomic.

INTRODUCTION Financial impacts on the economy or an impact of the economy on the sphere of finance may include both direct and indirect finance. And in this particular study, the effects of equity market on or from the economy are analyzed in such a big and influential CIS country as Russian Federation. Although there are many papers discussing relations between economic and financial variables, few studies deal with CIS countries. Such reasons might be mentioned as an underdeveloped stock market, lack of data, dependent market structure. However,

Russia has its own well-functioning stock market and well-functioning financial system. The paper will contribute further on the studies about CIS economies as well as to a simultaneous development of their financial markets and institutions. A stock market Index is an indicator of industrialization in a particular country. In some countries, it may include 200-300 corporations’ stock; in other countries, it may have fewer corporations. Russian RTS Index includes stock of 50 Russian corporations. Focusing on the stock market development for the country, it should be noted that the presence of a developed stock market promotes the inflow of foreign investments to the An Acad Bras Cienc (2017) 89 (3 Suppl.)

2545

MAYA KATENOVA et al.

national economy, which is extremely important for developing countries (Arefiev and Kuznetsov 2015). A stock investment tends to be pro-cyclical. And pro-cyclical investment behavior may accelerate the development in an economy. However, if investors behave in a counter-cyclical way to exploit the low price advantage during a recession, they may affect the economy in an opposite way, improving economic conditions. Therefore, the study attempts to discuss the effect of stock markets on and from the economy and to compare the magnitudes of impacts for policy considerations. The stock market is similar to the banking industry in terms of facilitating the transfer of funds from lenders to borrowers. The results may imply important policy recommendations on effective capital market monitoring through the direct stock market channel (Arefiev 2016). Hypothesis 1: There is a mutual causality between stock market and the economy in Russia Hypothesis 2: There is a unidirectional causality so that stock market affects the economy in Russian Federation Hypothesis 3: There is a unidirectional causality so that the economy affects the stock market in Russian Federation The increased liquidity in the economy will result in an enhanced economic boom. The coincidental or follow-up cyclical movements of financial variables upon economic developments are referred to as procyclicality of financial variables (Landau 2009), and increased procyclicality simply means cyclical fluctuations with broader amplitude. Due to a high cost of procyclicality on the real economic activity, policy makers promote countercyclical measures on financial variables to soften the business cycle (Athanasoglou et.al. 2013). Such an author as Rashid in 2008 discussed the idea of dynamic interaction between macroeconomic

variables and stock prices in Pakistan. The author employed cointegration technique in his paper in order to test the long-term relationship between macroeconomic variables and stock market indicators. The results of cointegration technique strongly prove the fact of cointegration between the stock prices and macroeconomic variables in Pakistan. The author mentioned the fact that an efficient and well-functioning stock market may facilitate economic growth in a country. This is direct finance issue, which was popularized by other authors as well. For example, such authors as Dimitrova (2005), Hsing (2004), Ibragim and Aziz (2003) and Hondroyiannis and Papapetrou (2001) studied relationships between macroeconomic variables and stock market performance. The results of those authors’ studies’ differ from each other. Dimitrova (2005) found that stock prices and output have negative relationships in the short run but positive in the long run. Ibrahim and Aziz argued that stock market is playing a predictive role for macroeconomic variables (Ibrahim and Aziz 2003). However, Hondroyiannis and Papapetrou have concluded an opposite view that domestic economic activity affects the performance of stock market (Hondroyiannis and Papapetrou 2001). The general consensus in the literature is that developed, sound, and functioning financial systems facilitate sustainable economic growth. This conforms to the reasoning of the new endogenous growth theorists. The causal relationship between financial depth and economic growth remains controversial and unclear despite the fact that it has been investigated extensively in the economic literature. Arguably, this divergence might emanate from differences in estimation techniques and data. In particular, results seem to be greatly determined by the choice of financial depth indicators (Chukwu and Agu 2009).

An Acad Bras Cienc (2017) 89 (3 Suppl.)

