Monetary Policy and Financial Conditions in Indonesia	285

the peak of the AFC (Iyke, 2018a). Agung, Juhro, and Harmanta (2016) argue that monetary policy alone is not sufficient to maintain macroeconomic stability and recommend complementary policies in Indonesia. In this regard, it is evident that understanding the evolution of the country’s financial conditions will go a long way in helping policymakers pre-empt future deterioration and enhance stability.

Second, the impact of monetary policy on financial conditions in Indonesia and other Asian economies is poorly understood. Debuque-Gonzales and Gochoco-Bautista (2017) examine this issue but use annual data. Policymakers and investors alike are arguably more interested in the reactions of markets at higher frequencies to policy surprises as evidenced in their decisions. For instance, monetary policy decisions are carried out on a quarterly basis. Similarly, firms announce their financial reports quarterly. Thus, a great deal of information is lost when annual data are used. We circumvent this problem by employing quarterly data. In addition, we deal with the well-known price and exchange rate puzzles when identifying monetary policy shocks by including commodity prices and using an alternative recursive ordering of the variables in the model.4

The main goal of monetary policy is to achieve macroeconomic and price (or monetary) stability. As argued by Juhro and Goeltom (2013), macroeconomic and price stability are tied to financial system stability in Indonesia because they are interlinked. Therefore, since financial conditions generally shape the direction of the economy (i.e. they serve as a leading indicator of business activities), our FCI would be a useful tool to enhance the decisions of participants in the Indonesian economy. We find that our FCI tracks financial conditions quite well. For instance, it captures the peaks of the AFC and the Indonesian banking crisis, the relatively stable period from 2000 until 2008, and the global financial crisis and its aftermath. This is consistent with previous FCIs. A unique feature of our FCI is that it is quarterly and thus offers near real-time development in financial conditions. We also find that monetary policy shapes the FCI. A contractionary monetary policy leads to unfavourable financial conditions within the first two quarters. Financial conditions then improve for nearly three quarters, before declining. This finding is robust to an alternative identification strategy. Our findings highlight the critical role of the monetary authority in shaping financial conditions in Indonesia.

The remainder of the paper is organized as follows. Section II presents the model specification and the data. Section III discusses the results. Section IV concludes the paper.

II.MODEL SPECIFICATION AND DATA A. Model Specification

This section outlines the approach used to construct the FCI. It also presents a

Vector Autoregressive (VAR) model to examine the effect of monetary policy on financial conditions.

4The price puzzle is a phenomenon whereby general prices react to a contractionary monetary policy shock by initially rising before falling (Sims, 1992). Christiano, Eichenbaum, and Evans (1999) recommend the inclusion of commodity prices to address this problem. The exchange rate puzzle arises when the exchange rate declines following a contractionary monetary policy shock (Cushman and Zha, 1997).

Monetary Policy and Financial Conditions in Indonesia	287

The policy shock is identified through the one-step-ahead forecast error, ut. Such a shock is structural and is transmitted to the entire economy. In practice, however, the decomposition of ut and an economically meaningful explanation of the structural shocks have remained a controversial topic. If we normalize ut into vt such that E[vt vt’ ]=In, then there exists a matrix A such that ut=Avt. The jth column of A denotes the instantaneous impact of the jth fundamental innovation on all the variables. This fundamental innovation has one standard error in size (Uhlig, 2005; Iyke, 2018b). Therefore, A is restricted by the variance–covariance matrix as follows:

(5)

Equation (5) indicates n(n-1)/2 degrees of freedom remaining in the model, which is not sufficient when identifying shocks to ut. There are several approaches to address this problem.6 Consistent with Sims (1986), we do so by restricting A to be a Cholesky factor of Σ. In other words, we use a recursive ordering of Yt when identifying shocks to ut.

B. Data

Our sample covers the period 1994: Q1 to 2018: Q4. To construct the FCI, we use various variables indicating specific aspects of the financial conditions in Indonesia. We use Bank Indonesia’s rate (IRATE)7 for the interest rate channel, the nominal effective exchange rate (NER) for the exchange rate channel, banking system claims on private enterprise (CREDIT) for the credit channel, the Jakarta Composite Index (JCI) and the MSCI Share Price Index (MSCI) for the equity channel, and the business confidence index (BCI) and the consumer confidence index (CCI) for the expectation or perception channel. In the VAR model, we use the manufacturing production index (MP), the growth in CPI, the FCI, the commodity price index (COM), NER, the short-term interest rate or monetary policy rate (STR), and the monetary base or money supply (M2). The movements of these variables are shown in Figure A1 in the Appendix and the summary statistics and further details on the variables are presented in Tables A1 and A2, respectively.

