Measuring economic uncertainty and crisis risk in the asian region’

The workshop, on 6 May 2019, focused on key results from a joint research project between ADB and a team of researchers from Macquarie Business School including Chi Truong, Jeffrey Sheen and Stefan Trueck. Attendants included a wide range of researchers and Higher Degree Research Students from the Department of Economics, Applied Finance, and Actuarial Studies and Business Analytics.

The workshop started with a presentation by Mr James Villafuerte (ADB) on the economic outlook and key risks for the world economy and the Asian region. According to Mr Villafuerte’s presentation, the global economy is forecast to continue to slow down in the next couple of years, with the growth of major industrial economies expected to reduce from 2.2% in 2018 to 1.9% in 2019 and 1.6% in 2020. This is mainly due to the decrease in global trade caused by tensions between major economies such as the US and China. In Asia, an economic slowdown is predicted for the majority of countries, with the economic growth of China predicted to decrease from 6.6% in 2018 to 6.3% in 2019 and 6.1% in 2020.

Main sources of risk include US-China trade tension, increasing leverage, a so-called hard Brexit, and other policy uncertainty. Trade policy uncertainty has increased to a record high level in January 2019, and the trade growth for ASEAN, China and Japan has slowed down, reaching negative territory in January 2019. Nevertheless, financial stress has stayed at a relatively low level, indicating that the market is not expecting a crisis soon. However, sudden movements of financial stress in the past suggest that it is important to closely monitor financial vulnerability.

Mr Villafuerte also pointed out that the ADB currently uses an Early Warning System model that was developed in 2000 for currency and banking crises, based on 45 economic and financial indicators. Although the model is useful in keeping policy makers vigilant about systemic risk, its drawbacks include sending too much noise and failing to capture the GFC. Also, its focus was on currency crisis, while key vulnerabilities have shifted to other areas. One of the key objective of the project conducted jointly with research from MQ was to improve the Early Warning System used by the ADB.

Dr Chi Truong then presented the results of the ADB-MQ project on a new Early Warning System. Dr Truong pointed out that the new framework has been developed based on the original ADB model, but introduced several new features. First, the new framework defines crises based on a financial stress index that can capture currency crises, banking crises, debt crises and financial crises. Second, a macro-finance dynamic factor model is used to summarize the information from a big, unbalanced and mixed frequency data set into four indices. This helps to capture not only information from domestic economic and financial variables, but also information from global economic and financial indicators.

The model is particularly useful in that it allows to capture information from frequently updated financial variables such as the prices of credit default swaps, and the default risk of major banks. Dr Truong pointed out that when testing the model, in-sample results suggest that the derived indicators provide additional explanatory power compared to standard indicators used in the original ADB model. He also emphasized that the new model yields a relatively high proportion of correct predictions, at the same time having a low noise to signal ratio in the out-of-sample test. Overall, the developed model provides a new milestone in crisis prediction for Asian economies and will be implemented by the ADB to monitor these risks in several countries in the near future.