The primary focus of this paper is that money supply is positively related to banking liquidity. Secondly, as an extension of this thesis statement, money supply will be impacted by higher banking liquidity, which will lead to a significant change in share prices. There has been no published consolidated comparative study of the four major Asian economies such as China, India, Korea and Japan, as that we put together in this paper whereby money supply impacts positively on banking liquidity and that banking liquidity affects share prices. An existing study by (Badarudin et al., 2013) shows similar results for G-7 economies. The relationship between money and share prices has been studied by both macro and finance economics. King (1966) showed that macroeconomic impact on share prices accounted for 52 per cent while industry and company factors accounted for just 10 and 38 per cent respectively almost five decades ago. Following this macro approach, our findings could be interpreted as suggesting that the source of the money supply impact is from liquidity, and liquidity impacts share prices. This is traceable to Friedman’s intuition that money supply is the source of liquidity changes, although there have been studies that pin-point some macroeconomic factors, such as inflation, changes in GDP, etc. as relevant for share price formation. In this study, our attempt was to link the primary source of share price changes in these four major Asian (and other) economies to liquidity and money supply. Alternative financial economic theories have suggested a set of variables as relevant key determinants of share price changes. Some of these variables are firm-specific factors (as in Lintner-Sharpe-Mossin’s famous CAPM, Solnik’s exchange rate, Breeden’s consumption; Fama’s price/book ratio, etc.). Only in Chen et al. (1986) did suggestions arise’ that GDP, output, inflation, and term structure are significant macroeconomic factors correlated with non-bank share prices. There is no study except that of Badarudin, op cit., that made a link between Friedman’s liquidity and bank share prices by using the advanced Generalized Method of Moments tests to find evidence in G-7 economies for both money supply effect on liquidity, and liquidity effect on bank share prices. This paper is a report of a further investigation of the money supply effect on liquidity, and liquidity effect on non-bank share prices in Asia. The link between money supply and non-bank share prices has yet to be studied. Friedman (1969) suggested that the macro factor, namely money-supplybased liquidity has a positive influence on asset (share) prices. His proposition of a negative money-supply effect on interest rate has been verified in a number of studies.2 Hamilton (1997) attempted to show a liquidity effect by using daily observations, while others (Pagan and Robertson, 1995; Goodfriend, 199; Lepper and Gordon, 1992; Edmond and Weill, 2005; and Thornton, 2007) have not been successful in finding evidence to support liquidity proposition. Most finance studies built on the firm-specific factors, to share prices; Markowitz (1952), mean variance portfolio; Sharpe (1964), equilibrium asset pricing model; and the widely known capital asset pricing model, CAPM, used finance concepts to connect these factors to share prices. Investors who wish to be successful, need to focus mostly on price-shaping macroeconomic factors, as has been researched by Musílek (1997) and Flannery & Protopapadakis (2002) who consider macroeconomic factors to be the major determinants of shares.
The empirical literature on the liquidity effect dates back at least to Cagan and Gandolfi (1969), Gibson (1970a; b), Leeper and Gordon (1992), Goodfriend (1997), Pagan and Robertson (1995), Christiano, Eichenbaum and Evans (1996), Hamilton (1997), Thornton (2001), Carpenter and Demiralp (2006) and Thornton (2007).
Our motivation in this study was to follow up on such validated statements for the dominant source of share price changes as coming from money supply and liquidity. The rest of the paper is organised as follows. The readers will find in Section 2 a very brief discussion of the money-supply theory, also its variations, and the discussion on (a) liquidity and (b) share price effects. Section 3 is devoted to explaining the hypotheses, methodology and data preparation steps (to correct for stationarity, multicollinearity, serial correlations, and heteroscedasticity), the development of test models for causality tests, the system of equations and the regression models.3 The findings are discussed in Section 4 before the paper ends with relevant comments in Section 5. 2. A Nexus of Interaction of Monetary, Liquidity and Share Price In this section, the nexus between monetary and liquidity impact on share prices from academic literature is explored. We first describe the liquidity effect proposition and its impact on monetary expansion, leading to a credit binge and surging share prices. This is demonstrated in the context of an endogenous money-supply framework. As demonstrated by Friedman (1969), the liquidity effect is the first of three effects on interest rates initiated by an unexpected exogenous change in money supply. Income and inflation are the other two effects of an increase in money supply. In the literature, it is generally accepted that money supply changes do lead to negative interest rate changes as some authors conclude. The linkage between money supply and interest rates has been recognized amongpolicy makers on the basis of evidence of its interest rate effect more than the unproven liquidity effect.4 The money stock is itself an asset in the portfolio of wealth-holders. Increases in the stock of money will also cause decreases in the benefits to holders of money from the last dollar of held. Changes in the supply of money are, therefore, a proxy for likely changes in return on money holdings. The risk-free interest rate has been assumed to be a function of the long-term bond rate. Alternately, it is also conceivable that the demand for money is a function of, among other interest rates, the yield on equities. Any increase in the supply of money will tend to cause all interest rates across the
We have tested with VIF for all variables. The results showed that all but 2 variables are multicollinear. Those two are money and liquidity. We felt that this would not seriously bias in the sense that the coefficient is correct but the signs of the coefficients, t and the F-statistics need to be interpreted carefully because the variances and standard errors of the estimates will increase. The inability to find conclusive evidence has led researchers like Pagan and Robertson (1995) to suggest that this could be due to different (a) definitions of money, (b) models, (c) estimation procedures and (d) sample periods. Duca (1995) adds bond funds to M2 and finds the expanded M2 more explainable of the missing M2 from 1990:3 to 1992:4. As an alternative to this approach, Mehra (1997) suggested redefining the opportunity cost of M2 to include the long-term bond rate to capture the increased substitution of mutual funds for bank deposits.
economy in the demand for money to fall. The speed with which the yield on other assets respond depends on the rate at which excess holdings of money balances are reduced. This provides a clue as to how the central bank uses statutory liquidity reserves to influence this to happen. The potential purchasers, such as institutions, dealers, and wealthy individuals, who form the bulk of the floating supply of corporate stocks, are responsive to changes in their money balances (as did the investors in the US market following QE). If so, then the returns on corporate stocks will be affected, and this is a liquidity effect. Thus, stock prices will be responsive to movements in money supply with a negative coefficient through this channel. The monetary theory is enhanced by the introduction of liquidity as it is the missing link between money and aggregate demand. Increases in liquidity can be observed during business upturns, also when money supply is eased (Quantitative Easing (QE) by the Fed in 2012) on strengthening investments, hence expanding the volume of money, thus also enhancing financial activities that led to an unprecedented share price increase in 2012 in response to QE. Research studies by post-Keynesian economists provide new insights on money being endogenously rather than exogenously determined which is an aspect that needs to be tested in this study. In theoretical as well as in empirical finance, the role of liquidity has been highlighted in recent policy debates especially after the credit splurge of the 1994-2004 that led to asset price bubbles before the Global Financial Crisis (Ariff et al., 2012). Despite its prominent role in conventional theories of the monetary policy transmission mechanism, there has been very little research evidence of a statistically significant or economically meaningful liquidity effect confirmed in the studies to-date. The main channel of influence of the money supply on dividends (from corporate earnings) is through the firm’s expected earnings, especially the expectation effect on share prices from the money supply. Although the current prices of common stock are affected by changing future dividends, the main effect of money supply is on the expected growth rate of dividends arising from the liquidity impact on the incomes of corporations. That influence leads to a permanent change in the earnings of firms, which may be thought of as a positive influence from undertaking positive NPV projects by the firms across the economy, given interest rate declines from money-supply increases. This also suggests that a proxy for earnings such as IPI (industrial production index) is a better variable than dividends since earnings are perfectly correlated with IPI and dividends are managerial variables not closely linked to changing earnings. Thus, the money supply and stock prices are positively related through this channel of macroeconomic influence.