6See, for example, Bernanke (1986), Blanchard and Watson (1986), Blanchard and Quah (1989), Uhlig

(2005), and Rubio-Ramírez, Waggoner, and Zha (2010). Each approach has its advantages and disadvantages.

7Note that, since 2005 (under the inflation targeting framework), Bank Indonesia has used different policy rates. From 2005 until mid-2016, the bank used the Bank Indonesia Certificate (Sertifikat Bank Indonesia). Then, since mid-2016, the bank has used a seven-day reverse repo rate. These rates are slightly different (i.e. the former is around 150 basis points higher than the latter). This does not imply that Bank Indonesia has pursued an expansionary monetary policy, since the two rates have the same term structure. There has been no change in policy stance.

288Bulletin of Monetary Economics and Banking, Volume 21, Number 3, January 2019

III. RESULTS

A. Measuring Financial Conditions

We begin our analysis by testing for unit roots in the indicators of financial conditions. These results are shown in Table 1. There is no strong evidence to reject the unit root null hypothesis. Therefore, we proceed to constructing the FCI by modelling the indicators in their first differences as linear functions of an unobserved factor. The unobserved factor is assumed to follow a second-order autoregressive process.

Table 1.

Tests for Unit Roots in FCI Constituents

This table reports unit root test results based on the Augmented Dickey-Fuller (ADF) and the Perron and Vogelsang (PV, 1992) breakpoint tests. The null hypothesis is that there is a unit root. The breakpoint type is an innovation outlier. The break point is selected by minimizing the Dickey-Fuller statistic. A maximum of 12 lags is included in these models. Finally, ** and *** denote, respectively, 5% and 1% significance levels.

		ADF test		PV test
	Variable	Zt-statistic (Lag)		Innovation outlier
		Constant	Constant and	t-statistic (Lag)	Break date
			Trend
	IRATE	-1.706(1)	-2.324(1)	-3.614(3)	2008M10
	lnBCI	-2.656(0)*	-4.316(0)***	-4.305(0)*	2009M01
	CCI	-3.876(4)***	-4.469(4)***	-7.621(0)***	2002M02
	lnCREDIT	-1.711(2)	-0.997(2)	-4.140(0)	2000M08
	lnNER	-2.112(2)	-2.510(1)	-5.546(2)***	1997M07
	lnJCI	-0.398(1)	-2.660(1)	-3.030(2)	2003M03
	lnMSCI	-2.210(0)	-1.734(0)	-3.938(0)	1998M07

Table 2 shows the maximum likelihood estimates of the dynamic factor model. Because two of the constituents of the FCI, the business confidence index (BCI) and the consumer confidence index (CCI) have a short time span (i.e. they start in 2000:Q1, whereas the others start in 1994:Q1), we estimate the dynamic factor model with and without these variables. The seven variables used for the dynamic factor model are IRATE, NER, CREDIT, JCI, MSCI, BCI, and CCI. Model

(1)contains all seven variables, whereas model (2) contains all seven except for BCI and CCI. Both models generally indicate some degree of persistence in the unobserved factor, since immediate past values of the factor are significant in the model. The unobserved factor appears to be a significant predictor of all indicators except CREDIT in model (1). The factors have less predictive power over NER, CREDIT, and MSCI in model (2). The estimated signs of the coefficients are generally consistent with conventional wisdom; that is, we could infer that high interest rates tend to signal bad financial conditions, an appreciating rupiah exchange rate signals good financial conditions, high equity returns signal good financial conditions, and good business conditions (perceptions and expectations) translate to good financial conditions.

290Bulletin of Monetary Economics and Banking, Volume 21, Number 3, January 2019

Figure 1. FCI Movement

This graph shows the movements of the FCI (with and without BCI and CCI) and interest rates (1994: Q1 to 2018: Q4).

30
20
10
0
-10		Interest rate growth		FCI without BCI and CCI
		FCI with BCI and CCI
-20
1995q1	2000q1	2005q1	2010q1	2015q1	2020q1

Figure 2. FCI for Indonesia

This graph shows the movements of the FCI for Indonesia (1994: Q1 to 2018: Q4).

