Indonesia Stock Exchange Resilience Toward Crisis; Study Case of Us And China Trade War

Purpose: This paper investigates the resilience’s of Indonesia stock Exchange toward the US and China trade war. Emphasis is put on the influence of US and China trade war since 2016. Design/methodology/approach : We analyze data using five major sectoral indexes in IDX (Consumer Goods Industry Sector, Agriculture Sector, Infrastructure, Utility and Transportation Sector, Miscellaneous Industry Sector, and Mining Sector) using volatility modelling of Autoregressive Distributed Lag (ARDL) & Generalized Method of Moments (GMM) methods to find the correlation and co-integration between selected sector US and China trade war. GMM method showed a significance short-run relationship and long-run relationship. Findings: We find inconclusiveness and no co-integration on correlated markets although the trade wars have impact for each sectoral index by using ARDL. Practical implications: The market disequilibrium takes nearly about one and half week to clear any disequilibrium toward certain shocks or impacts. Originality/value: This paper is original


I. INTRODUCTION
Indonesia and other Asian countries are among the countries with severe impact due to trade war between US and China. The trade war disrupts the resilience of currencies in Asian country and damaging the strength and resilience of stock market in Asian country. Several downturns happening relates with trade war particularly on International trade transaction between Indonesia with China and USA. Resilience toward crisis is the capacity of one country to regenerate itself after a particular shock. The study of economic resilience, however are still limited since the attempt to understand the concepts and its implications are varied. In this research, the interpretation of resilience narrowed down by describing the capacity of Indonesia Stock Exchange to adapt from the trade wars between US and China. (Bachelier, 2011), stated that in determining stock exchange movements are innumerable where past, present or even anticipated events and news often showing no apparent connection with its fluctuations, yet have repercussions on its course.
World Trade Organization, (2019) provides detailed analysis on the latest developments of world trade, according data release in 2019; the world leading merchandise trade in 2018 was China and the world leading commercial services trader was the United States of America (USA). As Indonesian economy require international trade thus making the development of Indonesia Stock Exchange influenced directly and indirectly toward US and China trade war. According to Amiti, M., Redding, S. J., & Weinstein (2019), the impact of trade war towards

II. LITERATURE REVIEW
The correlation between macroeconomic variables and stock market becoming subject of research since (Lafuente, J.A., Ordóñez, 2007); (Goeltom, 2008); (Cooray, A. & Wickremasinghe, 2007), they explained the impact of certain macroeconomic variables on stock prices. The issue of macroeconomic variables and its correlation with company's future earnings has showed a tendency that companies performances are influence by several correlated macroeconomic variables. Stock market return has been relying on the influence of macroeconomic variable since the linearity between economic growth with stock market performances are strong (Liu, Ming-Hua & Shrestha, 2008). Muradoglu, Gulnur & Taskin, (2000) claims that changes in stock prices is related to macroeconomic behavior and pattern that developed over certain times. The Arbitrage Price Theory (APT) championed by Ross (1976) also provides a theoretical framework for the relationship between stock prices and macroeconomic fundamentals by modeling them into linear functions where sensitivity toward changes represented by betaspecific factors, therefore stock returns are generally believed to be determined by some fundamental macroeconomic variables such as interest rates, inflation, exchange rates, and Gross Domestic Product rather than company individual performances (Siu, A., & Wong, 2004); (Chen, M. H., Jang, S. S., & Kim, 2007); (Kirui, E., Wawire, N. H. & Onono, 2014). Another study finds that there are a negative correlation between macroeconomic variables and economy due to lack of international trade ties with another countries (Spyrou, 2001); (Choudhry, 2001); (Hamrita, Mohamed, Essaied & Abdelkader, 2011).
The findings of the study were in contrast to evidence of a positive relationship stated by previous researchers. Kumara (2010) attempts to identify the impact of short term interest rate with period of time measured by 91 days, 182 days and 364 days treasury bill rates on stock prices in Srilanka. The result found that there is weak relationship between short term interest rate and stock price of Sri Lanka.
The other macroeconomics variables that influence to the stock price is oil price. Academic and practitioners explore the relationship between oil price shock and the macro economic variables after oil price shock at the end of 1980. Several papers have been examined the influence of oil price and stock price ranging from major European, Asian, and Latin American emerging markets. Their results indicate a significant short-run linkage between oil price changes and emerging stock markets (Papapetrou, 2001), (Basher, S.A. & Sadorsky, 2006); (Ackert, L. F. Church, B. K. & Deaves, 2003), (Stracca, 2004

