Measuring Economy‑wide Impacts of a Financial Shock

Abstract

I. Introduction The onset of the Asian financial shock in the summer of 1997 led to a simultaneous contraction of almost all sectors of production in the crisis-hit country. While the theory of propagation mechanisms from the financial to the real sector in business cycle has been explored (see, for example, Bernanke, Gertler, and Gilchrist 1996), empirical works that aim to identify and measure a shock impact within a general equilibrium framework are rare. This paper attempts to fill the gap. Using the specific case of Indonesia, Thorbecke (1998) and Azis (1998, 2000a) were among the first who attempted to adopt a general equilibrium model for such an analysis. While the former used the Social Accounting Matrix (SAM) multipliers, the latter traced the economy-wide impacts using Structural Path Analysis (SPA) and subsequently employed a price-endogenous CGE model with detailed specifications of the financial sector. (1) The limitation of previous SAM multipliers and SPA studies was the arbitrary manner with which the shock is introduced to the modelled economic system. Within these studies, the standard practice was to induce an artificial fall in the output of sectors that are known ex-post to contract during the crisis. Such an ad hoc method of introducing shock into the system does not capture the actual mechanics of the crisis which was triggered by movements in the financial variables (i.e., foreign capital) rather than in production. The problem is that sectoral output can decline because of numerous types of shocks, of which a financial turmoil is only one of them. Simply reducing the sectoral output artificially thus fails to recognize the origin of the crisis and neglects the linkage between financial sector and the rest of the economy. Another consequence of an ad hoc introduction of the shock is that it prohibits us from gauging the magnitude of the contraction if the crisis had been the only shock that occurred, ceteris paribus. Instead of determining endogenously the impact of the crisis on production, the decline in production is predetermined exogenously based on actual data as if the decline is all due to the financial crisis. This tacit assumption is likely to bias any study of impact estimation. The sources of the bias are the exclusion of other shocks that had nothing to do with the financial crisis (for example, the El Nino-induced drought that caused agriculture-crop failures, and massive haze problems that led to further decline in the agricultural output). Using a more sophisticated financial CGE model would be desirable. However, the needed data are often lacking, and capturing the intricate mechanisms of variables in such a model is far from easy. In this paper, we propose an alternative method to transcend the aforementioned limitation of the standard SAM-based approach without having to construct a CGE model. Specifically, we augment the standard SAM by incorporating a fairly detailed financial sector based on the flow-of-funds data, thus allowing financial variables to be the original source of the shock (a standard SAM condenses financial transactions into a single savings/investments account). While the concept of the flow-of-funds matrix is not new, our contribution is in the explicit use of such matrix in the SAM system. The construction of the flow-of-funds matrix is described in the Appendix. II. Methodology The standard inverse of (I - [A.sub.n]) from the following multiplier [M.sub.a] (1) [Y.sub.n] = [A.sub.n][y.sub.n] + x = [(I - [A.sub.n]).sup.-1] x = [M.sub.a]x, is a useful tool for estimating the impact of an exogenous shock on income of the endogenous accounts. It captures the direct and indirect effects of the shock. However, a multiplier analysis does not reveal the network of paths through which an injection is transmitted (Defourney and Thorbecke 1984). To identify the principal paths of transmission, we employ the SPA method. …