Financial Reforms,Capital Investment and Financial Intermediation in China

Abstract

China has witnessed remarkable changes in its capital investment and financial system since initiating economic and financial sector reforms more than three decades ago. However, there is a dearth of studies examining what impact these reforms have had on financial intermediation, measured by credit growth, in the country. This article addresses this vacuum and investigates the effect of financial sector and capital investment reforms on credit growth in China between 1986 and 2016. We examine how real interest rate (the financial reform indicator) and gross fixed capital formation (the economic capital investment indicator) are linked with financial intermediation in China. Our empirical results suggest that although gross fixed capital formation positively influences credit growth, there is no evidence that real interest rates influence credit growth in China. The main message is that credit has grown in China, not because of financial intermediation but because of the increased need to finance growing fixed capital investment.

JEL Classification: E43, E44, F65

Keywords

Financial reform credit growth real interest rate financial repression

Introduction

The financial sector and economic capital play a pivotal role in the functioning of an economy. To ensure general stability and development of the financial sector and fixed asset investment, reforms are required from time to time. Financial sector reforms refer to the process of partially or completely moving away from a system characterized by government regulations to one based on market determination. Reforms in capital investment refer to general policies dedicated towards encouraging productive capital investment. China has had a high capital investment and less regulated interest rates since the period of reforms.

These reforms are generally considered positive for the economy as they engender innovation and promote efficiency in the financial and economic system. However, the ills of a repressive financial system are well understood and documented in the literature. Not only is a repressive financial system inimical to economic growth and development, it also fails to engender a well-developed financial system. The concept of financial repression was first introduced by McKinnon (1973), who defined financial repression as financial policies that are characterized by high reserve requirement on bank deposits, interest rates regulation and compulsory allocation of resources.

Prior to the 1970s, the financial markets in many countries, especially developing countries like China, were highly repressed by the government. These repressive policies are thus expected to hinder financial deepening and impact economic growth negatively (McKinnon, 1973; Shaw, 1973). According to McKinnon (1973) and Shaw (1973), financial liberalization is the backbone of economic reforms, and this implies a higher level of real interest rate that equates demand and supply of savings. They further postulated that a rise in real interest rates would lead to increased financial activities, a more efficient use of savings via financial intermediation and an increase in savings to support credit growth. Over the past few decades, the world has become more integrated and this has led to the increasing scale of financial turbulence globally. The recent global financial crisis of 2007–2008 is worthy of mention because of its attendant impact on major economies around the globe. To reduce the risk or likelihood of financial crises, regulators continually introduce reform measures aimed at ensuring the general stability of the financial system. These measures are also expected to reduce economic volatility, increase transparency and greatly enhance the resilience of the financial system. In China, efforts to reform the financial system and capital investment started majorly with the adoption of widespread reforms around 1978. Prior to this period, the country had a highly repressed financial system and relatively lower ratio of capital investment.

Before 1978, interest rates on deposits and loans were directly fixed by the central government. In addition, the interbank market, bond market and stock market were non-existent. These reforms constituted part of the government’s plan to transform China from a centrally planned economy to one with a role for both the government and the private sector. The financial reforms were sustained into the 1990s but took a more gradual approach from the 2000s due to the crisis associated with non-performing loans, increased banking sector fragility and the Asian financial crisis. This crisis highlights the potential dangers of widespread financial reforms. In the aftermath of the crisis, the Chinese government tightened its control of the foreign exchange market. The Global Financial Crisis (GFC) also led to the introduction of several reform measures. In its wake, the government mobilized bank loans and introduced a massive stimulus package to support the economy. The importance of finance for growth was recognized by Chinese policymakers quite early and, consequently, the government undertook the task of building a modern financial system. These concerted efforts resulted in the phenomenal growth of financial infrastructure and a much better-developed financial system that we see in China today. Also, high capital investment in China is an important ingredient for China’s high rate of economic growth in the last decades.

However, despite successive years of financial sector reforms, the Chinese financial system is not fully reformed when compared to some of its emerging market peers such as Brazil, Russia, etc. To this day, the Chinese central bank still, to some extent, fixes rates for commercial banks and intervenes heavily in foreign exchange markets. Also, the Chinese authorities still exercise some measure of financial account controls, particularly over foreign direct investment, portfolio investment and debt financing. This has, however, not completely dampened the financial development of China. On the contrary, the financial sector in China has grown from a single-bank system to a well-developed system comprising several types of financial institutions, though much of the developments in the financial system are majorly in the banking sector.

For instance, China’s broad money supply is larger than that of the USA and the total bank deposits accounted for 190 per cent of GDP in mid-2011, while stock market capitalization was 80 per cent of GDP. Between 1986 and 2016, credit to the private sector as a per cent of GDP averaged 105.52 per cent, with a minimum of 65.33 per cent in 1986 and a maximum of 156.7 per cent in 2016. The ratio of broad money supply to GDP, which measures the depth of the financial sector, stood at 193 per cent in 2014, a figure that is well ahead of that of many emerging and developed economies. Though the Chinese financial system has been able to achieve remarkable developments in the last 30 years, more reforms are still needed to ensure complete market determined financial outcomes, especially for interest rates and exchange rates.

