Asia Pacific Institute of Research

Business sentiment indices play a crucial role in government's policy decisions, production planning in industries, and individual/institutional investment decisions. However, such indices require large-scale surveys, which are costly and laborious to conduct. As an alternative, this paper reports our ongoing work on estimating business sentiment based on abundant, ordinary newspaper articles. Our proposed framework is composed of a recurrent neural network for regression, a one-class support vector machine to filter out topically irrelevant articles, and fine-tuning for domain adaptation. The effectiveness of the framework is empirically demonstrated by experiments on Nikkei newspaper.

This study is an attempt to assess the impact of policy initiatives launched by Japan’s new Prime Minister Shinzo Abe on Japan’s real gross domestic product (GDP) in his first quarter in office. We use as a benchmark for measurement a counterfactual estimate of GDP. Since the Japanese economy is also in the midst of reconstruction from the 2011 Tohoku disaster in the first quarter of 2013, we first estimate the counterfactual GDP that would have materialized in the absence of that disaster. We will use a dummy variable method and the statistical method proposed by Cheng Hsiao and others. We check the validity of these methods with regard to the Kobe earthquake of 1995 and then estimate the post-disaster counterfactual GDP in the absence of the Tohoku disaster. We measure the impact of government policies as the difference between the actual and counterfactual GDP. By doing so, we conclude that government policies have failed to lift Japan’s GDP to the expected level. Even with the help of Abenomics, the gap remains in the range of 6 to 14 trillion yen per year.