A Risk management analysis to public debt sustainability in Indonesia

Purpose: In this paper we analyse the public debt sustainability of Indonesia from a risk management approach as initially proposed by Garcia and Rigobon (2004). This approach incorporates the effects of stochastic shocks to output, exchange rates, primary balance, interest rates and debt levels. It estimates the conditional means and variances from the data, simulates the different paths of the debt and measures risk of debt reaching high levels within a period of time. Originality: This study adopts a new risk management approach for fiscal sustainability in Indonesia. It introduces uncertainty into the analysis of the public debt in Indonesia. The results of this new approach provide useful and important guidance for policy makers. Design/methodology/approach: We estimate a VAR (Vector Auto Regression) to find the correlation pattern of the macroeconomic variables involved and use it to implement simulations. These simulations allow us to compute “risk probabilities”, i.e., probabilities that the simulated debt-to-GDP ratio exceeds a given threshold deemed risk. This approach involves three main steps. First, we estimate a VAR model and examine the impulse-responses obtained. Second, we perform simulation to obtain several debt paths. Finally, we use the simulated debt series to compute the distribution functions of debt ratios for each projection period. Further, we compute the risk probabilities that the simulated debt-to-GDP ratio exceeds a given threshold assumed to be risky. Findings: This study is still ongoing but our first analysis of the data has shown that the debt-to-GDP ratio has been decreasing since 2001, well below the average for emerging and advanced countries. Research limitations/implications: Compiling an adequate data set is a major challenge for this study. Available time series is not long enough to permit the use of more flexible lag structures in the estimated equations, which may potentially make the model estimation biased.