DIRECT AND INDIRECT FINANCE AND ECONOMIC GROWTH

As mentioned in Calderon and Liu, economic literature defines financial sector development as the improvement in quantity, quality, and efficiency of financial markets and intermediaries services (Calderon and Liu 2002). Financial sector development affects the real growth of output in different ways: the volume of investment increases and the next to improve is the volume of savings (Goldsmith 1969). In recent times, the financeeconomic growth nexus has attracted global attention especially in emerging and developing economies. There is still a divergence in views regarding the role of financial intermediaries and financial markets in facilitating sustainable economic growth in the long-term. Besides bank lending, there is an access to fund-raising in the capital market, which is direct finance. For example, Levine and Zervos argue that well-developed stock markets may be able to offer different kinds of financial services than banking systems and also may provide a different kind of impetus to investment and growth than the development of a banking system alone (Levine and Zervos 1996). The authors mentioned that increased stock market capitalisation may improve an economy’s ability to mobilise capital and diversify risk. They also mentioned that various measures of equity market activity are positively correlated with measures of real activity and that the association is particularly strong for developing countries. The conclusion made by the authors is that “stock market development explains future economic growth” (Alimpiev 2014). Hessling and Paul have mentioned that in the last decades, revolutionary changes in financial markets, instruments, and institutions have stimulated empirical and theoretical investigations into the interaction of the financial and the "real" side of economic systems (Hessling and Paul 2006). The authors mention that while a considerable body An Acad Bras Cienc (2017) 89 (3 Suppl.)

2546

of empirical investigation seems to provide evidence of positive correlations between stock market development and economic growth, there is no consensus in other social sciences as to whether there are two-way linkages, and if so, how to conceive a possible mechanism of interaction. Applying Luhmann's social theory to an analysis of the correlation between finance and economy was a new approach presented by the authors. MATERIALS AND METHODS The development of the economic situation in Russia under the influence of negative external economic factors showed that the economic model that was being created and improved in recent years proved unable to ensure the country's economic security in the face of sharp aggravation of the contradictions in the world financial system (Demeshev and Malakhovskaya 2016). According to a number of assessments, the current financial and economic crisis is managed by the world financial elite. Its main goal is to preserve the dominance of the current elite, to lay the foundations of a new model of world finance in the context of the ongoing global transformations: 1) the exhaustion of the explored hydrocarbon reserves in the near future; 2) the profound climate change leading to the transformation of the geopolitical space; 3) a significant change in the migration flows of the world's population; 4) the emergence of breakthrough technologies in the industry, capable of radically changing the direction and pace of human development. In fact, the organizers of the current crisis face the same tasks that were previously resolved by military force during the First and Second World Wars. In other words, today there is a Third World War and it is conducted by means of economic

2547

MAYA KATENOVA et al.

pressure, blackmail, speculation, and bribery. As a result, the geo-economic picture of the world is fundamentally changing, the needs and the interests of world powers associated with them are changing (Deng 2016). This study employed two macroeconomic variables and one stock market development indicator, which is the Russian Trading System Index (RTS index). The data is of time series format. The macroeconomic factors include money supply in terms of M2 aggregate, Industrial Production Index (IPI). Besides three variables mentioned, the credit spread is employed as an indicator of economic conditions. It has anticyclical nature: credit spread increases during recessions and decreases during an economic boom. The central model employed is VAR. As Bayramova mentioned the vector autoregression (VAR) model is one of the most popular models used for the multivariate time series analysis due to its flexibility and successful forecast capability (Bayramova 2010). All variables employed in this model are tested both on own lags and the lags of other variables. So, it is treated as a theoretical alternative to the structural models. VAR models were popularized in econometrics by Sims, who advocated a non-theoretical way of defining relationships between different time series (Sims 1980). EFFECTS OF MONETARY POLİCY Vector autoregressive (VAR) models were proposed by Sims and can be used to capture the dynamics and the interdependency of multivariate time series. It can be considered a generalization of a system of autoregressive regression models (Sims 1980).

