Formation Of Indonesia Economic Policy Uncertainty Index For 2014-2021
Puti Adani, Arief Wibisono Lubis

Universitas Indonesia

Abstract

Information contained in newspapers is one factor that determines decision-making at the individual level or on a larger scale. We try to use the information in newspapers to create a proxy for measuring uncertainty by forming the Indonesian Economic Policy Uncertainty Index for 2014-2021. The index was built by analyzing an extensive data set of newspaper articles containing the terms ^economics^ and ^uncertainty^ and similar terms. We were then searching for topics related to ^policy^ in the article, using the text-mining topic modeling method of the Latent Dirichlet Allocation model. Based on the frequency of article publication, an index with a mean of 100 and a standard deviation of 1 was formed for the main index and the index per policy topic. Then the movements and variations of the index are given meaning according to what, who, when, and the effect of a policy. In addition, we also make comparisons with pre-existing uncertainty indices, namely the VIX index, the Global Economic Policy Uncertainty (GEPU) index, and the World Uncertainty Index (WUI) for Indonesia. The IEPU index then consisted of 10 policy topics divided into eight main policy categories- monetary, fiscal, trade, domestic regulation, international regulation, geopolitics, energy/resources, and politics. This category was used to discuss variations in index movements that occur. The results for the comparison with the global VIX and GEPU indices show a negative correlation, whereas when compared with the WUI index for Indonesia, it produces a positive correlation.

Keywords: Economic- Policy- Uncertainty- Uncertainty Index- Text mining

Topic: Financial Management and Accounting