Compared to several decades earlier, there is no doubt that the Chinese financial system has witnessed several reforms and grown by leaps and bounds. But despite the phenomenal growth recorded, empirical examination of the impact of these reforms on credit growth, especially the effect of financial reforms—as hypothesized by McKinnon (1973) and Shaw (1973)—and the effect of reforms in capital investment remain relatively scarce. In addition, there is a dearth of empirical evidence on what these reforms portend for credit growth in China. The few available studies have mostly investigated the impact of financial reforms on economic growth and productivity and utilized measures of financial reforms like dummy variables, which do not capture all facets of financial reforms. To address this shortcoming, our article adopts real interest rates as financial reform variable and gross fixed capital formation as capital investment reform variable. We probe the impact of these variables on credit growth in China. The rest of the article is structured as follows: the second section presents the literature review, the third section describes the data and methodology, while the fourth and fifth sections present the empirical results and conclusion, respectively.

Literature Review

Economists have long acknowledged the central roles that the financial sector plays in the development process (Bencivenga & Smith, 1991; King & Levine 1993; Shaw, 1973). According to conventional wisdom and theory in this area, a well-developed and functional financial system promotes economic growth by providing a framework and solid institutions that support the pooling of savings, clearing and settling of payments and the reduction of information costs (Merton & Bodie, 1995). In the literature, many empirical and theoretical studies have been devoted to testing the McKinnon (1973) and Shaw (1973) hypothesis. The hypothesis centres around the effect of real interest rates on borrowers and lenders. According to McKinnon (1973), a financial system that is free from repression allows borrowers and savers to operate in line with the dictates of the market. In a highly repressed financial system, fixing interest rates will create a wedge between social and private returns and consequently shift savings towards the acquisition of real assets in addition to creating a bias towards current consumption. Several researchers have also expressed contrary opinions. Wijnbergen (1982) and Taylor (1988) postulate that higher interest rates may impede financial deepening which may adversely affect economic growth in developing countries. They further argue that the effect of higher interest rates also depends on the economic agent in question. A high interest rate discourages borrowers from accessing loans while, to the lender, the same will increase savings as savers will be able to get more income from their savings. Despite the various arguments for or against financial reforms, the widely held view is that its benefits far outweigh its potential dangers.

On the empirical side, several studies have tested this hypothesis and investigated related issues in the literature. The studies by Chung, Smith and Wu (2009) and Chandar, Patro and Yezegel (2009) reported that financial reforms stimulate healthy competition and promote financial market efficiency by ensuring that borrowers and savers operate in line with the dictates of the market. A similar study was also conducted by Khalaf (2011) for the case of Iraq, with empirical results supporting the widely held view that interest rate deregulation stimulates savings, increases financial assets and engenders financial deepening in the long run. In the case of Nigeria, Orji, Aguegboh and Anthony-Orji (2015) reported that financial liberalization promotes the real sector of the economy, while, on the contrary, Akingunola, Olusegun, Oluwaseyi and Olusoji (2013) reported that the effect of financial liberalization—proxied by real interest rate, the ratio of liquid liabilities and total deposits—on economic growth is not significant. In the study by Adam (2011) for the Ghanaian economy, a financial liberalization index constructed using the Principal Component Analysis was employed and empirical results reported positive and long-run relationships between financial liberalization and economic growth.

For the South African economy, Odhiambo (2010a) finds evidence that interest rate positively impacts financial development, but financial development was found not to impact economic growth and investment positively. Odhiambo (2011) also investigated the subject matter in South Africa, Lesotho, Zambia and Tanzania and empirical results reported that even though financial liberalization leads to financial development in all selected countries, it is only in Zambia that financial liberalization granger causes economic growth, while for the other countries it is economic growth that influences financial sector development. A related study by Hye and Wizarat (2013) also provided further evidence with empirical results revealing a negative relationship between interest rate and economic growth in the long run, and a positive relationship between the financial liberalization index and economic growth in the short run. In the empirical evidence by Aftab, Jebran, Ullah and Awais (2016), a negative relationship is reported between interest rate and private sector credit, while the study by Adeleye et al. (2017) provides evidence in support of the McKinnon (1973) and Shaw (1973) hypothesis.

For China, a few empirical studies have also been carried out. Li (1994) documents that the low interest rate ceiling in China has led to financial and investment inefficiency. A similar view was also reported in the empirical study by Lardy (1998), who finds that setting low lending rates results in excess demand for loans, which consequently leads to an inefficient financial system. Using Chinese provincial data, Park and Sehrt (2001) empirically examined whether financial reforms in the mid-1990s increased financial intermediation in China. Their results show that financial reforms do not improve the process of financial intermediation. In the study by Dong, Wang, and Xiangrong (2014), a nonlinear relationship between interest rate liberalization and bank liquidity creation was reported, where an improvement in interest rate liberalization increases bank liquidity creation at first, then decreases thereafter. Gang, Bose, and Chen (2015) examine the financial distortions–growth nexus in China by employing a general equilibrium model. They find that financial reform drives output gains and that lending by state-owned banks fails to reduce during the study period and lending by financial institutions also fails to respond to economic fundamentals. For capital investment and credit nexus, Chen and Kang (2018) investigate how fixed asset investment influences loan growth in China and find a positive relationship which signifies that more credit has flown to Chinese provinces relying more on fixed asset investment. However, their study is viewed from a provincial perspective and does not extend to as far back as the periods when China initiated a push for higher capital investment.