Generally, if Yt = (y1t, y2t…ynt)’ denotes a (nx1) vector of time series variables, the VAR (p) model would look as follows: Yt = C + Σpi=1AiYt-1+ΨDt+εt, t=1, …, T, Where Ai is an (nxn) coefficient matrix and εt is a (nx1) zero mean white noise vector process, C is a vector of constants and Dt is a vector of deterministic variables, such as linear trends, seasonal dummies. Since we examine the changes in linkages among three markets (Russia, the USA, and the EU) we will have to perform a trivariate VAR model. Such a model with 1 lag and no deterministic variable can be depicted as follows:

VAR models enable analyzing multiple time series since they do not require specifying which variables are endogenous and which are exogenous; however, VAR models have some drawbacks. First, in a VAR model, all variables should be stationary, but financial stock/index price series are usually non-stationary; therefore, the VAR model shall be transformed into Vector Error Correction Model (VECM), which drops out the requirement regarding the stationarity of the data. Finally, it is not easy to determine the appropriate lag length in a VAR model (Dikareva and Barannikov 2014). To choose the optimal lag length, two methods are commonly applied. One of them is the likelihood ratio (LR) and another one is the information criteria, such as Akaike (AIC) and Schwartz Bayesian Information Criteria (SBIC). The best model is the one that either maximizes LR or the one that minimizes information criteria. Out of these two methods, information criteria model is more powerful (Brooks, 2002). Out of the two specifications of the information criteria method, we prefer the SBIC specification because it is more An Acad Bras Cienc (2017) 89 (3 Suppl.)

DIRECT AND INDIRECT FINANCE AND ECONOMIC GROWTH

parsimonious, while the AIC will choose on average a model with too many lags. The use of vector autoregressions by Christopher Sims himself was aimed at analyzing the relationships between various macro variables. First of all, he was interested in one of the most discussion macroeconomic issues - the question of the impact of monetary policy on business activity. The source of this discussion was the work of Milton Friedman and Anne Schwartz, in which they argued that the observed high correlation between the money supply and output indicates a unidirectional impact of monetary shocks on the real sector. In turn, this meant that business fluctuations have, first of all, a monetary nature. In addition, the shocks of monetary policy were considered as changes in monetary aggregates (Engle and Granger 2015). In the works of Christopher Sims, devoted to the analysis of monetary policy, one can see a change in his views on the issue under study. In the work of 1972, the existence of not only a high correlation of money supply and output but also a unidirectional relationship characterizing the impact of monetary policy on business activity was revealed (in fact, Sims agreed with Friedman's and Schwartz’s point of view). Conclusions were made on the basis of the main methodological innovation of this work - a direct test for the existence of a cause-effect relationship. But already in the work of 1980, it was shown that when the interest rate is added to the number of regressors, the explanatory power of the variable money supply is significantly reduced. This conclusion contradicted the monetarist concept, according to which monetary policy is responsible for exogenous shocks of the money supply, generating fluctuations in business activity (Lebedeva 2015). In the work of 1983, this idea was further developed: Christopher Sims argues that fluctuations in money reserves occur after (and in response) to a An Acad Bras Cienc (2017) 89 (3 Suppl.)

2548

change in the interest rate. This can be interpreted as a reflection of the endogenous cyclical dynamics of the money supply itself, a concept that became widespread after data in the US after the 1960s and for other countries became available that did not confirm the conclusions of the monetarist theory. This explains why, since 1980, Sims identifies the shocks of monetary policy as innovation (unsystematic changes) in the interest rate, although, in his work of 1972, a monetary proposal was made as a tool of the Fed policy (Polbin 2014). Christopher Sims notes that in equilibrium models with rational expectations one can get a false conclusion about the high predictive power of the money supply. Seven years later, this hypothesis was strongly confirmed: Sims proposed a model of the RBC class (real business cycle), in which the policy of the central bank is presented in the form of a rule for the growth of money supply. However, despite the fact that only performance shocks have a significant impact on the actual release in the model, it predicts a causal relationship according to Granger in the direction from money to release. And this connection is confirmed by the evidence: in the work on the basis of not too strict identification limitations, it was shown that the shocks of monetary policy have a statistically significant and quite strong impact on the real output (Pestova and Mamonov 2016). Alternative to Granger's test. Strictly speaking, none of these two tests allows you to reliably determine the cause or effect of a particular variable dynamics. In particular, the variable xt is called the Granger cause for yt, if the information about the xt lags reduces the prediction error yt + s, s > 0. In this sense, the term "cause-effect relationship", established in Russian literature, is not an absolutely accurate translation of English word "causality" (Niyazbekova 2014).

2549

MAYA KATENOVA et al.

