| Description |
1 online resource (57 pages) |
| Series |
IMF Working Paper ; WP/19/273 |
|
IMF working paper ; WP/19/273.
|
| Contents |
Cover -- Contents -- I. Introduction -- II. Data And Methodology -- A. Data -- B. Methodology: Word Vector Representation and Semantic Clustering -- C. Abstract Term Clusters for Five Different Sentiments -- D. Negative and Positive Sentiment -- E. Aggregated Sentiment Indices -- III. Evaluating "TEXT-BASED" EWIS -- A. Evaluation Metrics -- B. Benchmark index -- C. Prediction Performance -- IV. Robustness Checks -- A. Economic term clusters -- B. Different Time Windows -- C. Different Weights within Clusters -- D. Topic Modeling -- V. Limitations And Extensions -- A. Limitations |
|
3. Sentiment Indices Performance Compared to a Random Benchmark (12m -- Boxes -- 1. Word Vector Representations, Semantic Clustering, and Sentiment -- 2. Constructing Sentiment Indices based on Semantic Clustering -- 3. Negative Sentiment in Argentina in 2018 and 2019 -- Appendixes -- 1. Semantic Clusters -- 2. EWS -- 3. Time Series -- 4. All Metrics for All Countries and All Indices -- References -- References |
|
B. Extensions -- VI. Conclusion -- Tables -- 1. Sample of Crises Episode -- 2. Seed and Related Terms for Fear Language -- 3. Summary Evaluation of -- 4. F2 Scores for Specific Sentiment -- 5. F2 Scores for Aggregated Sentiment -- 6. Results Selected Economic Terms Clusters -- 7. F2 Scores for Selected Seed Terms for All Countries -- 8. Results for Selected Seed Words for Alternative Time Windows -- 9 Summary Results for Selected Topics -- Figures -- 1. Sentiment Indices Performance Compared to a Random Benchmark (24m -- 2. Crises (Start to Peak) and Crisis Sentiment Signals |
| Summary |
We construct sentiment indices for 20 countries from 1980 to 2019. Relying on computational text analysis, we capture specific language like "fear", "risk", "hedging", "opinion", and, "crisis", as well as "positive" and "negative" sentiments, in news articles from the Financial Times. We assess the performance of our sentiment indices as "news-based" early warning indicators (EWIs) for financial crises. We find that sentiment indices spike and/or trend up ahead of financial crises |
| Notes |
Print version record |
| Subject |
Economic indicators.
|
|
Economic indicators
|
| Form |
Electronic book
|
| Author |
Simpson, Sean, author
|
|
Ulybina, Daria, author
|
|
International Monetary Fund, issuing body.
|
| ISBN |
151352318X |
|
9781513523187 |
|