A number of researchers have argued that the lack of empirical support is due to the Fed’s attachment to interest rate targeting in one form or another. They argued that innovations to monetary aggregates, M1, reflect shocks to money demand rather than to money supply. Consequently, the inability to find the liquidity effect is due the nability to isolate a statistically significant variable that reflects the exogenous policy actions of the Fed. The cointegration tests conducted for the Asian countries, indicated that IPI and GDP are cointegrated in the long-run and that IPI can be used as a proxy for earnings in our share price equations.
The theoretical framework presented by monetarists for a relationship to exist between money supply and stock prices may be viewed from the Simple Quantity Model (SQT) or the more sophisticated Portfolio Theory (PT). The SQT states that an increase in money supply results in a change in the equilibrium position of money held across the economy with respect to other non-money assets (for example shares) in the portfolio balance of asset holders. This process alters the demand for other assets that compete with money balances so that liquidity must affect assets through the rebalancing of the portfolio whenever money supply is altered by regulators. SQT of money states: M.V. = P. Q where, M is the total amount of money in circulation in an economy during the period, say a year; P the corresponding price level; P.Q is the nominal money value of output; V is the velocity of money in final expenditures; and Q is an index of the real value of final expenditures. An increase in money supply is expected to increase excess supply of money balance, which in turn leads to excess demand for shares. Share prices are expected to rise as a result of the money-supply increases. This channel of interaction had been described quite some time back as a direct channel for the first time in Sprinkel (1964). He explicitly tested a model incorporating SQT as an asset pricing model. As the supply of money expands, the portfolio of desired versus actual cash holding is thrown out of balance. Since the stock of money must be held by some agents, the prices of other assets as well as goods and services for consumption are bid up to a new equilibrium level. SQT and PT are still in vogue although the question of how the money supply influences the asset prices has newer interpretations, as for example, in Badarudin et al. (2011). Therefore, the relationship between the money supply and stock prices is said to be positive through this adjustment mechanism on assets. In summary, the most plausible explanation for a valid relationship between money supply changes and stock price changes conditional on liquidity effect seems to be a combination of the SQT and the PT. Finance literature has focused more narrowly on the pricing of shares. Building on the work of Markowitz’s (1952) mean variance portfolio, the capital asset pricing model, CAPM, a widely accepted asset pricing theory, was next introduced. These and later theories have identified firm-specific variables to generate a body of knowledge that is collectively termed the asset pricing models: most tests of these models confirm, consistent with King (1966), that firm-specific factors do not explain the full amount of share price changes. Almost all these variable except the term structure variable, are narrowly chosen fundamental factors associated with firms; variables such as industrial production such factors for asset returns, the arbitrage pricing theory (APT) h a branch of research incorporating macro factors into the asset pri
(1986) comes to mind as representing this approach. Such a newe index (IPI), income, inflation, liquidity from money supply, etc. This approach as “Macroeconomic Factor (MFM). brings macroeconomic factors as dominant factors in Model” asset pricing (consistent with King, 1966). Therefore, anyone investing in stocks (shares) should pay attention to these factors. Shortly after acknowledging the importance of such A very relevant forhasthis research is the portfolio mo factors for asset returns, the arbitrage pricing theorypaper (APT) become a focal point of a branch of research incorporating macro factors into the asset pricing where, using the term of the is author, the left sidecould is thehold return to mo assumption thisas model that wealth literature. Chen et al. (1986) comes to in mind representing thisindividuals approach. Such a newer model may be referred to as “Macroeconomic Factor Model” (MFM). anticipated right side common is thefor return to(somewhat stockisasset; Pt is the stock different from money andchang bond A very relevant paper this research the portfolio model ofKeyne’s Cooperpercentage (1970); the assumption in this model is that individuals could hold wealth in two assets determine the quantities of money assets individuals will plus hol forms, moneylevel and common stock (somewhat different from Keyne’s and ; r ∗t isstock the anticipated real pecuniary return of stocks (dividend bond). The marginal returns of stock assets determine the quantities of assets s individuals will hold. MNPS Abalanced portfolio is saidthetomarginal be balanced when marginal when returns to the holding these two assets are prices); t is the marginal pecuniary return to the j-th asset (the ris returns to holding these two assets are equal.
incorporated into its pecuniary Sreturns). MNPStM is implicitly a functio s (1.2) (1.2) MNPSM t − P = MNPSt + rt money except for returns on alternative assets. An underlying assumption i where, using the term of the author, the left side is the return to money asset and the right side is the return to stock asset; is the anticipated percentage M,S cancel each other. the difference betw income t change in general priceeffect levelisontheMNPS anticipated real pecuniary returnThus, of stocks s (dividend plus changeM,S in stock prices); MNPSt is the marginal pecuniary return primarily function ofinto money. In thisreturns). model, money change MNPS t ofisj-th to the j-th asset (the risk assets is aincorporated its pecuniary M MNPSt is implicitly a function of demand for money except for returns on and prices. Theonresult is that mon adjustments through the MNPS t schedules alternative assets. An underlying assumption is that the positive income effect M,S M MNPSt cancel each other. Thus, the difference between MNPSt and MNPStM,S stock returns. is primarily a function of money. In this model, money changes induce portfolio adjustments through the MNPSt schedules and prices. The result is that money supply leads to stock returns.