The survey of the literature reveals the dearth of studies addressing the impact of the identified reforms on credit growth in China on a macro level over the last decades. To this end, our study addresses this gap in the literature. By doing so, we contribute to the existing literature on the Chinese economy.

Data and Methodology

Data

The data samples cover annual time series from 1986 to 2016. Data on real interest rate (INTR), domestic private sector credit/GDP (CR), inflation rate (INFL), GDP per capita growth (PGDPG) and growth rate of gross fixed capital formation (DINV) were sourced from the World Development Indicators of the World Bank (2017), while data on financial system deposit (FSD) was from the Global Financial Development Indicators of the World Bank (2017). In this article, we proxy credit growth with domestic credit to the private sector provided by financial institutions as a percentage of GDP. This measure captures the amount of credit facilities available to firms and households, excluding public sector credits. Real interest rate refers to the lending rate adjusted for inflation. Financial system deposits capture the volume of financial liquidity, per capita GDP growth measures economic growth, while the growth rate of gross fixed capital formation measures the growth rate of domestic investment and real sector activities.

In the empirical framework, we specify credit as a function of real interest rate and gross fixed capital formation. This is can be expressed as:

C R_{t} = f (I N T_{t}, D I N V_{t})

(1)

where CR_t denotes credit and INT_t is the rate of interest.

Following Adeleye et al. (2017), we introduce four additional variables namely, inflation (INFL), financial system deposit (FSD), per capita GDP growth (PGDPG) and gross fixed capital formation (DINV). Accounting for these variables, we respecify Equation (1) as:

C R_{t} = f (I N T_{t}, D I N V_{t}, I N F L_{t}, F S D_{t}, P G D P G_{t})

(2)

To test our hypothesis, and the broader predictions of McKinnon (1973) and Shaw (1973), both credit growth and real interest rates serve as dependent variables. It should be noted that the inclusion of the explanatory variables in this study is in line with previous studies and follows from the Keynesian and neoclassical financial liberalization theories (Molho, 1986), as well as the analytical and theoretical frameworks of McKinnon (1973) and Shaw (1973). According to McKinnon (1973) and Shaw (1973), the increase in the pool of funds from depositors encourages more financial intermediation. Thus, an increase in financial system deposits implies more loanable funds being available for financial intermediation and, hence, more supply of credit. Per capita GDP growth is included to capture economic growth, in line with Chen and Kang (2018). The inclusion of per capita GDP growth helps to capture two strands of argument in the finance–growth literature: the demand-following hypothesis, which posits that economic growth leads to financial development, and the supply-leading hypothesis which posits that finance leads to growth (Chuah & Thai, 2004; Patrick, 1966; Robinson, 1952). Gross fixed capital formation is included because it captures the real sector. Inflation is included to control for the average price level in the economy.

Empirical Modelling

Unit Root Test

Before carrying out the empirical analysis, the first step is to determine the stationary properties of the variables. In this study, the unit root test is conducted using several unit root tests such as the Augmented Dickey Fuller (ADF), Phillips–Perron (PP), ADF–GLS, NG–Perron and Kwiatkowski–Phillips–Schmidt–Shin (KPSS) tests. The Augmented Dickey Fuller approach accounts for the autocorrelation of the first differences of a series in a parametric fashion by estimating additional nuisance parameters while the Phillips–Perron unit root test is robust even in the presence of serial correlation (Gujarati & Porter, 2009). The ADF test has weak power when the sample size is small; hence, the Phillips–Perron (PP), ADF–GLS, NG–Perron and Kwiatkowski–Phillips– Schmidt–Shin (KPSS) unit root tests will help provide more reliable results. Given the difficulty in establishing that all variables are I(1), as the variables exhibit a mixture of I(0) and I(1) properties (see Table 2), we employ the ARDL bounds testing, described below, to investigate the extent of cointegration and the type of long run relationships among the variables.

The ARDL Bounds Testing

We employ the Auto Regressive Distributed Lag (ARDL) model of Pesaran, Shin and Smith (2001) to analyse the financial reform–credit growth nexus in China. This is with a view to test the major assertions of the McKinnon (1973) and Shaw (1973) hypothesis. Following Adeleye et al. (2017), we formulate two ARDL models where both credit growth and real interest rate serve as dependent variables. In the models in Equations (3) and (4), credit growth is the dependent variable, while for the models in Equations (5) and (6), real interest rate is the dependent variable. The models are stated as follows:

\begin{array}{l} \begin{matrix} Δ \ln C R_{t} = α_{0} + \sum_{j = 1}^{p} ϑ_{j} Δ \ln C R_{t - j} + \sum_{j = 0}^{q} δ_{j} Δ I N T R_{t - j} \\ + \sum_{j = 0}^{m} φ_{j} Δ I N F L_{t - j} + \sum_{j = 0}^{a} γ_{j} Δ F S D_{t - j} + \sum_{j = 0}^{b} θ_{j} Δ P G D P G_{t - j} \end{matrix} \\ \begin{matrix} + \sum_{j = 0}^{c} σ_{j} Δ D I N V_{t - j} + β_{1} \ln C R_{t - 1} + β_{2} I N T R_{t - 1} + β_{3} I N F L_{t - 1} \\ + β_{4} F S D_{t - 1} + β_{5} P G D P G_{t - 1} + β_{6} D I N V_{t - 1} + μ_{t} \end{matrix} \end{array}

(3)