10
5
0
-5
	FCI with BCI and CCI		FCI without BCI and CCI
-10
1995q1	2000q1	2005q1	2010q1	2015q1	2020q1

B. Impact of Monetary Policy on Financial Conditions

Financial conditions are not independent of monetary policies. The actions of monetary authorities tend to shape financial conditions. For instance, a tight monetary policy leads to credit shrinkage in the economy. This, in turn, leads to firms cutting down production, layoffs, declines in demand for goods and services, and reductions in business confidence. Similarly, an expansionary monetary policy leads to expansions in credit, production, employment, the demand for goods and services, and inflationary pressures, among others. Good financial conditions, if

292Bulletin of Monetary Economics and Banking, Volume 21, Number 3, January 2019

note that such a relationship tends to reverse the impact of expansionary monetary policy. We document that expansionary monetary policy is linked with favourable financial conditions for the first few quarters. In the medium term, our findings appear to corroborate theirs, in that financial conditions appear to decline, perhaps due to the reduction in risk-taking activities and credit facilities.

Figure 3. Response of FCI to Monetary Policy Shocks

This figure shows the response of financial conditions to a contractionary monetary policy shock of one standard deviation in size, which is identified as the innovation in the short-term interest rate, ordered last in Cholesky decomposition. The FCI is ordered first, followed by lnCPI. The three lines denote the 16% quantile, the median and the 84% quantile of the posterior distribution.

Response to Cholesky One S.D. Innovations ± 2 S.E.

1,2
0,8
0,4
0,0
-0,4
-0,8
1	2	3	4	5	6	7	8	9	10

Figure 4. Response of FCI to Monetary Policy Shocks – Alternative Ordering

This figure shows the response of financial conditions to a contractionary monetary policy shock of one standard deviation in size, which is identified as the innovation in the short-term interest rate, ordered last in the Cholesky decomposition. The FCI is placed first, followed by lnMP. The three lines denote the 16% quantile, the median and the 84% quantile of the posterior distribution.

Response to Cholesky One S.D. Innovations ± 2 S.E.

,8
,6
,4
,2
,0
-,2
-,4
1	2	3	4	5	6	7	8	9	10

Monetary Policy and Financial Conditions in Indonesia	293

IV. CONCLUSION

We create a new FCI and analyse the effect of monetary policy on financial conditions in Indonesia. There are, so far, only limited FCI studies on Asian economies. These studies are based on a panel of Asian economies; however, these countries are interlinked through trade and, therefore, analysis of the unique attributes of their FCIs becomes highly tasking within a single framework. We address this issue by solely focusing on Indonesia.

Indonesia has undergone substantial changes in terms of financial conditions, making it appealing for this study. The country is among the three that were most affected by the AFC. It has also, in recent times, experienced the sharpest depreciation in its currency since the peak of the AFC. Good FCIs would enhance authorities’ abilities to pre-empt future deterioration in financial conditions. In addition, there is little understanding of the impact of monetary policy on financial conditions in Indonesia and other Asian economies. Previous attempts have used annual data, which might not be appealing, because policymakers and investors are arguably more interested in the reactions of markets to policy surprises at higher frequencies, as evidenced in their decisions. We address this point by employing quarterly data.

We find that our FCI tracks financial conditions quite well. For instance, it captures the peaks of the AFC and the Indonesian banking crisis, the relatively stable period from 2000 until 2008, and the global financial crisis and its aftermath. This is consistent with previous FCIs. A unique feature of our FCI is that it is quarterly and thus offers near real-time development in financial conditions. We also find that monetary policy shapes the FCI. A contractionary monetary policy leads to unfavourable financial conditions between the first and second quarters. Financial conditions then improve for nearly three quarters, before declining. This finding is robust to an alternative identification strategy. Our findings highlight the critical role of the monetary authority in shaping financial conditions in Indonesia. In this case, a significant countercyclical monetary policy impact on financial conditions opens up room to augment the standard monetary policy rule by incorporating an unexpected development (deviation) of financial conditions.

REFERENCES

Agung, J., Juhro, S. M., & Harmanta, T. (2016) Managing Monetary and Financial Stability in a Dynamic Global Environment: Bank Indonesia’s Policy Perspectives (October 2016). BIS Paper No. 88f. Available at SSRN: https://ssrn. com/abstract=2861312.