By re-arranging this equation, it could be shown that the stock return is:
By re-arranging this equation, it could be shown that the stock return is:
S rst = (MNPSM t − P ) − MNPSt
Thus, Cooper’s model is equivalent to the asset pricing model in finance. It would be interestingThus, to examine the link between liquidity from supply Cooper’s model is equivalent to money the asset pricing model in fin affecting the stock prices using the insights provided by Cooper. Friedman’s proposition could be extended as moneythe supply an influence assetmoney supply a interesting to examine link having between liquidity on from prices, namely non-bank share prices in this study. Most of these studies use the Monetary Portfolio (MP) model developed prices using the insights provided by Cooper. Friedman’s proposition cou by Brunner (1961), Friedman and Schwartz (1963) and Cagan (1972) as their starting premise. An investor is assumed as reaching an equilibrium position money he supply influence on asset prices, namely in which, in general, or shehaving holds aannumber of assets including money in anon-bank share p portfolio of assets. A monetary disturbance such as an unexpected increase (or decrease) in the growth rate of the money supply causes disequilibrium in asset Most of studies the money Monetary Portfolio (MP) model deve portfolios. Investors attempt to these rebalance their use desired positions, which are transmitted as monetary changes to the financial markets at large. Over
(1961), Friedman and Schwartz (1963) and Cagan (1972) as their star investor is assumed as reaching an equilibrium position in which, in general time, studying this issue then lapsed until the emergence of the Global Financial Crisis, which has been diagnosed as having been caused by liquidity surges that created imbalance in the financial sector and the real sector: (Ariff et al., 2012). 3. Hypothesis, Methodology and Data 3.1
It is an empirical question whether principal economic indicators such as industrial production, inflation, interest rates, Treasury bill rate, liquidity, and money supply are significant explanatory factors for share returns. Studies of Hardouvellis (1987), Keim (1985), Litzenberger and Ramaswamy (1982) have shown that share prices are affected by macroeconomic factors up to 50 per cent on average, along the same line as King (1966). In addition, if economic variables are significantly and consistently priced in share returns, they should be cointegrated with share returns. If there is no significant cointegrating relationship between macroeconomic variables on one hand and share returns, the conclusion of a linkage is violated. This study will use cointegration and Granger causality tests to investigate the relationship between share returns and underlying macroeconomic variables. An alternative approach, which has certain advantages over both the EngleGranger (1987) OLS and the Johansen (1991) maximum likelihood procedures in cointegration analysis, has been proposed by (Stock and Watson, 1993). Their method improves the Engle-Granger OLS by coping with a small sample and dynamic sources of bias. The Johansen method, being a full information technique, is exposed to the problem that parameter estimates in one equation are affected by any misspecification in other equations. The Stock Watson method is, by contrast, a robust single equation approach which corrects regressor endogeneity with the inclusion of leads and lags of first differences of the regressors, and for serially correlated errors by a GLS procedure. Lag and lead terms included in DOLS regression have the purpose of making its stochastic error term independent of all past innovations in stochastic regressors. In addition it has the same asymptotic optimality properties as the Johansen distribution. The procedure advocated by Stock and Watson (1993) involves thr estimation of long-run equilibria via dynamic OLS (DOLS), which corrects for potential simultaneity bias among regressors. DOLS entails regressing one of the I(1) variables on other I(1) variables, the I(0) variables, and the lags and leads of the first difference of the I(1) variables. The essence of incorporating the first difference variables and the associated lags and leads is to obviate simultaneity bias and small sample bias inherent among regressors. Standard hypothesis testing can be undertaken using the robust standard errors derived via the procedure recommended by Newey and West (1987). It is hypothesized that money supply (MS) is endogenously determined by economic activity as mediated via the deposit-taking institutions (Badarudin et al., 2013). The literature on post-Keynesian theory on endogenous money is extensive.7 Economic activity is proxied by real gross domestic product (Y); liquidity (LQ) is endogenously determined by money supply (MS) and share prices (SP) endogenously by liquidity (LQ). Money supply (MS) is also determined by share prices (SP), inflation (CPI), real GDP (Y) and Treasury bill rate (TBR). Liquidity is determined by real GDP (Y), money supply (MS) and lending rate (LR). A system of equations, comprising 3 simultaneous equations of stock returns (SP) and liquidity (LQ), is developed to be solved endogenously as follows:8 SPit = f [LQit-, MSit+, IPIit+] LQit = f [MSit+, Yit+, LRit-] MSit = f [LQit+, Yit+, TBRit-, SPit+, CPIit+, CPI(1) it+] where SPit is aggregate share price index, LQit is liquidity as proxied by reserve money, MSit is money supply, IPI is industrial production index, Y is real GDP, LR is lending rate, TRB is Treasury bill rate and CPI is inflation. All variables are in logs. These theoretical models are operationalized as follows: ln SPit = a0+ a1 ln LQit + a2 ln MSit + a3 ln IPIit + eit (1.7) ln LQit = b0+ b1 ln MSit + b2 ln Yit + b3 LRit + zit (1.8) ln MSit = c0+ c1 ln Yit + c2 ln LQit + c3 ln SPit + c4 TBRit + c5 lnCPIit + vit (1.9) Two separate sets of hypothesis tests will be conducted to verify the above models. The first set of tests will be conducted on whether money supply is endogenously determined. 3.3
MS causes GDP (suggesting money is exogenous) or there is bidirectional causality (endogenous). Since no tests have been done on this assumption for other countries than Japan (Badaruddin et al., op cit), it is important to test this condition first. It is hypothesised under the alternative hypotheses that there may be unidirectional or bidirectional causality from real GDP to money supply. The second set of hypotheses shown below will be used to test hypotheses
2 to 3: H2 : H3 :
MS causes liquidity: this follows Friedman’s proposition which is still not verified for other countries than Japan. Liquidity causes share prices. This is to test for a bi-directional causality.
It is hypothesised under the alternative hypothesis that money supply causes liquidity and that liquidity causes share price changes.
The data transformation and tests on the data to ensure that the regressions are not spurious, are shown in Appendix 1. Findings on Causality Tests
The results on causality are presented in this sub-section. Since this test has yet to be reported in the literature for China, India and Korea, we report these and the results for Japan as well. The results of the Granger causality tests are shown in Table 1 below after data transformation using unit root testing. This table summarizes the test results of the pair of variables for the Granger causality tests. For MS, GDP (important for Hypothesis 1), they indicate bidirectional causality for India and Japan. For China, MS to GDP is instantaneous with no lead and lag relationship, and for Korea, it shows a unidirectional relationship. Hence, these results indicate that money is endogenous in India and Japan, for China it is instantaneous, while for Korea the impact is from money supply to GDP. These relationships reflect the degree of development of financial institutions in these Asian countries, with India and Japan having a high degree of integration with the real sector. For China, the relationship is segmented while in Korea the relationship is from money supply to GDP. For MS, liquidity (Hypothesis 2), the relationship is unidirectional for all countries (except Japan, which is bidirectional). In India and Korea, Liquidity leads money supply, which reflects the use of statutory reserves by the Central Bank to affect liquidity in the banking system. In Japan, there is a bidirectional impact of money supply and liquidity, while in China, the impact is from money supply to liquidity reflecting a multiplicity of instruments (exchange rates, open market operations and rediscount rate) to control liquidity besides relying on statutory reserves8.
| China | India | Korea | Japan | ||
|---|---|---|---|---|---|
| MS | GDP | Instant | MS <--> GDP | MS --> GDP | MS <--> GDP |
| MS | LQ | MS --> LQ | LQ --> MS | LQ --> MS | MS <--> LQ |
| MS | SP | MS --> SP | Instant | MS --> SP | Instant |
| LQ | GDP | Instant | LQ <--> GDP | LQ --> GDP | LQ <--> GDP |
| SP | GDP | SP --> GDP | Instant | SP --> GDP | GDP --> SP |
| SP | LQ necessarily stable over time. | LQ --> SP variables, LSPRICE, LRLQ AND LRM2. This was conducted in a VAR framework using the Toda-Yamamoto procedure, which is extolled as the best. This result (though not reported in this paper) confirms the earlier univariate and money supply, liquidity and money supply and money supply and GDP. Hence, it is suggested that there is a causal link among the variables confirmed. | LQ --> SP As a robustness test we also did a multivariate test using 3 endogenous Xie (2004) notes that the PBC has been using growth rates of money and bank lending as explicit intermediate targets. The relationship of these aggregates to real activity has not been finding of bidirectional causality between share prices and liquidity, share prices | SP --> LQ | LQ --> SP |
As a robustness test we also did a multivariate test using 3 endogenous variables, LSPRICE, LRLQ AND LRM2. This was conducted in a VAR framework using the Toda-Yamamoto procedure, which is extolled as the best.