In particular, the results of calculations carried out on post-war American data showed that at the four-year horizon only 4% of the dispersion of the industrial production index is due to the innovations of the monetary aggregate. In the model, real cash balances are not included in the utility function of households, the restriction of cash prepayment is also not introduced. However, monetary policy in the model may affect the economy due to the transaction costs of absorbing liquidity. The impact of monetary policy on real indicators occurs only in the case of a predictable policy affecting inflation expectations and nominal interest rates. Unexpected policy, on the contrary, is neutral (McElroy and Findley 2014). However, Christopher Sims regards this relationship as a purely statistical law, rather than a reflection of the structural economic relationship. From his point of view, due to the fact that interest rates and monetary aggregates are directly related to investment decisions, they change very quickly when new information arrives. In other words, in statistical models, money and interest rates have an explanatory ability for release for the same reasons as asset prices (Martynov 2015). A feature of the work of 1992 is the consideration of a group of five developed countries, which allows the author to find similarities and differences in the transmission mechanism in the United States, Japan and in some European countries. And although the general trends in the interrelation of variables between countries in the sample are easily traced, in some cases inexplicable from the point of view of macroeconomic theory the effects that later became known as “the riddle of prices" and "the riddle of the exchange rate" arise. Their essence consists in the following: tightening monetary policy, considered as an increase in the short-term interest rate, leads to an increase in prices and depreciation of the national currency, while,

according to theory, such a policy should reduce the price level and cause a sharp rise in the price of the currency. The most logical, according to Sims, the explanation of this counterintuitive regularity is the following: the central bank, having reason to expect a rise in prices, increases the interest rate. If this change is not strong enough to completely eliminate the rise in prices, then when conducting econometric research there is a false conclusion that it is the tightening of monetary policy that causes a surge in inflation. The explanation for the exchange rate is similar. An implicit conclusion from these considerations is that even the use of interest rate innovations as an indicator of exogenous monetary policy may not be entirely correct if the specification of the model does not fully reflect the information set of participants in economic processes. The continuation of this work of Christopher Sims was an article co-authored with Eric Liper and Tao Zah, which showed that only a small part of the dispersion of output and prices is explained by monetary shock, while a significant portion of the variation of monetary instruments is due to the systematic reaction of monetary authorities to State of the economy. It is interesting to note that the article suggests several ways of identifying VAR, and the estimation of model parameters is based on the Bayesian method (see below), which allows authors to include up to 18 variables in the model without fear of curse of dimensionality. The work also demonstrated that different ways of identifying the model lead to different quantitative estimates of the effects of monetary policy, but in any case, the importance of monetary shock as a source of recessions remains low (Lomivorotov 2015). RESULTS AND DISCUSSION The data set covers ten years period from 2006 until 2015 on monthly basis. The more frequent the An Acad Bras Cienc (2017) 89 (3 Suppl.)

DIRECT AND INDIRECT FINANCE AND ECONOMIC GROWTH

data, the more accurate results can be obtained from the model. This fact served as the motivation for taking monthly data. The study employed industrial production index as an economic variable and Russian RTS Index as an indicator of stock market performance in Russian Federation. Money Supply was employed as an additional economic parameter. Credit spread was employed as an indicator of economic conditions and it serves as an additional macroeconomic variable. Monthly data were collected in aggregate forms during January 2006 – December 2015 (120 observations) and also for two sectors of industrial productions in levels, stock market index value, money supply and credit spread. The industrial production data were obtained from the Russian Tsentrobank data. These monthly data

2550

are used for Vector Autoregressive (VAR) analysis to identify the direction of causality between the economic and financial variables. The level data are first transformed into logarithmic forms for the industrial production, and stock market index values. The first differenced data, and in some cases monthly seasonal differenced data, are used from the logarithmic data to assure stationarity of data sets in the VAR analysis (Sergeeva and Tayebehsadat 2016). The null hypothesis for ADF cannot be rejected both at a 1% and a 5% level of significance for IPI, Money Supply, RTS Index, and Credit spread. The log values of variables were taken. Only after this step was performed did all data show stationarity at both significance levels.