Volatility Modeling
Modeling and forecasting stock return volatility is central to modern finance because risk volatility increased due to market uncertainty and the attempt from market participant to manage asset pricing, asset allocation and risk management. Two approaches generally used are the GARCH and stochastic volatility (SV) models. In their standard forms, the ensuing volatility processes are stationary and weakly dependent with autocorrelations that decrease exponentially.
Wild bootstrapped automatic variance ratio (WBAVR) test Let Yt be an asset return at time t, where t = 1,2,...,T. Choi (1999) AVR test statistic takes the following form:

ARDL Bound Testing Approach
To examine the long run relationship among the five sectoral indexes, this research employs the ARDL bound testing approach to co-integration which involves estimating the conditional error correction version of the ARDL model (Pesaran, M. H., Shin, Y. & Smith, 1996). The choice of ARDL approach in this research is based on the consideration of efficient and unbiased co-integration analysis; it can be applied to a small sample size study and therefore conducting bound testing will be appropriate for the present study, secondly it estimates the short and long run components of the model simultaneously, removing problems associated with omitted variables and autocorrelation. Finally, it can distinguish between dependent and independent variable.
The null of no co integration in the long run relationship is defined by H0:λ1 =λ2 =λ3 = λ4 = λ5 = 0 is tested against the alternative of H1:λ1 ≠ λ2 ≠ λ3 ≠ λ4 ≠ λ5 ≠ 0, by the means of familiar F-test. However, the asymptotic distribution of this F-statistic is non-standard irrespective of whether the variables are I (0) or I (1). Pesaran, M. H., Shin, Y. & Smith (1996)have tabulated two sets of appropriate critical values. One set assumes all variables are I(1) and another assumes that they are all I(0). This provides a bound covering all possible classifications of the variables into I(1) and (I)0 or even fractionally integrated.

Generalized Method of Moments
This research investigates the short and long run relationship among Indonesia Stock Market sectoral index of Consumer Goods Industry Sector (log_JKCONS), Agriculture Sector (log_JKAGRI), Infrastructure, Utility and Transportation Sector (log_JKINFA), Miscellaneous Industry Sector (log_JKMISC) and Mining Sector (log_JKMING). The research estimates them by GMM estimation, where the error correction terms are incorporated in the models. The GMM can be used to estimate the model parameters and test the set of moment conditions that arise during period of research. Study on five variables Johansen-Juselius co-integration test, the VECM representation can then be reformulated in a simple matrix form as follows:   Akaike (1974) information criterion (AIC) is used to determine the lag length incorporation in the entire tests of this research. It is important to note that for GMM estimator to be identified; there must be at least as many instrumental variables Z as there are parameters Ɵ. Following (Lee, Byung-Joo & Lee, 1997), this study used lags of explanatory variables as the instrumental variables. These variables were opted for use because of the difficulty in finding other instrument variables, as the research utilizes daily data and for an extended period. As seen from table 2, each variable consists of 2556 data where mean value represents the average value of the variables within the research period. The standard deviation represents the average value by which individual data differs from the mean. The higher standard deviation, the more volatile the selected variables which conclude to higher risk within the variables. In the concept of investment, low standard deviation is more preferable.