Xie (2004) notes that the PBC has been using growth rates of money and bank lending as explicit intermediate targets. The relationship of these aggregates to real activity has not been necessarily stable over time.
This result (though not reported in this paper) confirms the earlier univariate finding of bidirectional causality between share prices and liquidity, share prices and money supply, liquidity and money supply and money supply and GDP. Hence, it is suggested that there is a causal link among the variables confirmed. 3.5
A description of the data sources and variables was given before we reported on the results for hypotheses 2 and 3. The Datastream database was used for sourcing all data while the International Financial Statistics (IFS) database of the International Monetary Fund (IMF) was used to verify the consistency check for the macroeconomic variables to ensure that there were no errors. The data were quarterly series for the period 1968–2012, where the number after the colon refered to the quarter. Data for all four Asian countries were compiled and it is important to note that income was included as an explanatory variable in some equations specified above. Real gross domestic product (GDP) was used as a proxy for income and since only quarterly data were available for income, the highest frequency was quarter. As a proxy for earnings, we combined the literature and decided on the use of the industrial production index (IPI) which is highly correlated with earnings of firms and real income. Hence, we used the log change of IPI as a proxy for earnings in the equation for asset pricing: if IPI goes up, the earnings of the firms go up or vice versa. Liquidity is another difficult variable to specify. There are three alternative proxies: bid-ask spread used in market studies (Amihud and Mendelson, 1986); volume of transactions (Amihud, 2002; Chordia, Subrahmanyam and Anshuman, 2001), all for individual stock liquidity studies. Based on the study by Gordon and Leeper (2002) we used reserve money as the correct variable for financial sector liquidity. This appeared as the right choice using reserve money because, if the banking system has more money in the central bank, liquidity declines, and if it keeps less reserve, liquidity goes up. Therefore, liquidity is inversely related to reserves. As for the data for money supply, M2, values were used.9 The Treasury bill rate and the bank lending rate were the domestic 3-month Treasury-bill rate and the lending rate respectively. In Datastream, the MSCI stock index values were reliably reported and suitable to be used for share returns, SP, and computed as log change. The consumer price index was used as a proxy for inflation, INF. The bank lending rate, LR, deposit rates, TBR, and real gross domestic product, RGDP, were also obtained. All variables were seasonally adjusted where available and transformed to logarithmic form, with the exception of interest rates, which were the local 3-month Treasury bill rate, TBR and the lending rate, LR.
The choice of monetary aggregate has been discussed earlier and its implications on the demand for money have been discussed in Pagan and Robertson (1995) and Duca (1995) on finding the liquidity effect and for the stock market in Parhizgari (2012) on the share price effect.
The asset pricing theory as discussed in Section 2 suggests a relationship between share prices and corporate dividend streams (earnings in this study proxy it) growing at g-rate of growth. In the long run, the g and dividends depend on earnings of the corporations, which directly depends on IPI. Although we were testing the relationship between liquidity and share prices, there was a need to control the effect of earnings changes in the equations. For this, we used the IPI after some initial tests using cointegration. The cointegration tests conducted for the Asian countries, indicated that IPI and GDP were cointegrated in the long-run and that IPI could be used as a proxy for earnings in our share price equations. So, we specified this (earnings and money supply) as control variables in our liquidity equation for share prices. The table below reveals the expected direction of the tested factors in affecting the dependent variables. It also includes our finding, the details of which will be shown in the next section.
| SP | + |
|---|---|
| + | |
| TBRate | CPI + |
| + | |
| CPI(+1) | + |
| + | |
| DUM(GFC) | |
| + | |
| + |
DUM (GFC) and DUM (Regime) are used as control variables in the equations for Korea and Japan because these two countries are more integrated with the world economies. China and India were more isolated in the earlier years and the variables DUM(GFC) and DUM(Regime) are not included.
The results of these carefully run tests are presented in this section. After discussing the descriptive statistics, the data transformation test results are summarised and interpreted. The causality test results are presented next before presenting the results of single equations and then the Dynamic Ordinary Least Squares (DOLS) equations results. 4.1
| DLSPRICE | LCPI | LR | LRGDP | LRIPI | LRLQ | LRM2 | LSPRICE | TBR | |
|---|---|---|---|---|---|---|---|---|---|
| China | |||||||||
| Mean | 0.01 | 4.61 | 5.86 | 4.69 | 2.61 | 8.93 | 10.38 | 5.19 | 2.54 |
| Median | 0.00 | 4.63 | 5.58 | 4.70 | 2.68 | 8.77 | 10.33 | 5.16 | 2.25 |
| Std. | 0.14 | 0.21 | 0.66 | 0.38 | 0.29 | 0.41 | 0.39 | 0.39 | 0.67 |
| Dev. | |||||||||
| Jarq- | 0.11 | 3.89 | 12.59 | 2.34 | 6.91 | 4.14 | 2.14 | 2.00 | 15.72 |
| Bera | |||||||||
| Obs | 48 | 48 | 48 | 48 | 48 | 48 | 48 | 48 | 48 |
| India | |||||||||
| Mean | 0.03 | 4.61 | 4.56 | -4.44 | -0.73 | 5.32 | 4.56 | 6.98 | 11.95 |
| Median | 0.03 | 4.56 | 4.53 | -3.50 | -0.81 | 5.36 | 4.33 | 6.00 | 12.00 |
| Std. | 0.12 | 0.27 | 0.32 | 7.47 | 0.37 | 0.21 | 0.69 | 1.55 | 1.33 |
| Dev. | |||||||||
| Jarq- | 0.19 | 3.12 | 3.78 | 0.42 | 6.30 | 4.88 | 7.12 | 39.39 | 1.93 |
| Bera | |||||||||
| Obs | 62 | 62 | 62 | 62 | 62 | 62 | 62 | 62 | 62 |
| Korea | |||||||||
| Mean | 0.02 | 4.14 | 8.75 | 3.99 | 3.76 | 10.14 | 12.75 | 3.93 | 11.66 |
| Median | 0.02 | 4.24 | 8.50 | 4.12 | 3.87 | 10.27 | 13.06 | 4.20 | 11.83 |