TABLE I Unit Root Test Unit root and stationarity Critical values IPI Money Supply RTS Index Credit spread

Unit root test Critical values IPI Money Supply RTS Index Credit spread

ADF (1%)

ADF (5%)

PP (1%)

PP (5%)

-3.48 1.26 -1.10 -1.66 -1.89

-2.88

-3.48 1.02 1.08 1.01 -1.24

-2.88

TABLE II Unit Root Test ADF (5%) PP (1%) -2.88 -3,48

ADF (1%) - 3.48 -4.98 -11.82 -10.96 -4.98

As mentioned earlier, the information criteria method is the best way to identify the proper lag length. Table 3 demonstrates the results of the information criteria method, including LM test,

An Acad Bras Cienc (2017) 89 (3 Suppl.)

PP (5%) -2,88

-3.99 -11.62 -11.04 -4.79 Akaike (1974), Schwarz’ Bayesian (1978) information criterion, and Hannan-Quinn information criterion as well. The results help to determine the optimal lag length by looking at the smallest information criterion. The best lag length is

2551

MAYA KATENOVA et al.

the one in which AIC and SBC show the lowest numbers. The minimal value is obtained at lag 1.

Lag 0 1 2 3 4 5 6 7 8 9 10

Log L 1109.76 1326.76 1224.88 1511.42 1578.45 1661.22 1709.54 1753.67 1874 1889.56 1844.55

IPI Money Supply RTS Index CSP

Lag 1 has proven to be the best option.

TABLE III Lag length selection LR AIC NA -21.89 112.76 -24.65 41.89 -22.88 29.06 -25.76 46.54 -27.54 24.78 -25.32 24.57 -23.89 21.09 -23.99 29.67 -23.54 42.09 -23.78 21.38 -23.98

SC -22.01 -22.67 -21.76 -21.02 -20.11 -20.98 -19.33 -18.93 -17.04 -17.84 -16.02

TABLE IV Residuals’ correlation (from VAR model) IPI Lending KASE Index 1 0,07 1 0,09 -0,08 1 0,03 -0,05 0,01

HQ -22.89 -23.43 -23.28 -22.95 -22.11 -22.67 -20.98 -22.76 -20.65 -20.99 -20.72

CSP

1

Dependent variables are listed in the left column and independent variables are listed in the top row. Standard errors in ( ) and t-statistics in { }

There are 120 monthly observations, which cover a ten-year period of time starting in January 2006 and finishing in December 2015. First of all, overall IPI was employed with overall aggregate M2, (money supply), overall credit spread, and overall RTS Index. After that, industrial data was employed based on two separate sectors: the oil and gas sector and the service sector. The data on money supply in a particular sector was obtained with the

RTS Index of the particular sector, the credit spread of the particular sector, and the IPI of the particular sector (Shchepeleva 2014). From the results below, it is obvious that all coefficients are largely insignificant. The results of the VAR model are presented below. The results are not supportive of any causality between the industry of finance and the economy in the Russian Federation (Shevelev 2017).

An Acad Bras Cienc (2017) 89 (3 Suppl.)

DIRECT AND INDIRECT FINANCE AND ECONOMIC GROWTH

2552

TABLE V VAR Estimation Output1

Log_IPI

Log_MS Log_RTS_ Index Log_Credit spread

Log_IPI

Log_MS

Log_RTS_ Index

Log_Credit spread

-0.102934 (0.09394) {-1.102077}

0.092835 (0.09782) {0.949039}

-0.0192932 (0.0189837) {1.016303}

0.0276565 (0.0298272) {0.927224}

0.126525 (0.128726) {0.982902} 0.156253 (0.168726) {0.926075} -0.103287 (0.109827) {0.940452}

0.129876 (0.129827) {1.000377} -0.109827 (0.107283) {-1.023713} -0.097263 (0.099927) {0.97334054}

-0.138726 (0.138246) {-1.003472} -0.117627 (0.117526) {1.000859} -0.0935653 (0.0953627) {0.981152}

-0.129876 (0.127267) {-1.020500} 0.092787 (0.097982) {0.946980} 0.0786562 (0.079765) {0.9860991}

1

Dependent variables are listed in the left column and independent variables are listed in the top row. Standard errors in ( ) and tstatistics in { } An Acad Bras Cienc (2017) 89 (3 Suppl.)

2553

MAYA KATENOVA et al.

The results of the industrial IPI based on the industrial RTS Index and the industrial credit spread

are presented below. The first industry observed is the oil and gas industry.

