![Sample our Humanities journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5939/TOC-Subject-Sample-2021-Humanities.jpg"})
Abstract
This study tests Korea’s two endogenous money supply channels: the accommodationist and structuralist channels. The main analytic tools are the Fully Modified OLS, Maki’s structural break test, Error Correction Model, and Impulse Response Function (IRF) analysis. The primary purpose of this study is to propose analytical methods to compare the two endogenous money supply channels. The analytic models also consider the impact of fiscal policy and the current account balance on reserves. The test results reveal a unilateral causality from loan to reserve and a bilateral causal relationship between loan/reserve and Lf/deposit ratio. These results suggest that both accommodationist and structuralist channels coexist in Korea. However, the structuralist channel explains a relatively small portion of the endogenous money supply process. The IRF analysis does not provide evidence for the pressure on the interest rate spread exerted by financial institutions’ portfolio adjustment. This result can be attributed mainly to the central bank’s growing accommodative attitude toward the reserve demand.
Notes
1 In this regard, Palley (Citation1994, 74) states as follows: “Both Structuralists and Accommodationists believe that the money supply is endogenously affected by bank lending.”
2 Here, “→” represents the direction of causality.
3 The “money supply-focused period” was set from January 1987 to June 1997, and the “interest rate-focused period” was set from January 2001 to June 2017.
4 Liquidity Coverage Ratio (LCR) = Liquidity assets / Net cash outflows over one month.
5 Net Stable Funding Ratio (NSFR) = Available Stable Funding / Required Stable Funding.
6 If the explanatory variables are strictly exogenous with respect to error terms, both ω12 and λ12 are 0. In this case, the β estimated by OLS can be both efficient and unbiased.
7 For the entire Error Correction Model to converge toward a cointegration relationship, the sign of an adjustment coefficient (λ) must be the opposite of the sign taken by the dependent variable located at the cointegration equation.
8 Most empirical analyses below were conducted through Eviews 12. Meanwhile, Maki (Citation2012) cointegration test was performed through Gauss 21. The data and work files supporting this study’s findings are available on request from the author.
9 A typical General-to-Specific procedure selects an abbreviated model that minimizes information criteria. Meanwhile, Brüggemann and Lütkepohl (Citation2001) showed that excluding variables via the AIC criterion is equivalent to eliminating variables insignificant at the 15–20% significance level based on the t statistic. Therefore, explanatory variables with t statistic less than 1.3 ∼ 1.4 were sequentially eliminated when implementing the General-to-Specific procedure in this study.
10 Although most of the data set’s starting point is December 2001, due to some outliers of loan data influenced by the credit card bubble, which had existed in late 2001 and early 2002, the starting point of the data was set as mentioned above.
11 To check the robustness of the above unit root test result, another unit root test statistic called “Intercept break min-t” suggested by Perron (Citation1997), was also used. Even in this case, loan_lresb followed the I(1) process.
12 As lloan was found to be weakly exogenous via the VECM test, the cointegration equation having lres as a dependent variable was estimated. A weakly exogenous variable is a variable that does not contribute to the cointegration relationship. See Appendix 2.
13 Considering that both netg and netx are I(0) series, level variables were used.
14 An outlier of sizable fiscal deficit occurred in this period due to the large-scale conversion of public funds for financial restructuring (after the IMF Crisis) into government bonds. See Lee and Kang (Citation2007, 27).
15 It was found that when switching Eq. (1)’s dependent variable to d(lloan), the t-statistic of the adjustment coefficient (which took a value of –1.1727) was insignificant and showed a theoretically meaningless sign.
16 According to Lee and Yoon (Citation2001), since the IMF crisis, the monetary authority has purchased foreign exchange to stabilize the exchange rate in case of an accumulation of balance of payments (BOP) surplus and then neutralized the resulting fluctuation in monetary base through “monetary stabilization bond” trade.
17 The more accommodative the central bank becomes, the less need for financial institutions to rely on liability management arises.
18 However, in the cases of Eqs. (5)–(6), coefficients of fiscal balance and current account balance display opposite signs compared to Eqs. (2)–(3) because the reserve is the denominator of the dependent variable.
19 In March 2003, financial market instability escalated, with concerns over credit card lending distress. In addition, in October 2005, the BOK ended its low-interest-rate stance and began to raise its base rate to terminate the interest rate reversal with the United States. The base rate started at 3.25% and eventually climbed to 5.25% in August 2008, just before the global financial crisis.
Additional information
Notes on contributors
Wonik Park
Wonik Park is an associate research fellow at the Innovation Procurement Research Center in the Korea Institute of Procurement, Seoul, South Korea and received Ph.D. in Economics from Korea University. This study draws on the second chapter of the author’s Ph.D. dissertation. [email protected]