| Std. | 0.12 | 0.49 | 4.23 | 0.62 | 0.84 | 0.65 | 1.20 | 0.91 | 6.30 |
| Dev. | |||||||||
| Jarq- | 7.51 | 9.26 | 25.21 | 10.69 | 8.84 | 9.62 | 13.42 | 12.34 | 19.27 |
| Bera | |||||||||
| Obs | 137 | 137 | 137 | 137 | 137 | 137 | 137 | 137 | 137 |
| Japan | |||||||||
| Mean | 0.01 | 4.18 | 4.03 | 4.96 | -0.04 | -1.09 | 1.13 | 3.83 | 3.34 |
| Median | 0.02 | 4.48 | 4.21 | 5.92 | -0.08 | -1.11 | 1.27 | 4.18 | 4.15 |
| Std. | 0.08 | 0.56 | 0.61 | 2.27 | 0.15 | 0.82 | 0.78 | 1.03 | 2.47 |
| Dev. | |||||||||
| Jarq- | 24.06 | 37.77 | 31.15 | 23.32 | 32.93 | 6.83 | 19.43 | 19.87 | 25.10 |
| Bera | |||||||||
| Obs | 209 | 209 achieved during the bull phase of 1:2009 and a minimum of -39% (Korea) during the bear phase of 4:1997 of the market correction. | 209 Table 3 provides the summary of the descriptive statistics of the variables used in the regression. The variables were transformed into logs (except for LR and TBR) and first differenced (to get the return for SP). The Jarque-Bera (JB) test indicated that most variables for China and India were normal (JB of <5.9 for variables to be considered normal) while in the case of Korea and Japan, the JB tests for normality indicated non-normality (JB >5.9). Most of these variables were skewed (> 0, for normality should be close to 0). A quick read of the values of these variables suggest that these were what one would expect in the panel of Asian countries’ (China, India, Korea and Japan) economies. For example, the inflation rate averaged to 4.6 per cent for China and India, higher than the 4.1 per cent for Korea and Japan. The Treasury bill rate the and lending rate over the test period were (2.5 per cent and 5.9 per cent) for China, (7 per cent and 12 per cent) for India, (11 per cent and 8.8 per cent) for Korea and (3.3 per cent and 5.5 per cent ) for Japan. These first moment values were as reported for the Asian countries. The mean of difference in LSPRICE or the share price returns was 1-3 per cent pa for the Asian countries, with a maximum return of 36% (China) | 209 Note: S.D. is standard deviation. LSPRICE, LRM2, LRIPI, LRGDP, LCPI, TBR, LR and LRLQ are Stock price index, Money supply, Industrial production index, Income, Inflation, Treasury bill rate, Lending rate and Reserve money respectively. All variables are in logarithmic form except for TBR and LR. | 209 | 209 | 209 | 209 | 209 |
LR
LRM2
TBR
China Mean
Median
Std. Dev. JarqBera Obs
India Mean
Median
Std. Dev. JarqBera Obs
Korea Mean
Median
Std. Dev. JarqBera Obs
Japan Mean
Median
Std. Dev. JarqBera Obs
Note: S.D. is standard deviation. LSPRICE, LRM2, LRIPI, LRGDP, LCPI, TBR, LR and LRLQ are Stock price index, Money supply, Industrial production index, Income, Inflation, Treasury bill rate, Lending rate and Reserve money respectively. All variables are in logarithmic form except for TBR and LR.
Table 3 provides the summary of the descriptive statistics of the variables used in the regression. The variables were transformed into logs (except for LR and TBR) and first differenced (to get the return for SP). The Jarque-Bera (JB) test indicated that most variables for China and India were normal (JB of <5.9 for variables to be considered normal) while in the case of Korea and Japan, the JB tests for normality indicated non-normality (JB >5.9). Most of these variables were skewed (> 0, for normality should be close to 0). A quick read of the values of these variables suggest that these were what one would expect in the panel of Asian countries’ (China, India, Korea and Japan) economies. For example, the inflation rate averaged to 4.6 per cent for China and India, higher than the 4.1 per cent for Korea and Japan. The Treasury bill rate the and lending rate over the test period were (2.5 per cent and 5.9 per cent) for China, (7 per cent and 12 per cent) for India, (11 per cent and 8.8 per cent) for Korea and (3.3 per cent and 5.5 per cent ) for Japan. These first moment values were as reported for the Asian countries. The mean of difference in LSPRICE or the share price returns was 1-3 per cent pa for the Asian countries, with a maximum return of 36% (China) achieved during the bull phase of 1:2009 and a minimum of -39% (Korea) during the bear phase of 4:1997 of the market correction. 4.3
We first discuss the results obtained for each single equation and then the cointegration regressions before presenting the Dynamic OLS (DOLS) results. With cointegration, estimator consistency and accurate information on the extent of parameter uncertainty occur without the need for instruments. One advantage of this over instrumental variable estimators is that this avoids potential problems involved with poorly correlated instruments, within estimation procedures such as the generalized method of moments (see e.g. Bound, Jaeger and Baker, 1995). Estimations with the Johansen (1991) method and the dynamic ordinary leastsquares (DOLS) technique of Stock and Watson (1993) generate economically plausible estimates. These coefficients allowed us to obtain estimates of two well-known structural elasticities, the short-run liquidity and interest and the long-run income response to demand for money. The results for each of the three equation for four countries are shown in order. The statistics presented in Table 4 indicate that the dependent variable share price - in the first equation is determined by reserve money (that is liquidity LQ), money supply MS and proxy for earnings IPI and IPI(-1). All the variables are significant in terms of the t-statistics. The liquidity impact on money supply in the third equation is quite significant reflecting that reserve increases will lead to a decline in money supply. In the second equation, liquidity is determined by money supply, income or real GDP and lending rate LR. All the variables except for money supply (LRM2) and lending rate (LR) are significant given their t-statistics at the same acceptance levels of 0.01, 0.05 and 0.10. Money supply in the third equation is determined by income (RGDP), reserve money (LQ), share price (SP), Treasury bill rate (TBR) and inflation
