Econometric Trump Tweet
Venue: LKSLIB Project Room 4-2 Time: 20-Apr-2019 15:00 - 16:00 (1hrs)
- 4.16: Data collecting
- 4.18: Task Distribution
- 4.20: Determined data & methodlogy
- 4.24: initial result
- 4.27: report writting & presentation prepartion
http://www.trumptwitterarchive.com/archive
- positive
- negative
- positive
- negative
- Trump
s tweets
effect on companyies is more likely a Belated effort (does not change companyies)- postive & negative comparsion
- Possibile Reason
- Limitation
- small sample
- may influence industry more
- etc
- Introduction
- Literature Review
- Data
- Data Source
- Data processing
- Classfication
- Methodology
- Event Study
- Benchmark
- Results:
- Benchmark parameter
- AAR 以及 CAAR
- 可能要除去 earning announcement day 影响再讨论?
- Discussion Bias limitation
- Conclusion
- 确定只做 individual stock impact
- 重新增加并审视了应 drop 的关键词
- 丰富了审查的格式(多了一栏 mentioned company name)
- 讨论了可以作为 positive negative 判断的依据
- 清理出 891 条 tweet
- positive/ negative judgement:
- Positive Normal type
- Thanks for investment/ investing
- Great meeting with CEO/ manger
- Plant/ Job created
- Negative Normal type
- Firms Move to Mexico,
- prices are out of control
- Threating firms due to Mexico
- fake news (CBS ABC CNN NBC)
- Bad acts, fines and penalties
- Positive Normal type
- 在周四晚上下课前,给 tweet 区分好属性
- 周四下课后小聚一会儿,讨论细致分工
- 说明:为了让大家对 Trump tweet 有感觉,每个人有要看的工作量
辛苦夏老师 🌝
- 清理指南:
- 待清理数据位置:TrumpTweetsProject/code/tweets_data.xlsx (注意,是 excel文件,不是csv)
- 在第4列(E列)输入数字:1 or -1
- 1: postive event type
- -1: negative event type
- 在5列(F列)输入 event tweet 梗概
- 无关的事件不做标记
- 整理名单 以 excel 的 row index 为序号(不是 A 列)
- 1-200:ZPH
- 201-400: ZZM
- 401-600: Lance
- 601-800: SL
- 801-891: ZLC
positive negative judgement 仅供参考,可能还有新的类型,具体问题具体分析,有疑问可以另行标注后续讨论