By reparametrizing Equations (3) and (5), we obtain short-run coefficients by estimating an error correction model associated with the long-run estimation. These are specified in Equations (4) and (6):

\begin{matrix} Δ \ln C R_{t} = α_{0} + \sum_{j = 1}^{p} ϑ_{j} Δ \ln C R_{t - j} + \sum_{j = 0}^{q} δ_{j} Δ I N T R_{t - j} + \sum_{j = 0}^{m} φ_{j} Δ I N F L_{t - j} \\ + \sum_{j = 0}^{a} γ_{j} Δ F S D_{t - j} + \sum_{j = 0}^{b} θ_{j} Δ P G D P G_{t - j} + \sum_{j = 0}^{c} σ_{j} Δ D I N V_{t - j} \\ + ψ E C M_{t - 1} + μ_{t} \end{matrix}

(4)

\begin{array}{l} \begin{matrix} Δ I N T R_{t} = α_{0} + \sum_{j = 1}^{p} ϑ_{j} Δ I N T R_{t - j} + \sum_{j = 0}^{q} δ_{j} Δ \ln C R_{t - j} + \sum_{j = 0}^{m} φ_{j} Δ I N F L_{t - j} \\ + \sum_{j = 0}^{a} γ_{j} Δ F S D_{t - j} + \sum_{j = 0}^{b} θ_{j} Δ P G D P G_{t - j} + \sum_{j = 0}^{c} σ_{j} Δ D I N V_{t - j} \end{matrix} \\ \begin{matrix} + β_{1} I N T R_{t - 1} + β_{2} \ln C R_{t - 1} + β_{3} I N F L_{t - 1} + β_{4} F S D_{t - 1} \\ + β_{5} P G D P G_{t - 1} + β_{6} D I N V_{t - 1} + μ_{t} \end{matrix} \end{array}

(5)

\begin{matrix} Δ I N T R_{t} = α_{0} + \sum_{j = 1}^{p} ϑ_{j} Δ I N T R_{t - j} + \sum_{j = 0}^{q} δ_{j} Δ \ln C R_{t - j} + \sum_{j = 0}^{m} φ_{j} Δ I N F L_{t - j} \\ + \sum_{j = 0}^{a} γ_{j} Δ F S D_{t - j} + \sum_{j = 0}^{b} θ_{j} Δ P G D P G_{t - j} + \sum_{j = 0}^{c} σ_{j} Δ D I N V_{t - j} \\ + ψ E C M_{t - 1} + μ_{t} \end{matrix}

(6)

Δ denotes the first difference operator, ₀ is the drift component, (p,q,m,a,b,c) are the lag lengths, which will be chosen using lag selection criterion, and µ_t is the error term. The coefficients (ϑ_j,δ_j,ɸ_j,γ_j,θ_j,σ_j) represent the short-run effects, while the coefficient ψ associated with error correction mechanism (ECM) allows for adjustment back to the long-run equilibrium following short-run deviations.

Empirical Results

Table 1 presents results of the descriptive statistics and correlation matrix. From the table, it is reported that the mean and median of the variables have values within the minimum and maximum values, thus implying that all the variables exhibit a high level of consistency. For the correlation matrix, results reveal that the linear dependence among the variables is not exact.

Table 1.

Descriptive Statistics and Correlation Matrix Result

	CR	INTR	INFL	FSD	PCGDP	DINV
Mean	107.4647	1.928006	67.65292	34.24097	8.600403	11.39356
Median	109.1584	2.729220	71.28211	38.25000	8.782185	10.46976
Maximum	156.7061	7.348178	102.0000	50.76000	13.63634	34.31494
Minimum	73.98362	−7.97725	22.70805	17.82000	2.393612	−13.6934
Std. Dev.	22.28277	3.439623	23.61834	11.17895	2.565855	7.714959
CR	1
INTR	0.296685	1
INFL	0.886585	0.207668	1
FSD	0.868891	0.147074	0.884085	1
PCGDP	−0.12422	−0.54889	−0.00058	0.10733	1
GFCF	−0.01597	−0.30156	−0.07068	0.016766	0.750466	1

Source: Authors calculations.

To ascertain the stationarity properties of the variables, this study employs graphical illustrations as well as standard unit root tests. From the graphical illustration in Figure 1, it is seen that credit growth (CR) and financial system deposit (FSD) exhibit unit roots, while real interest rate appears stationary. Tables 2 and 3 present results of the unit root tests using Augmented Dickey Fuller (ADF), Phillips–Perron (PP), ADF–GLS, NG–Perron and Kwiatkowski–Phillips–Schmidt–Shin (KPSS), both at level and at first difference. From the unit root test result, it is observed that real interest rate (INTR), per capita GDP growth (PGDPG) and gross fixed capital formation (DINV) are stationary at level while other variables become stationary at first difference. Hence, since we have a combination of I(0) and I(1) variables, the ARDL technique becomes an appropriate analytical tool. The main advantage of the ARDL technique is that it can be applied especially when variables have different orders of integration, that is, irrespective of whether the variables are I(0) or I(1). The technique also improves the small sample properties of the estimates and allows different optimal lags to be used for the variables.

Figure 1.

Source: The author.

Table 2.