(CPI). The significant relationship between LQ and money (Panel B) in all the four countries (B1, B2, B3 and B4) is as per Friedman’s proposition highlighted in the Granger causality test. Except for inflation, all the variables are significant given t-statistics at the 0.01, 0.05 and 0.10 levels of acceptance. The income elasticity of money (Panel C) is less than one, for all four countries (C1, C2, C3 and C4) given we used M2: in endogenous money supply studies the use of M3 quasi money elicits a larger effect because M2 does not reflect the transaction demand for money. Tables 4A, 4B & 4C are a summary of the results from running single equation regressions on selected individual Asian countries. These countries are China, India, Korea and Japan. The t-statistics are in square brackets, while . ***, **, * denote significance at the 0.01,0.05 and 0.1 levels respectively. The model parameters reported are adjusted R2, std error, mean, std dev. of dependent variable and sum of square residuals. The reported results are mostly in line with the expected signs except for some exceptions. Tables 4A, 4B & 4C: Results of Estimation Using Single Equation
| Share Price | China | India | Korea | Japan |
|---|---|---|---|---|
| Panel A: DV is Share Price | (Coefficients with t-statistics in square brackets) | |||
| C | 1.78 [0.89] | 11.43*** [10.59] | -8.55*** [-5.54] | 0.01 [0.05] |
| LRLQ | 1.68*** [2.76] | 2.41*** [30.76] | 0.77*** [5.86] | -1.04*** [-8.78] |
| LRM2 | -1.46** [-1.91] | -0.96*** [-5.02] | 0.41*** [2.39] | 2.48*** [21.98] |
| LIPI | 0.20 [0.70] | 0.004 [1.19] | 1.43* [1.57] | -1.16 [0.70] |
| DUM (GFC) | -0.33*** [-2.52] | -0.65*** [-6.96] | ||
| DUM (Regime) | Model Parameters | 0.08 [0.66] | -0.26*** [-4.14] | |
| Adjusted R2 | 0.3859 | 0.9776 | 0.8782 | 0.9347 |
| Std Error | 0.3216 | 0.3040 | 0.3252 | 0.2685 |
| Mean | 5.1947 | 2.1087 | 3.9203 | 3.8298 |
| SD of dep var | 0.3926 | 2.0306 | 0.3252 | 1.0349 |
| Sum of Square Res | 4.4483 | 18.8477 | 13.854 | 14.417 |
Korea
Japan
(Coefficients with t-statistics in square brackets) [0.89]
Group 2
DUM (Regime)
Model Parameters Adjusted R2
Std Error
Mean
SD of dep var
Sum of Square Res
(continued)
| Share Price | China | India | Korea | Japan |
|---|---|---|---|---|
| Panel A: DV is Share Price | (Coefficients with t-statistics in square brackets) | |||
| C | 1.78 [0.89] | 11.43*** [10.59] | -8.55*** [-5.54] | 0.01 [0.05] |
| LRLQ | 1.68*** [2.76] | 2.41*** [30.76] | 0.77*** [5.86] | -1.04*** [-8.78] |
| LRM2 | -1.46** [-1.91] | -0.96*** [-5.02] | 0.41*** [2.39] | 2.48*** [21.98] |
| LIPI | 0.20 [0.70] | 0.004 [1.19] | 1.43* [1.57] | -1.16 [0.70] |
| DUM (GFC) | -0.33*** [-2.52] | -0.65*** [-6.96] | ||
| DUM (Regime) | Model Parameters | 0.08 [0.66] | -0.26*** [-4.14] | |
| Adjusted R2 | 0.3859 | 0.9776 | 0.8782 | 0.9347 |
| Std Error | 0.3216 | 0.3040 | 0.3252 | 0.2685 |
| Mean | 5.1947 | 2.1087 | 3.9203 | 3.8298 |
| SD of dep var | 0.3926 | 2.0306 | 0.3252 | 1.0349 |
| Sum of Square Res | 4.4483 | 18.8477 | 13.854 | 14.417 |
[-4.72] DUM (Regime)
Model Parameters Adjusted R2
Std Error
Mean
SD of dep var
Sum of Square Res
| Share Price | China | India | Korea | Japan |
|---|---|---|---|---|
| Panel A: DV is Share Price | (Coefficients with t-statistics in square brackets) | |||
| C | 1.78 [0.89] | 11.43*** [10.59] | -8.55*** [-5.54] | 0.01 [0.05] |
| LRLQ | 1.68*** [2.76] | 2.41*** [30.76] | 0.77*** [5.86] | -1.04*** [-8.78] |
| LRM2 | -1.46** [-1.91] | -0.96*** [-5.02] | 0.41*** [2.39] | 2.48*** [21.98] |
| LIPI | 0.20 [0.70] | 0.004 [1.19] | 1.43* [1.57] | -1.16 [0.70] |
| DUM (GFC) | -0.33*** [-2.52] | -0.65*** [-6.96] | ||
| DUM (Regime) | Model Parameters | 0.08 [0.66] | -0.26*** [-4.14] | |
| Adjusted R2 | 0.3859 | 0.9776 | 0.8782 | 0.9347 |
| Std Error | 0.3216 | 0.3040 | 0.3252 | 0.2685 |
| Mean | 5.1947 | 2.1087 | 3.9203 | 3.8298 |
| SD of dep var | 0.3926 | 2.0306 | 0.3252 | 1.0349 |
| Sum of Square Res | 4.4483 | 18.8477 | 13.854 | 14.417 |
TBRate
(continued)
| Share Price | China | India | Korea | Japan |
|---|---|---|---|---|
| Panel A: DV is Share Price | (Coefficients with t-statistics in square brackets) | |||
| C | 1.78 [0.89] | 11.43*** [10.59] | -8.55*** [-5.54] | 0.01 [0.05] |
| LRLQ | 1.68*** [2.76] | 2.41*** [30.76] | 0.77*** [5.86] | -1.04*** [-8.78] |
| LRM2 | -1.46** [-1.91] | -0.96*** [-5.02] | 0.41*** [2.39] | 2.48*** [21.98] |
| LIPI | 0.20 [0.70] | 0.004 [1.19] | 1.43* [1.57] | -1.16 [0.70] |
| DUM (GFC) | -0.33*** [-2.52] | -0.65*** [-6.96] | ||
| DUM (Regime) | Model Parameters | 0.08 [0.66] | -0.26*** [-4.14] | |
| Adjusted R2 | 0.3859 | 0.9776 | 0.8782 | 0.9347 |
| Std Error | 0.3216 | 0.3040 | 0.3252 | 0.2685 |
| Mean | 5.1947 | 2.1087 | 3.9203 | 3.8298 |
| SD of dep var | 0.3926 | 2.0306 | 0.3252 | 1.0349 |
| Sum of Square Res | 4.4483 | 18.8477 | 13.854 | 14.417 |
DUM (Regime)
Model Parameters Adjusted R2 Std Error Mean
SD of dep var
The Stock-Watson Dynamic OLS estimates of the 3 equations are summarized in Table 5 (China), Table 6 (India), Table 7 (Korea) and Table 8 (Japan). The equations were estimated including up to j=±1 leads and lags, the insignificant lags and leads were dropped. The long-run variance shown replaces the usual estimator for the residual variance of Vit with an estimator of the long-run variance of the residuals. If we had used OLS as an estimator, the value for the long-run variance would be SE of the regression.
The statistics indicate that the dependent variable share price, SP, in the first equation, Panel A, is determined by reserve money LQ, money supply MS and earnings of firms IPI and IPI(-1); the last two being control factors. All the variables except earnings were statistically significant in terms of the t-statistics at the same acceptance levels normally used. This confirmed our findings of a long-run liquidity elasticity impact of 3.1 on share prices. In the second equation for liquidity (Panel B), the money supply effect was evident with an elasticity of 1.75, and was statistically significant, also after controlling for the effects of income, lending rate and stochastic trend. All the variables except income and lending rate were significant in terms of the t-statistics at the same levels normally used.