A. Preliminary Evidence: AR (p) Model
An AR (p) model was fitted to the returns of five selected industries to ensure the pre-whiten residual before testing the evidence of non-linearity. Stock market returns modeled as autoregressive time series with random disturbances having conditional heteroscedastic variances. Fitting an AR (p) model to the series is necessary to find whether each selected variables respond with volatilities. Five selected industries following the low order autoregressive. Having fitted an AR (p) model, it is now necessary to examine an adequate and useful functional form for the data generating process. In the case of stock price during trade war of US and China, the dynamics volatility of the stock price shown in table 3 where The existence of linearity in six selected stock market of this research shows that stock market is open for gaining abnormal return. Moreover, the existence of linearity in stock market also showed the capability of return prediction that lead to inefficiency. The case of US and China trade war considered as event studies and lead to the opportunities of several investor in gaining abnormal return. From the result in table 5, when we placed log_JKCONS as the dependant variable, the f-stat is 2.324 which is lower than the lower bound of I(0) series, which imply that there is no co-integration among the variables. The same result of f-stat lower than I(0) lower bound also were found when log_JKAGRI and log_JKINFA are the dependant variables, which also imply that at these variables, there were no co integration or no long run relationship. At the other hand, the result for log_JKMING and log_JKMISC are considered to be inconclusive due to the f-stat in between the lower bound I(0) and the upper bound I(1). Notes: ***, ** and * represent significance at the 1%, 5% and 10% level, respectively. ECT t-¹ is derived by normalizing the co integrating vectors on the dependent variable, producing residual r. by imposing restrictions on the coefficients of each variable and conducting wald test, we obtain F-statistics for each coefficient in all equations. Figures in the (.) and [.] represents t-statistics and probabilities for F-statistics, respectively. The optimal lag-length included in the models is based on the Akaike Information criteria (AIC). DW is durbin watson d test for autocorrelation and J-stats is the Hansen's J-Statistics test for correct specification (over identifying restrictions) of the model. Lag length is set at 3 For selected index, the Vector Error Correction Model (VECM) based on the GMM estimation is appropriate technique because this model distinguishes between short and long run dynamics linkages among selected index. Both short and long run linkages are important to see the impact of US and China trade war to five major indexes in Indonesia Stock Exchange. The existence of the short run multivariate Granger Causalities on selected index indicated by the significance of the F-statistics through joint test of lagged differences, while on the long-run shown by the significance of the t-statistics test ECT. The VECM analysis is conducted on the baseline model containing all five indexes. The result showed that all the Error Correction Terms (ECTs) are significant for the five being reviewed. The ECTs coefficients ranging from 0.59 to 0.98 suggest that on average, the selected index takes nearly about one and half week to clear any disequilibrium. Disequilibrium in indexes also considered as distraction for the stability of the stock price.

V. CONCLUSION
In this paper, we explored the co-integration on short and long run relationship of U.S.A and China trade war of 5 sectoral indexes with highest value transaction between with those countries. The result concludes that there is no co-integration between selected industries either short or long runs. The result finds that in long term, the trade war does not deliver influence towards Indonesia trading system particularly in stock market. However, the real value of trading (number of export and import) will be significantly influenced by the trade war since Indonesia has strong bilateral trading ties with both countries. This result is similar with the result found by (Abiad, A. Baris, K. Bernabe, J. A. Bertulfo, D. J., Romance, S. C. Feliciano, P. N. Maria singham, M. J. & Blackman, 2018) with conclusion that trade conflicts have mildly positive effect towards developing Asian countries as the region benefits from trade redirection in electronics and textiles.
However, due to the fact that both countries between U.S.A and China experienced short term business relationship even though the trade war has taken place. On the contrary, the long term effect will impact the most when both countries starting to enforce regulatory policy against each other. The result showed that all the Error Correction Terms (ECTs) are significant for the five being reviewed. The ECTs coefficients ranging from 0.59 to