Result of Unit Root Tests (ADF and PP)

	ADF			PP
Variables	Level	First Diff.	Status	Level	First Diff.	Status
lnCR	0.079302	−5.01842	I (1)	0.187417	−5.00499	I (1)
	0.9586	0.0003***		0.9671	0.0003***
INTR	−3.08905		I (0)	−3.16725		I (0)
	0.0382			0.0322**
INFL	−2.55781	−4.88768	I (1)	−2.64958	−5.24025	I (1)
	0.1131	0.0005***		0.095	0.0002***
FSD	−1.41989	−3.64197	I (1)	−2.39479	−5.12547	I (1)
	0.5582	0.0112***		0.1519	0.0003***
PGDPG	−2.52715	−4.2763	I (1)	−2.01128	−6.39897	I (1)
	0.1194	0.0023***		0.2807	0.0000***
DINV	−3.32086		I (0)	−2.88637	−9.68314	I (1)
	0.0228**			0.0588	0.0000***

Source: Authors calculations.

Note: *** and ** indicate significance at 1% and 5% level, respectively.

Having adopted the ARDL approach as the appropriate technique for this study, we proceed to determining the optimal lag length to be used for the analysis. Due to its small sample properties, the optimal lag length is selected based on the outcome of the Akaike information criterion (AIC). The criterion is more appropriate and performs well when we are dealing with small samples; however, it is inconsistent when the samples are large (Acquah, 2010). In addition, all lag selection criteria agree on the choice of three as the optimal lag length, and this holds for when credit growth and real interest rate are the dependent variables. Thus, the lag length of three is chosen for this study based on the outcome of the Akaike Information Criterion. Table 4 presents the results of the lag selection criteria.

Table 3.

Result of Unit Root Tests (ADF–GLS, NG–Perron and KPSS)

Variables	ADG-GLS			NG-Perron			KPSS
Variables	Level	First Diff.	Status	Level	First Diff.	Status	Level	First Diff.	Status
lnCR	40.94269	1.595466***	I (1)	26.9855	1.70098***	I (1)	0.702447	0.131903###	I (1)
INTR	2.115819	1.620805***	I (1)	2.1948	1.72381***	I (1)	0.099776###		I (0)
INFL	190.6743	2.652451**	I (1)	9.66534	2.74859**	I (1)	0.694869	0.16715###	I (1)
FSD	55.79831	1.277018***	I (1)	0.49554***		I (0)	0.654971	0.193192###	I (1)
PGDPG	2.826067**		I (0)	2.87474**		I (0)	0.099261###		I (0)
DINV	2.010929**		I (0)	2.08311	1.71752***	I (1)	0.06958###		I (0)

Source: Authors calculations.

Note: *** and ** indicate significance at 1% and 5% level, respectively. ### Indicates the failure to reject the null hypothesis at 1% level of significance.

Table 4.

Lag Length Selection Criteria

Lag	LogL	LR	FPE	AIC	SC	HQ
0	−424.295	NA	2804160.	31.87367	32.16164	31.95930
1	−297.989	187.1195	3750.931	25.18436	27.20011	25.78375
2	−234.14	66.21342	746.4980	23.12150	26.86503	24.23465
3	−114.882	70.67160*	7.226167*	16.95422*	22.42553*	18.58112*

Source: Authors calculations.

Note: * Indicates lag order selected by the criterion; LR, FPE, AIC, SIC and HQ indicate sequential modified LR test statistic, Final Prediction Error, Akaike Information Criterion, Schwarz Information Criterion and Hannan–Quinn, respectively.

Having chosen an appropriate lag length, the next step is to perform the ARDL bounds test for cointegration. Table 5 presents the bounds test result, where the null hypothesis is no cointegration. For the case where credit growth (CR) is the dependent variable, the F-statistic (11.04) is above the upper critical bound (4.15) at the 1 per cent level of significance, while for the case where real interest rate (INTR) is the dependent variable, the F-statistic (7.71) is above the upper critical bound (4.15) at the 1 per cent level of significance. Thus, in both cases, a long-run relationship among the variables exists since the null hypothesis of no cointegration is rejected at the 1 per cent level of significance. We can therefore conclude from the bounds test that a long-run relationship should exist among the variables. Having established cointegration, we next estimate the cointegrating relationship to capture the type of long-run relationship that binds the variables together.

Table 5.

ARDL Bounds Test

F-statistic	No. of Regressors	Critical Values	L(0)	U(1)
Dependent variable: Credit growth (CR)
11.0377	5	10%	2.08	3
		5%	2.39	3.38
		1%	3.06	4.15
Dependent variable: Interest rate (INTR)
7.7094	5	10%	2.08	3
		5%	2.39	3.38
		1%	3.06	4.15

Source: Authors calculations.

The results of the ARDL estimation when credit growth and real interest rate are the dependent variables are presented in Tables 6 and 7, respectively. In Table 6, empirical results reveal that in the long run, real interest rate and financial system deposit have a statistically insignificant positive effect on credit growth, while inflation has an insignificant negative effect on credit growth. Interestingly, increases in fixed capital investment are found to be linked with an increase in credit growth in China.

Table 6.