Single equations were also estimated separately for each of the Asian countries to cross-check on the differences of the OLS and DOLS results. Those results are not reported here. Since DOLS results obviate the simultaneity bias and small sample bias inherent, we report only these results: the OLS results are available for inspection, if needed.
| Panel B: LRLQ |
| Variable |
| LRM2 |
| LRGDP |
| LR |
| C |
| @TREND |
| R-squared |
| Adjusted R-squared |
| S.E. of regression |
| Durbin-Watson stat |
| Panel C: LRM2 |
| Variable |
| LRGDP |
| LRLQ |
| LSPRICE |
| TBR |
| LCPI |
| C |
| @TREND |
| R-squared |
| Adjusted R-squared |
| S.E. of regression |
| Durbin-Watson stat |
| (Liquidity) |
| Coefficient |
| 1.75*** |
| 1.80 |
| 0.07* |
| -16.40*** |
| -0.07*** |
| 0.980754 |
| 0.972935 |
| 0.063988 |
| 0.501774 |
| (Money Supply) |
| Coefficient |
| 0.79*** |
| 0.20*** |
| -0.03* |
| -0.04*** |
| -0.96*** |
| 9.15*** |
| 0.02** |
| 0.999463 |
| 0.998944 |
| 0.011652 |
| 1.673374 |
| t-Statistic |
| 5.91 |
| 1.67 |
| 1.73 |
| -3.48 |
| -2.44 |
| Mean dep var |
| S.D. dep var |
| Sum of sqd residual |
| Long-run variance |
| 8.928131 |
| 0.388953 |
| 0.131025 |
| 0.007587 |
| t-Statistic |
| 2.69 |
| 7.98 |
| -1.83 |
| -9.43 |
| -15.29 |
| 9.71 |
| 2.49 |
| Mean dep var |
| S.D. dep var |
| Sum of sqd residual |
| Long-run variance |
| 10.38248 |
| 0.367329 |
| 0.003259 |
| 6.06E-05 |
In the demand for money equation for China, long-run income elasticity was found to be 0.79, less than 1 as M2 were used with money and interest elasticity of -0.04. Apart from the long-run price elasticity, liquidity elasticity and share price elasticity, which had the opposite sign, all other elasticity had the predicted sign and were highly significant. The coefficient in the linear trend was statistically different from zero at the acceptable levels, indicating that there was a deterministic time trend common to both LRM2 and its determinants. The DOLS for the demand for money is robust to various departures from the standard regression assumptions in terms of residual correlation, hetroscedasticity, misspecification of functional form and non-normality of residuals. Stability tests conducted by plotting CUSM and CUSMSQ (not shown here) suggest that estimated models were stable over the sample period.
the variables except earnings (a control variable) were statistically significant in terms of the t-statistics at the usual acceptance levels of significance. Our finding of a long-run liquidity elasticity impact of 3.9 on share prices was also confirmed for India. In the second equation for liquidity (Panel B), the money supply effect was evident with an elasticity of 3.86, which was also statistically significant; this was after controlling for the effects of income, lending rate and stochastic trend. All the variables except lending rate were significant with large t-statistics significant at the levels normally used in tests.
| Panel B: LRLQ |
| Variable |
| LRM2 |
| LRGDP |
| LR |
| C |
| @TREND |
| R-squared |
| Adjusted R-squared |
| S.E. of regression |
| Durbin-Watson stat |
| (Liquidity) |
| Coefficient |
| 3.86*** |
| 1.99*** |
| 0.02*** |
| -8.73*** |
| -0.02*** |
| 0.993506 |
| 0.991630 |
| 0.032747 |
| 1.214469 |
| t-Statistic |
| 3.64 |
| 7.98 |
| 2.53 |
| -18.06 |
| -3.42 |
| Mean dep var |
| S.D. dep var |
| Sum of sqd residual |
| Long-run variance |
| -0.725948 |
| 0.357943 |
| 0.048257 |
| 0.001431 |
| Panel C: LRM2 |
| Variable |
| LRGDP |
| LRLQ |
| LSPRICE |
| TBR |
| LCPI |
| C |
| @TREND |
| R-squared |
| Adjusted R-squared |
| S.E. of regression |
| Durbin-Watson stat |
| (Money Supply) |
| Coefficient |
| -0.98 |
| 0.55*** |
| 0.02 |
| -0.004 |
| -1.15*** |
| 9.57*** |
| 0.03*** |
| 0.995222 |
| 0.992511 |
| 0.018110 |
| 1.044157 |
| t-Statistic |
| 1.58 |
| 5.04 |
| 0.41 |
| -0.44 |
| -8.06 |
| 10.66 |
| 4.06 |
| Mean dep var |
| S.D. dep var |
| Sum of sqd residual |
| Long-run variance |
| 5.325013 |
| 0.209261 |
| 0.012135 |
| 0.000402 |
In the demand for money equation for India, long-run income elasticity was found to be -0.98, and interest elasticity was -0.004 with correct signs. Besides income elasticity, the long-run inflation elasticity and the coefficient in the linear trend had signs opposite to the theoretical signs. This could be due to an economy with high inflation, a well-documented fact for this economy, and the demand for money was thus seriously affected by inflation. The coefficient in the linear trend was statistically different from zero at conventional levels, indicating that there was a deterministic time trend common to both LRM2 and its determinants.
The statistics indicate that the dependent variable share price, SP, in the first equation, was determined by reserve money LQ, money supply MS, earnings of firms IPI and IPI(-1) and dummy for global financial crisis; the last three being control factors. All variables were statistically significant in terms of the t-statistics as used in acceptance for significance. These findings confirmed our a priori expectations of a long-run liquidity elasticity impact on share prices, with an elasticity of 1.3.
| Panel B: LRLQ |
| Variable |
| LRM2 |
| LRGDP |
| LR |
| DUM |
| DUM1 |
| C |
| @TREND |
| R-squared |
| Adjusted R-squared |
| S.E. of regression |
| Durbin-Watson stat |
| (Liquidity) |
| Coefficient |
| -0.97*** |
| 4.33*** |
| 0.028*** |
| -0.17 |
| 0.24*** |
| 8.29*** |
| -0.02*** |
| 0.982527 |
| 0.976606 |
| 0.120366 |
| 0.767867 |
| Panel C: LRM2 |
| Variable |
| LRGDP |
| LRLQ |
| LSPRICE |
| TBR |
| LCPI |
| DUM |
| DUM1 |
| C |
| @TREND |
| R-squared |
| Adjusted R-squared |
| S.E. of regression |
| Durbin-Watson stat |
| (Money Supply) |
| Coefficient |
| 2.29*** |
| -0.12*** |
| 0.05*** |
| -0.012* |
| 0.63*** |
| 0.28*** |
| 0.03 |
| 4.13*** |
| -0.02*** |
| 0.999549 |
| 0.999197 |
| 0.032645 |
| 0.437193 |
| t-Statistic |
| -3.50 |
| 5.47 |
| 2.23 |
| -1.22 |
| 2.32 |
| 6.67 |
| 3.86 |
| Mean dep var |
| S.D. dep var |
| Sum of sqd residual |
| Long-run variance |
| 9.873795 |
| 0.786957 |
| 1.753034 |
| 0.029771 |
| t-Statistic |
| 14.06 |
| -1.80 |
| 1.99 |
| -1.51 |
| 4.02 |
| 7.58 |
| 1.16 |
| 9.22 |
| -10.53 |
| Mean dep var |
| S.D. dep var |
| Sum of sqd residual |
| Long-run variance |
| 12.77017 |
| 1.152148 |
| 0.077794 |
| 0.001909 |
In the second equation for liquidity, the money supply effect was evident with an elasticity of -0.97 and was statistically significant, after controlling for the effects of income, lending rate, dummies (regime changes and global finance crisis) and stochastic trend. All the variables except for income and lending rate were significant in terms of the t-statistics at the same levels normally used. In the demand for money equation for Korea, long-run income elasticity was found to be 2.29 and interest elasticity was -0.012. Apart from trend coefficient which had the opposite sign, all the other control variables (liquidity, share price, inflation and dummy variable for global finance crisis and regime changes) had the predicted signs and were highly significant. The coefficient in the linear trend was statistically different from zero at conventional levels, indicating that there was a deterministic time trend common to both LRM2 and its determinants.