Bernanke, B. S. (1986). Alternative Explanations of the Money–Income Correlation. In: Brunner, K., & Meltzer, A. (Eds.), Real Business Cycles, Real Exchange Rates, and Actual Policies. Carnegie-Rochester Conference Series on Public Policy, vol. 25. North-Holland, Amsterdam.

Blanchard, O., & Watson, M. (1986). Are All Business Cycles Alike? In: Gordon, R.J. (Ed.), The American Business Cycle. University of Chicago Press, Chicago, 123–160.

Blanchard, O. J., & Quah, D. (1989). The Dynamic Effects of Aggregate Demand and Supply Shocks. American Economic Review, 79, 655–673.

294Bulletin of Monetary Economics and Banking, Volume 21, Number 3, January 2019

Brave, S., & Butters, R. (2011). Monitoring Financial Stability: A Financial Conditions Index Approach. Economic Perspectives, 35, 22–43.

Canova, F. (2007). Methods for Applied Macroeconomic Research. Princeton, NJ: Princeton University Press.

Christiano, L., Eichenbaum, M., & Evans, C. (1999). Monetary Policy Shocks: What Have I Learned and to What End? In: Woodford, M., Taylor, J.B. (Eds.), Handbook of Macroeconomics. North-Holland, Amsterdam, 65–148.

Cushman, D. O., & Zha, T. (1997). Identifying Monetary Policy in a Small Open Economy Under Flexible Exchange Rates. Journal of Monetary Economics, 39, 433-448.

Debuque-Gonzales, M., & Gochoco-Bautista, M. S. (2017). Financial Conditions Indexes and Monetary Policy in Asia. Asian Economic Papers, 16, 83-117.

Dudley, W., & Hatzius, J. (2000). The Goldman Sachs Financial Conditions Index: The Right Tool for a New Monetary Policy Regime. Global Economics Paper No. 44.

Freedman, C. (1994). The Use of Indicators and of the Monetary Conditions Index in Canada. Frameworks for monetary stability: policy issues and country experiences, 458-476.

Gauthier, C., Graham, C., & Liu, Y. (2004). Financial Conditions Indexes for Canada. Bank of Canada Working Paper No. 2004–22.

Geweke, J. (1977). The Dynamic Factor Analysis of Economic Time Series Models. In Latent Variables in Socioeconomic Models, ed. D. J. Aigner & A. S. Goldberger, 365–383. Amsterdam: North-Holland.

Goldstein, M. (1998). The Asian Financial Crisis: Causes, Cures, and Systemic Implications. Peterson Institute, 55.

Goodhart, C., & Hofmann, B. (2001). Asset Prices, Financial Conditions, and The Transmission of Monetary Policy. In conference on Asset Prices, Exchange Rates, and Monetary Policy (March, 2001), Stanford University (2-3).

Guichard, S., & Turner, D. (2008). Quantifying the Effect of Financial Conditions on US Activity. OECD Economics Department Working Paper No. 635.

Guichard, S., Haugh, D., & Turner, D. (2009). Quantifying the Effect of Financial Conditions in the Euro Area, Japan, United Kingdom, and United States. OECD Economics Department Working Paper No. 677.

Iyke, B.N. (2018a) A Test of the Foreign Exchange Market Efficiency in Indonesia. Bulletin of Monetary Economics and Banking (forthcoming).

Iyke, B. N. (2018b). Assessing the Effects of Housing Market Shocks on Output: The Case of South Africa. Studies in Economics and Finance, 35, 287-306.

Juhro, S. M., & Goeltom, M. (2013) The Monetary Policy Regime in Indonesia (November 1, 2013). Macro-Financial Linkages in Pacific Region, Akira Kohsaka (Ed.), Routledge, February 2015. Available at SSRN: https://ssrn.com/ abstract=2875631.

Koop, G., & Korobilis, D. (2014). A New Index of Financial Conditions. European Economic Review, 71, 101-116.

Mayes, D., & Virén, M. (2001). Financial Conditions Indexes. Bank of Finland Discussion Paper No. 2001–17.

Monetary Policy and Financial Conditions in Indonesia	295

Rubio-Ramírez,J.,Waggoner,D.,&Zha,T.(2010).StructuralVectorAutoregressions: Theory of Identification and Algorithms for Inference. Review of Economic Studies, 77, 665-696.