ARDL Long-run and Short-run Estimates (dependent variable: credit growth)

Variable	Coefficient	Std. Error	t-Statistic	Prob.
Long-run rstimates
INTR	0.077755	0.050056	1.553354	0.1388
INFL	−0.00116	0.005934	−0.19619	0.8468
FSD	0.021461	0.013081	1.640626	0.1192
PCGDP	−0.11289	0.050868	−2.21927	0.0404**
DINV	0.038157	0.018076	2.110895	0.0499**
C	4.349211	0.199493	21.80132	0.0000***
Short-run parsimonious estimates
D (lnCR(−2))	−0.21518	0.071777	−2.99797	0.0081***
D (INFL)	0.02195	0.004518	4.858294	0.0001***
D (INFL(−1))	−0.00861	0.003551	−2.42587	0.0267**
ECM(−1)	−0.24688	0.024147	−10.2241	0.0000***
R ²	0.88272
Adjusted R²	0.86232

Source: Authors calculations.

Note: *** and ** indicate significance at 1% and 5% level, respectively.

These results are not surprising given that the Chinese financial system is yet to be fully reformed. For instance, Chinese central bank still regulates rates and the authorities exercise strict capital account controls, particularly over foreign direct investment, portfolio investment and debt financing. Thus, the lack of conclusive evidence to show that real interest rate and financial system deposit have a positive effect on credit growth is because the Chinese financial system, despite being well developed, is yet to witness full financial reforms. Empirical results also reveal that per capita GDP growth has a negative effect on credit growth at the 5 per cent level of significance while gross fixed capital formation (DINV) has a positive effect on credit growth also at the 5 per cent significance level in the long run.

The error correction mechanism (ECM) which captures the speed of adjustment of the model to equilibrium in the event of shocks is found to be negative and statistically significant (−0.24688). The value indicates that about 25 per cent disequilibrium of the previous year’s shocks is corrected back to the long-run equilibrium in the current year. Short-run results also reveal that the current and previous year’s inflation rates (at lag 2) have a positive and negative effect, respectively, on credit growth, while credit growth in the current period is negatively influenced by credit growth in the previous period (at lag 2). The other short-run coefficients are not reported because they are not significant. Lastly, the R² and adjusted R² are in the right magnitude. For the R², the coefficient is approximately 88 per cent, indicating that a higher variation in credit growth is explained by the variables in the model, while the adjusted R² indicates that the variables in the model explain about 86 per cent of the total variation in credit growth. This study does not find any statistically significant evidence that financial reform measured by real interest rate is positively linked with credit growth. To this end, the assertion by McKinnon (1973) and Shaw (1973) that higher real interest rate brings improvement in the process of financial intermediation, cannot be supported in China. However, the apparent disconnect between credit growth and interest rate in China can still be sustained in the Chinese economy because the capital investment, which China has engaged in in recent years, has been a significant positive driver of credit growth as our results have shown. Thus, even if we find no evidence that financial reform has driven credit growth in China, our result suggests that the rapid expansion in China’s fixed capital investment is an important channel that supports credit growth in the economy.

Table 7.

ARDL Long-run and Short-run Estimates (Dependent Variable: Interest Rate)

Variable	Coefficient	Std. Error	t-Statistic	Prob.
Long-run estimates
lnCR	−7.476239	4.459113	−1.67662	0.1158
INFL	0.160173	0.0353	4.537466	0.0005***
FSD	−0.188898	0.0492	−3.83937	0.0018***
PCGDP	−0.439904	0.304086	−1.44664	0.1700
DINV	0.037337	0.079185	0.471516	0.6445
C	38.24189	19.50723	1.960396	0.0702*
Short-run parsimonious estimates
D (lnCR)	11.56938	2.729173	4.239153	0.0008***
D (lnCR(−1))	−9.974395	3.154424	−3.16203	0.0069***
D (INFL)	−1.127046	0.091014	−12.3832	0.0000***
D (PCGDP)	0.247906	0.101771	2.435923	0.0288**
D (PCGDP(−1))	0.207335	0.095706	2.166372	0.0480**
D (DINV)	−0.190798	0.027878	−6.84397	0.0000***
D (DINV(−1))	−0.055995	0.024796	−2.25825	0.0404**
ECM(−1)	−0.947663	0.10793	−8.78033	0.0000***
R ²	0.982453
Adjusted R²	0.975435

Source: Authors calculations.

Note: *** and indicate significance at 1% and 5% level, respectively.

Turning now to the second empirical exercise, Table 7 reports the empirical estimates when real interest rate is the dependent variable which will help us test whether a higher inflation rate results in real interest rate reduction, as postulated by McKinnon (1973) and Shaw (1973). From the results, it is revealed that inflation has a positive effect on the real interest rate in the long run. This is contrary to the short-run results where the effect is negative. Empirical results also reveal that financial system deposit has a negative effect on real interest rate in the long run. As the empirical results report a positive relationship between real interest rate and inflation in the long run, it follows that there is no evidence in support of the McKinnon (1973) and Shaw (1973) hypothesis that higher inflation brings about real interest rate reduction.

The ECM which captures the speed of adjustment of the model to equilibrium in the event of shocks is found to be negative and statistically significant (−0.947663). The value indicates that 95 per cent disequilibrium of the previous year’s shocks is corrected back to the long-run equilibrium in the current year. The adjustment coefficient reported is relatively high, indicating a fast adjustment process. Lastly, the R² and adjusted R² and Durbin–Watson and F-statistics are of the right magnitude. For the R², the coefficient is approximately 98 per cent, indicating that a higher variation in real interest rate is explained by the explanatory variables in the model, while the adjusted R² indicates that the variables in the model explain about 96 per cent of the total variation in real interest rate. By and large, the results reported in this study, especially with regards to the relationship between financial reform and credit growth, are in line with previous studies by Shingjergji and Hyseni (2015) and Aftab et al. (2016), but contrary to the findings by Loayza and Romain (2004), Obamuyi & Olorunfemi (2011), Odhiambo (2010a, 2010b, 2011), Chipote, Mgxekwa, and Godza (2014), Orji, Aguegboh, and Anthony-Orji (2015) and Adeleye et al. (2017). Our finding that increases in gross fixed capital formation leads to credit growth increases is similar to that of Chen and Kang (2018) who show that fixed asset is positively related to credit and more credit has flown to provinces relying more on fixed asset investment in China.