statistics indicate that the dependent variable share price, SP, in the first equation, was determined by reserve money LQ, money supply MS, earnings of firms IPI and IPI(-1). The control dummy for global financial crisis, along with the other last two were significant; this is the consistent with the known fact that the crisis did affect Japan. All the variables were statistically significant in terms of the t-statistics at the same levels used in this research. The Japan findings helped to confirm our a priori expectations of a long-run liquidity elasticity impact on share prices, although the signs were different with an elasticity of -1.4. In prior studies by others (for example the G-7 study), it has also been shown that the results were similar.
| Panel B: LRLQ |
| Variable |
| LRM2 |
| LRGDP |
| LR |
| DUM |
| DUM1 |
| C |
| @TREND |
| R-squared |
| Adjusted R-squared |
| S.E. of regression |
| Durbin-Watson stat |
| Panel C: LRM2 |
| Variable |
| LRGDP |
| (Liquidity) |
| Coefficient |
| 1.61*** |
| -1.22*** |
| -0.01 |
| -0.14 |
| -0.06 |
| 1.85*** |
| 0.004 |
| 0.981806 |
| 0.975494 |
| 0.120868 |
| 0.422899 |
| (Money Supply) |
| Coefficient |
| 0.90*** |
| t-Statistic |
| 3.12 |
| -2.37 |
| -0.32 |
| -0.81 |
| -0.46 |
| 1.19 |
| 1.19 |
| Mean dep var |
| S.D. dep var |
| Sum of sqd residual |
| Long-run variance |
| -1.092434 |
| 0.772106 |
| 2.147546 |
| 0.041554 |
| t-Statistic |
| 15.41 |
| (continued) |
| LRLQ |
| LSPRICE |
| TBR |
| LCPI |
| DUM |
| DUM1 |
| C |
| @TREND |
| R-squared |
| Adjusted R-squared |
| S.E. of regression |
| Durbin-Watson stat |
| 0.15* |
| 0.05*** |
| -0.08 |
| -0.17*** |
| 0.08*** |
| -0.08 |
| -2.05*** |
| 0.003*** |
| 0.999153 |
| 0.998680 |
| 0.026686 |
| 0.284517 |
| 1.85 |
| 2.16 |
| -0.84 |
| -3.97 |
| -0.98 |
| -1.67 |
| -7.61 |
| 1.73 |
| Mean dep var |
| S.D. dep var |
| Sum of sqd residual |
| Long-run variance |
| 1.134701 |
| 0.734582 |
| 0.090443 |
| 0.001712 |
Note: Superscripts are p-values. ***, **, * denote significance at the 1, 5 and 10 percent levels respectively. C is a constant for each equation. SP is stock price index, LQ is liquidity as proxy by reserve money, MS is M2 as proxy for money supply, IPI is industrial production index as proxy for earnings, Y is real GDP as proxy for income, TBR is treasury bill rate, CPI is inflation and LR is lending rate.
In the second equation for liquidity (Panel B), the money supply effect was found as per our a priori expectation of money supply with an elasticity of 1.61, which was statistically significant, after controlling for the effects of income, lending rate, dummies (regime changes and global financial crisis) and stochastic trend. All the variables except the income and lending rate were significant given the size of the t-statistics and the probability level used in this study. In the money demand equation (Panel C) for Japan, the long-run income elasticity was found to be 0.90, close to most findings of 1 and interest elasticity of -0.08. All the other control variables (liquidity, share price, inflation and dummy variable for global finance crisis and regime changes) had the predicted signs and were highly significant. The coefficient in the linear trend was statistically different from zero at conventional levels, indicating that there was a deterministic time trend common to both LRM2 and its determinants. 6. Conclusion This paper is a report of an investigation on the (a) money supply effects on interest rate and (b) banking liquidity as well as (c) banking liquidity effect on share price using Asian data from China, India, Korea and Japan. The literature on money supply effect has been widely followed in policy circles, yet propositions (b) and (c) have yet to verified. By adopting all the refinements needed to obtain robust parameter estimates and by using the system of equations (DOLS) developed for this study, the results reported in this paper are useful new findings on the money supply and share pricing literature. The data for the Asian countries used cover the period of 1968 to 2012, which were quarterly series on all variables. The variables were transformed to ensure that there was no spurious parameter estimates as an improvement to prior studies. Friedman’s 1969 propositions were used; we added an asset pricing equation to these propositions in order to extend the test for a liquidity– credit surge - effect on share prices. Further, we control led for monetary regime changes and the effect of global crisis by specifying dummy variables to correct structural breaks in the time series. The results (using the Granger and Toda Yamamoto tests) reconfirmed the already known evidence that money supply was endogenous and that there was bidirectional causality from money to interest rate as already confirmed in the studies on the post-Keynesian endogenous money theory. They indicated bidirectional causality for all the variables, MS, GDP (important for Hypothesis 1) and Liquidity (Hypothesis 2). All variables had bidirectional impact on one another except for MS to GDP for China (instantaneous with no lead and lag relationship) and Korea, which showed a unidirectional respectively. These new findings reported here relate to (a) the effect of money supply on banking liquidity that has yet to be confirmed and (b) the banking liquidity effect on share prices. We showed that these effects are significant as they weretested using DOLS to correct for simultaneity and endogeneity, thus, confirming Friedman’s second proposition on money effect on banking liquidity. We extended that proposition via a share-pricing equation to test and confirm the banking liquidity effect on share prices. Our data limitation was simply that this research used quarterly series since GDP data were only available as quarterly series and earnings which were available on an annual basis were proxied by industrial production index. Author information Chung Tin Fah, PhD, is a senior lecturer at Taylor's University, Subang Jaya, Malaysia. He is a committee member of RAM Rating and Bursa Market Participants. This topic was part of his PhD thesis and he has taught and researched on this topic for several years, publishing articles in journals. He lectures and conducts training in economics and finance. The author wishes to record his gratitude to the two anonymous referees and the editors of the Journal for providing review comments on an earlier version of this paper. Chung Tin Fah may be contacted at: E-mail: tinfah@yahoo.com.