TABLE VI VAR Estimation Output based on oil and gas industry2

Log_IPI

Log_MS Log_RTS_ Index Log_Credit spread

Log_IPI

Log_MS

Log_RTS_ Index

Log_Credit spread

-0.076567 (0.078726) {-0.972576} 0.054565 (0.050865) {1.072741} 0.078456 (0.073826) {1.062715} -0.086456 (0.089283) {-0.968337}

0.092787 (0.092673) {1.001230} 0.060687 (0.060127) {1.009313} 0.068745 (0.06172) {1.113820} -0.072938 (0.079287) {-0.919924}

0.075476 (0.072576) {1.039958} 0.067567 (0.067816) {0.996328} 0.095643 (0.095624) {1.000198} 0.082783 (0.086276) {0.959514}

-0.082687 (0.083182) {-0.994049} 0.085643 (0.085532) {1.001298} 0.084738 (0.08384) {1.010711} -0.0728736 (0.072736) {-1.000000}

TABLE VII VAR Estimation Output based on servicing industry (Russian data)2

2

Dependent variables are listed in the left column and independent variables are listed in the top row. Standard errors in ( ) and tstatistics in { } An Acad Bras Cienc (2017) 89 (3 Suppl.)

DIRECT AND INDIRECT FINANCE AND ECONOMIC GROWTH

Log_IPI

Log_MS

Log_RTS Index

2554

Log_Credit spread

-0.0634654 0.0854323 0.06453212 0.0643235 Log_IPI (0.062973) (0.0801723) (0.0639847) (0.059182) {-1.0078192} {1.065609} {1.00855548} {1.0868760} 0.0534678 0.0625432 0.07543456 0.0954565 Log_MS (0.052198) (0.062187) (0.0763728) (0.089123) {1.024327} {1.0057279} {0.98771499} {1.0710647} 0.0856453 0.0756434 0.0598766 0.0876566 Log_RTS (0.088473) (0.079828) (0.059827) (0.089827) Index {0.968039} {0.947580} {1.000829} {0.975838} 0.0665456 0.0698456 0.0693234 0.0687567 Log_Credit (0.069827) (0.069182) (0.069283) (0.062738) spread {0.953007} {1.0095921} {1.000583} {1.09593389} From the results above, it is obvious that all level. A unit root test was employed again to check coefficients are largely insignificant again as it was at the first difference. The results showed that it is with the previous results. nonstationary at the first level. The conclusion to be A co-integration test as an additional tool drawn from the test is that there are no long-term The error terms of the cointegration regressions relationships among variables tested at the first level were taken and checked on a unit root test. All of data and at the first difference (Shimanovsky 2014a). them turned out to be non-stationary at the first TABLE VIII The results of unit root test at the first level for error term (Russian data) 1% -1.095423 5% -2.112453 Residual of equation (1) Log_IPI = α 0 + α1Log_Lendingt-1 + α2 Log_RTS _ Index + α 3 Log_Credit Spread + ε it

Residual of equation (2) Log_Lending = α 0 + α1 Log_IPI + α 2 Log_MS+ + εit

An Acad Bras Cienc (2017) 89 (3 Suppl.)

10%

-2.096435

1%

-2.115645

5%

-3.012543

10%

-3.034643

2555

MAYA KATENOVA et al.

There are 120 monthly observations, which cover a ten-year period of time starting in January 2006 and finishing in December 2015. First of all, overall IPI was employed with money supply (M2), and overall RTS Index. After that, industrial data was employed based on two separate sectors: the oil and gas sector and the service sector. From the results below, it is obvious that all coefficients are largely insignificant. The results do not support the hypothesis of mutual causality between financial markets and the economy in Russian Federation (Shimanovsky 2014b). CONCLUSION The results of the study reject all three hypotheses. Based on the results of the study, it can be concluded that stock market does not affect the economic growth in Russia. At the same time, the results support the evidence that the economy does not affect stock market in Russia. The view of causality is not supported in this particular study. The following limitations deserve particular attention. The study employed only ten-year period of time. VAR was employed as a central model in this study. The study employed different industries such as oil and gas and servicing industry besides overall aggregate data. The following theoretical implications can be taken into consideration. Some other models can be employed such as Ordinary Least Squares, Generalized Method of Moments (Sukhanova and Shirnaeva 2015). Such variables as exchange rate, Consumer Price Index can be added in a further research. Also, it should be suggested that the empirical links between stock market development and economic growth warrant further investigation in emerging economies. Focusing on this particular issue and examining the impact of financial liberalization on stock market volatility and the effects of the latter on