Robustness Check

To ensure that results obtained in this study are not spurious, we perform some diagnostic tests. We employ diagnostic tests such as the Lagrange multiplier (LM) test for autocorrelation, functional form, test for normality and heteroscedasticity, and two stability tests which are cumulative sum (CUSUM) and cumulative sum of squares (CUSUMsq) of the residuals. These are indicated in Figures 2 and 3, respectively. The results of the diagnostic tests, reported in Table 8 below, reveal that our empirical models pass the diagnostic tests for serial correlation, functional form, normality and heteroscedasticity for both cases where credit growth and interest rate are dependent variables. Thus, the models in this study do not have misspecification issues. We also assess the stability of the estimated ARDL model with tests of CUSUM and CUSUMsq. The plots of the CUSUM and CUSUMSQ, also shown below, confirm that the models are stable as the graphs lie within the required critical bounds at the 5 per cent level of significance. These results are included in Appendix A.

Figure 2.

Source: The author.

Figure 3.

Source: The author.

Table 8.

ARDL Model Diagnostic Test

Test Statistic	F-statistics	p-Value
Dependent variable: Credit growth (CR)
Serial Correlation	0.627562	0.5505
Functional Form	0.008488	0.928
Normality	0.700330*	0.704572
Heteroscedasticity	0.771393	0.6831
Dependent variable: Interest rate (INTR)
Serial correlation	2.339399	0.2995
Functional form	0.35072	0.5657
Normality	0.148454*	0.928461
Heteroscedasticity	1.028206	0.4611

Source: Authors calculations.

Note: * Denotes JB statistic and not F-statistic.

Conclusion and Policy Implications

This study has examined the effect of financial and capital investment reforms on credit growth in China. This was with a view to ascertain if these reforms, which have led to less regulated interest rates and an increase in capital investment, have influenced credit growth in China. This also enables us to test if the reform postulates of Mckinnon (1973) and Shaw (1973) hold in China. Most studies have concentrated on the links between financial reforms, capital investment and economic growth in China, with little known about the effects of financial reforms on credit in China.

Using an appropriately specified ARDL model, our empirical analysis reveals that while gross fixed capital formation positively influences credit growth, there is no evidence to support the view that higher real interest rate improves the process of financial intermediation, measured by credit growth, in China. In addition, contrary to the postulates of McKinnon (1973) and Shaw (1973), a rise in inflation does not reduce real interest rate in the long run. Based on the reported results, it is reasonable to posit that a less regulated financial system has not improved financial intermediation and, hence, credit growth in China, partly because the Chinese financial system still faces a significant amount of regulations, in spite of progress made in the past, and is not completely reformed. Perhaps there exists a reform threshold which reforms must attain before they begin to significantly enhance financial intermediation and spur credit growth. One policy implication is thus that, for the Chinese financial system to bring immense benefits to the economy, especially as it regards the process of financial intermediation, there is a need to embark on coordinated reforms and, in particular, increased financial liberalization.

The positive link between fixed capital investment and credit growth is an interesting outcome. Even though financial reforms via real interest rates show no evident impact on financial intermediation, our results provide evidence that the process of financial intermediation has been supported by the growth in China’s fixed capital investment over the past decade. In some measure, this explains why China has been able to sustain the apparent disconnect between credit growth and financial reform over the past decade. Thus, even if we find no evidence that financial reform has driven credit growth in China, our result advances the view that rapid expansion in China’s fixed capital investment is an important channel that supports credit growth in the economy.

Footnotes

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The author received no financial support for the research, authorship and/or publication of this article.

Appendix

References

Acquah

H. D.

(2010). Comparison of Akaike information criterion (AIC) and Bayesian information criterion (BIC) in selection of an asymmetric price relationship. Journal of Development Agricultural Economics, 2, 1–6.

Adam

M. A.

(2011). Financial openness induced growth and poverty reduction. The International Journal of Applied Economics and Finance, 5(1), 75–86.

Adeleye

, Osabuohien

, & Bowale

(2017). The role of institutions in the finance-inequality nexus in sub-Saharan Africa. Journal of Contextual Economics, 137(2017), 173–192.

Aftab

, Jebran

, Ullah

, & Awais

(2016). The impact of interest rate on private sector credit: Evidence from Pakistan. Jinnah Business Review, 4(1), 47–52.

Akingunola

R. O.

, Olusegun

A. A.

, Oluwaseyi

, & Olusoji

S. G.

(2013). The effect of the financial liberalisation on economic growth. International Journal of Academic Research in Economics and Management Sciences, 2(1), 2226–3624.

Bencivenga

V. R.

, & Smith

B. D.

(1991). Financial intermediation and endogenous growth. Review of Economic Studies, 58(2), 195–209.

Chen

, & Kang

J. S.

(2018). Credit booms—Is China different? Washington, DC: International Monetary Fund.