Satria, D., & Juhro, S. M. (2011). Risk Behavior in the Transmission Mechanism of Monetary Policy in Indonesia. Bulletin of Monetary Economics and Banking, 13, 243-270.

Sargent, T. J., & Sims,C. A. (1977). Business Cycle Modeling without Pretending to Have too much a Priori Economic Theory. In New Methods in Business Cycle Research: Proceedings from a Conference, ed. C. A. Sims, 45–109. Minneapolis: Federal Reserve Bank of Minneapolis.

Shinkai, J.-I., & Kohsaka, A. (2010). Financial Linkages and Business Cycles of Japan: An Analysis Using Financial Conditions Index. OSIPP Discussion Paper No. 2010-E-008. Osaka School of International Public Policy.

Sims, C. A. (1986). Are Forecasting Models Usable for Policy Analysis. Minneapolis Federal Reserve Bank Quarterly Review Winter, 2–16.

Sims, C. A. (1992). Interpreting the Macroeconomic Time Series Facts. European Economic Review, 36, 975–1011.

Stock, J. H., & Watson, M. W. (1991). A Probability Model of the Coincident Economic Indicators. In Leading Economic Indicators: New Approaches and Forecasting Records, ed. K. Lahiri and G. H. Moore, 63–89. Cambridge: Cambridge University Press.

Swiston, A. (2008). A U.S. Financial Conditions Index: Putting Credit where Credit is Due. IMF Working Paper No. 08/161.

Osorio, M. C., Unsal, D. F., & Pongsaparn, M. R. (2011). A Quantitative Assessment of Financial Conditions in Asia (No. 11-170). International Monetary Fund.

Perron, P., & Vogelsang, T. J. (1992). Nonstationarity and Level Shifts with an Application to Purchasing Power Parity. Journal of Business & Economic Statistics, 10, 301-320.

Uhlig, H. (2005). What are the Effects of Monetary Policy on Output? Results from an Agnostic Identification Procedure. Journal of Monetary Economics, 52, 381- 419.

Yamazawa, I. (1998). The Asian Economic Crisis and Japan. The Developing Economies, 36, 332-351.

296Bulletin of Monetary Economics and Banking, Volume 21, Number 3, January 2019

APPENDIX

Figure A1. Variables Used for Constructing FCI and the VAR Model

This figure shows the behaviour of the variables used in constructing the FCI and the VAR model. The first seven graphs are the financial condition indicators used in the FCI model. The last seven (including lnNER) graphs are those variables used in the VAR model to examine the impact of monetary policy shocks on financial conditions. The maximum sample period employed is from 1994: Q1 to 2018: Q4.

80		Interest Rate
60
40
20
0
1995	2000	2005	2010	2015

Log of Business Confidence Index

5,0

4,8

4,6

4,4

4,2

4,0

1995

2000

2005

2010

2015

298Bulletin of Monetary Economics and Banking, Volume 21, Number 3, January 2019

Figure A1. Variables Used for Constructing FCI and the VAR Model (Continued)

Log of Credit to Private Sector

12
10
8
6
4
1995	2000	2005	2010	2015

Log of MSCI Index

9
8
7
6
5
1995	2000	2005	2010	2015
12			FCI
8
4
0
-4
-8
1995	2000	2005	2010	2015

Monetary Policy and Financial Conditions in Indonesia	299

Figure A1. Variables Used for Constructing FCI and the VAR Model (Continued)

4,8		Log of Manufacturing Production Index
4,6
4,4
4,2
4,0
3,8
1995	2000	2005	2010	2015
5,0		Log of CPI
4,5
4,0
3,5
3,0
2,5

1995	2000	2005	2010	2015
		Log of Commodity Price Index

9,0
8,5
8,0
7,5
7,0
6,5
1995	2000	2005	2010	2015

300Bulletin of Monetary Economics and Banking, Volume 21, Number 3, January 2019

Figure A1. Variables Used for Constructing FCI and the VAR Model (Continued)

Log of Nominal Effective Exchange Rate

6,5
6,0
5,5
5,0
4,5
4,0
1995	2000	2005	2010	2015

60		Monetary Policy Rate
50
40
30
20
10
0
1995	2000	2005	2010	2015

Log of M2

16
15
14
13
12
11
1995	2000	2005	2010	2015