investment and growth seem to us to be a promising avenue for future research as well. The most important practical implication is that stock market does not promote economic growth in Russian Federation and the economy does not affect the stock market. Assessing the situation on the Russian stock market during the development of crisis phenomena, and taking into account the lack of effect from the actions of regulatory bodies, it can be assumed that in the near future the Russian stock market will completely fall under the influence of world exchange centers, which will sharply negatively affect the sovereignty of the country's economy generally (Vashcheliuk et al. 2015). Even today, foreign investors have a determining influence on the state of the Russian stock market. According to expert estimates, they account for the bulk of turnover on shares of Russian companies (in foreign markets - almost entirely, on Russian sites - up to half). Trade in Russian assets goes to foreign markets. In the conditions of full convertibility of the ruble and the increasing integration of Russia into the world economy, the influence of foreign investors on the Russian market will intensify. At the same time, such an impact can be not only positive but also negative, especially taking into account the existing weaknesses in the architecture of the Russian stock market (Yershov and Kadreva 2015). Under these conditions, the question arises of the possibility and expediency of preserving the national financial sovereignty of the country, its economic security, primarily due to the qualitative enhancement of competitiveness of the Russian financial market and the formation of an independent financial center on its basis capable of concentrating a wide range of financial instruments, demand for financial instruments by domestic and foreign investors. But before talking about possible An Acad Bras Cienc (2017) 89 (3 Suppl.)

DIRECT AND INDIRECT FINANCE AND ECONOMIC GROWTH

options for the development of the Russian financial system in the future, it is necessary to highlight the features and principles of the world financial centers’ organization that can serve as the basis for the formation of national financial institutions in Russia. The key link in the financial market of any modern state is the stock market, which creates conditions for effective mutual organized trade in financial instruments between its participants. The purpose of the stock market functioning - as well as of all financial markets - is to provide a mechanism for attracting investments in the economy by establishing the necessary contacts between those who need money and those who would like to invest surplus income. The stock market is a free market, and it will fulfill its tasks of constant maintenance of economic growth only if there is complete freedom of investments’ movement (Zimin 2014). In the global economic community, global stock markets dominate. They are characterized by the largest capitalization, significant volumes of trading, the highest liquidity. In such markets are circulating securities and derivative financial products of the most famous corporations, as well as depositary receipts and secondary placements of foreign companies. Obligatory participants in markets of this scale are stock exchanges - organizers of civil transactions with securities. The clients of the exchange are both national and foreign investors from all over the world. REFERENCES ALIMPIEV, Y. (2014). Methodological problems of modeling the financial and monetary transmission. Journal of Economic Regulation, 5(2), 124-132. AREFIEV, N. (2016). Partial identification of monetary rule using restrictions on lagged effects. Economic Journal of the Higher School of Economics, 20(3), 500-512.

An Acad Bras Cienc (2017) 89 (3 Suppl.)