Chandar

, Patro

D. K.

, & Yezegel

(2009). Crises, contagion and cross-listing. Journal of Banking and Finance, 33(9), 1709–1729.

Chipote

, Mgxekwa

, & Godza

(2014). Impact of financial liberalization on economic growth: A case study of South Africa. Mediterranean Journal of Social Sciences, 5(23), 1–8.

10.

Chuah

H. L.

, & Thai

(2004). Financial development and economic growth: Evidence from causality tests for the GCC countries (IMF Working Paper WP/04/XX: 1–47). Washington, D.C., United States.

11.

Chung

K.H.

, Smith

W.T.

, & Wu

T.L.

(2009). Time diversification: Definitions and some closed-form solutions. Journal of Banking & Finance, 33(2009), 1101–1111.

12.

Dong

, Wang

, & Xiangrong

(2014). Interest rate determination in China: Past, present, and future (HKIMR Working Paper, No. 04/2014).

13.

Gang

, Bose

, Chen

, (2015). Prediction of financial distress: An empirical study of listed Chinese companies using data mining. European Journal of Operational Research, 241(1), 236–247. Available at: https://dx-doi-org.web.bisu.edu.cn/10.1016/j.ejor.2014.08.016

14.

Gujarati

D.N.

, & Porter

D.C.

(2009). Basic econometrics (5th edn.). New York, NY: McGraw Hill Inc.

15.

Hye

Q. M.

, & Wizarat

(2013). Impact of financial liberalization on economic growth: A case study of Pakistan. Asian Economic and Financial Review, 3(2), 270–282.

16.

Khalaf

A. H.

(2011). Impact of financial liberalization on financial depth in Iraq. The Review of Finance and Banking, 3(2), 67–78.

17.

King

R. G.

, & Levine

(1993). Finance and growth: Schumpeter might be right. Quarterly Journal of Economics, 108(3), 713–737.

18.

Lardy

(1998). China’s unfinished economic revolution. Washington, DC: Brookings Institution.

19.

(1994). Financial repression and economic reform in China. Westport, CT: Praeger Publisher. Available at: https://ssrn.com/abstract=2696611

20.

Loayza

, & Romain

R. G.

(2004, September). Financial development, financial fragility, and growth. Retrieved from: https://ssrn.com/abstract=859565 or https://dx-doi-org.web.bisu.edu.cn/10.2139/ssrn.859565

21.

McKinnon

R. I.

(1973). Money and capital in economic development. Washington, DC: Brookings Institution.

22.

Merton

R. C.

, & Bodie

(1995). A conceptual framework for analyzing the financial environment. In Crane

D. B.

, Froot

K. A.

, Mason

S. P.

, Perold

, Merton

R. C.

, Bodie

, Sirri

E. R.

& Tufano

(Eds.), The global financial system: A functional perspective (pp. 3–31). Boston, MA: Harvard Business School Press.

23.

Molho

(1986). Interest rate, savings and investment in developing countries: A reexamination of the McKinnon-Shaw hypothesis. IMF Staff Papers, 33(1), 90–111.

24.

Obamuyi

T. M.

, & Olorunfemi

(2011). Financial reforms, interest rate behaviour and economic growth in Nigeria. Journal of Applied Finance & Banking, 1(4), 39–55.

25.

Odhiambo

N. M.

(2010a). Interest rate deregulation, bank development and economic growth in South Africa: An empirical investigation. International Business & Economics Research Journal, 9(11), 1–12.

26.

Odhiambo

N. M.

(2010b). Interest rate reforms and credit allocation in Tanzania: An application of the ARDL bounds testing approach. International Business & Economics Research Journal, 9(5), 23–32.

27.

Odhiambo

N. M.

(2011). The impact of financial liberalisation in developing countries: Experiences from four SADC countries. Addis Ababa: Organisation for Social Science Research in Eastern and Southern, Africa (OSSREA).

28.

Orji

, Aguegboh

, & Anthony-Orji

O.I.

(2015). Real sector output and financial liberalisation in Nigeria. Journal of Infrastructure Development, 7(2), 136–150.

29.

Park

, & Sehrt

(2001). Tests of financial intermediation and banking reform in China. Journal of Comparative Economics, 29(1), 608–644.

30.

Patrick

H. T.

(1966). Financial development and economic growth in underdeveloped countries. Economic Development and Cultural Change, 14(1), 174–189.

31.

Pesaran

M. H.

, Shin

, & Smith

R. P.

(2001). Bounds testing approaches to the analysis of level relationship. Journal of Applied Econometrics, 16(3), 289–326.

32.

Robinson

(1952). The generalization of the general theory, in her the rate of interest and other essays. London: Macmillan.

33.

Shaw

E. S.

(1973). Financial deepening in economic development. New York, NY: Oxford University Press.

34.

Shingjergji

, & Hyseni

(2015). The impact of macroeconomic and banking factors on credit growth in the Albanian banking system. European Journal of Economics and Business Studies, 2(1), 113.

35.

Taylor

(1988). Varieties of stabilization experience: Towards sensible macroeconomics in the third world. Oxford: Clarendon Press.

36.

Wijnbergen

(1982). Stagflation effect on monetary stabilization policies: A quantitative analysis of South Korea. Journal of Development Economics, 10(2), 133–169.

37.

The World Bank (2017). World development indicators (WDI) database. Washington, DC: Author.