2556

AREFIEV, N. K. (2015). From correlation to causation: Computer Science Approaches. Economic Journal of the Higher School of Economics, 19(3), 457-496. ATHANASOGLOU, P., IOANNIS, D., & MANTHOS, D., (2013). Bank procyclicality and output: issues and policies, Working Paper, Unpublished, Munich Personal RePEc Archive. BAYRAMOVA, B. (2010). Empirical test on macroeconomic factors and stock market analysis: Case of Kazakhstani stock market, Dissertation. United Kingdom: Lund University. CALDERON, С. &. (2002). The Direction of Causality between Financial Development and Economic Growth. Working Papers, Central Bank of Chile(184). CHUKWU, J. &. (2009). Multivariate Causality between Financial Depth and Economic Growth in Nigeria. African Review of Money, Finance and Banking, 7-21. DEMESHEV, &. B. (2010). Macroeconomic Forecasting with a Litterman’s BVAR Model. Economic Journal of the Higher School of Economics, 20(4), 691-710. DENG, M. (2016). Bayesian Variable Selection in a Large Vector Autoregression for Origin-Destination Traffic Flow Modelling. Spatial Econometric Interaction Modelling, 199-223. DIKAREVA, I. B. (2014). Complex analysis of the mechanism of functioning of the Russian stock exchanges. Polythematic network electronic scientific journal of the Kuban State Agrarian University, 96, 1-27. ENGLE, R. G. (2015). The co-integration and error correction: representation, estimation, and testing. Applied Econometrics(3), 106-135. HESSLING, P. (2006). The Global System of Finance: Scanning Talcott Parsons and Niklas Luhmann for Theoretical Keystones. The American Journal of Economics and Sociology, 65(1), 189-218. HONDROYIANNIS, G. P. (2001). Macroeconomic influences on the stock market. Journal of Economics and Finance, 25(1), 33-49. IBRAHIM M.H., A. H. (2003). Macroeconomic variables and the Malaysian equity market: A view through rolling subsamples. Journal of Economic Studies, 30(1), 6-27. LANDAU, J. (2009). Procyclicality – what it means and what could be done, Unpublished. Bank of Spain’s conference on procyclicality and the Role of Financial Regulation. LEBEDEVA, K. (2015). An empirical analysis of the Russian financial markets’ liquidity. Review of Business and Economics Studies(3), 5-31. LEVINE, R. &. (1996). Stock market development and long-run growth. World Bank Economic Review, 10(2), 323-339.

2557

MAYA KATENOVA et al.

LOMIVOROTOV, R. (2015). The use of Bayesian methods for analysis of monetary policy in Russia. Applied Econometrics, 2 (38), 41-62. MARTYNOV, S. (2015). Indexes of the stock market in the system of indicators of the financial market status. The territory of Science, 4, 191-196. MCELROY, T. &. (2014). Fitting constrained vector autoregression models. Empirical Economic and Financial Research, 451-470. NIYAZBEKOVA, S. (2014). The stock market creation and development in the Russian Federation and the Republic of Kazakhstan. Bulletin of Orenburg State Agrarian University, 1, 188-191. PESTOVA, A. &. (2016). The assessment of the impact of various shocks on the dynamics of macroeconomic indicators in Russia and the development of conditional forecasts based on the BVAR model of the Russian economy. Economic politics 4(11), 56-92. POLBIN, A. (2014). Macroeconomic effects from investment’s efficiency growth: the analysis of a general equilibrium model. The Russian Entrepreneurship, (13), 64-72. RASHID, A. (2008). Macroeconomic variables and stock market performance: testing for dynamic linkages with a known structural break. Savings and Development, 32(1), 77-102. SERGEEVA, I. &. (2016). The impact of financial market industry on industry growth: an evidence from Russia and Iran. Bulletin of Perm University, 4 (31), 66-73. SHCHEPELEVA, M. (2014). Measuring Systematic Risk in the Financial Sector. Bulletin of MGIMO University, 6, 117125. SHEVELEV, A. (2017). Bayesian approach to evaluate the impact of external shocks on Russian macroeconomics indicators. The World of Economics and Management, 17(1), 26-40. SHIMANOVSKY, D. (2014a). Taking into account non-price competition in the course of forecasting Russian corporate crediting market. Bulletin of Perm University, (4), 24-31. SHIMANOVSKY, D. (2014b ). The use of indicators non-price lending terms in the process of short-term forecasting of the Russian credit market. Gaudeamus, (2), 208-211. SIMS, C. (1980). Macroeconomics and Reality. Econometrica, 48, 1-48. SUKHANOVA, Y. &. (2015). The use of econometric methods for research of stabilization processes of the Russian economy. Regional development: electronic scientific and practical journal, 4 (8), 1-6.

VASHCHELIUK, N. P. (2015). Estimation of the Monetary Policy Shock’s Influence on the Russian Economy. Economic Journal of the Higher School of Economics, 19(2), 169-198. YERSHOV, E. &. (2015). Organized Savings Modeling of Russian Households’: Macro-approach, Credit Consideration. Economic Journal of the Higher School of Economics, 19(3), 349-385. ZIMIN, V. (2014). The investment activities of the stock market. The army, and society, 42(5), 1-10.

An Acad Bras Cienc (2017) 89 (3 Suppl.)