What Types of Events Provide the Strongest Evidence that the Stock Market is Affected by Company Specific News - PowerPoint PPT Presentation

oshin
slide1 l.
Skip this Video
Loading SlideShow in 5 Seconds..
What Types of Events Provide the Strongest Evidence that the Stock Market is Affected by Company Specific News PowerPoint Presentation
Download Presentation
What Types of Events Provide the Strongest Evidence that the Stock Market is Affected by Company Specific News

play fullscreen
1 / 37
Download Presentation
What Types of Events Provide the Strongest Evidence that the Stock Market is Affected by Company Specific News
295 Views
Download Presentation

What Types of Events Provide the Strongest Evidence that the Stock Market is Affected by Company Specific News

- - - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript

  1. What Types of Events Provide the Strongest Evidence that the Stock Market is Affected by Company Specific News Calum Robertson, Shlomo Geva, Rodney Wolff AusDM2006

  2. Outline • Introduction • Data • Methodology • Results • Conclusions and Future Work

  3. Outline • Introduction • Data • Methodology • Results • Conclusions and Future Work

  4. Introduction • The efficient market hypothesis • Is stock market an efficient market? • Types of News • Macroeconomic news: relatively infrequent, scheduled • Company specific news: more frequent, unpredictable • Purpose of this paper • Identify events which occur with a high correlation to the occurrence of real time news

  5. Outline • Introduction • Data • Methodology • Results • Conclusions and Future Work

  6. Data – Source and Collecting Period • Source • Bloomberg Professional service • S&P 100, FTSE 100, and ASX 100 • Collecting Period • Stock: 1st July 2005 – 1st September 2006 • News: 1st May 2005 – 31st August 2006 • : the period of data collection • : the business time scale

  7. Data – stock dataset • The stock time series • The date time series • The price time series • The volume time series • The tick time series

  8. Data – news dataset • Articles set for a stock (s) • The news time series

  9. Outline • Introduction • Data • Methodology • Results • Conclusions and Future Work

  10. Preprocessing (1/3) • The return time series for a stock (s) • The change in volume time series for a stock (s)

  11. Preprocessing (2/3) • The change in ticks time series for a stock (s) • The volatility time series for a stock (s)

  12. Preprocessing (3/3) • Grouped stocks set • Divided time windows

  13. How to define events (1/2) • The Event Point Process (EPP) • F – a generalized time series, one of the return, volume, tick, and volatility time series • x – the specified threshold, a threshold of 10% means that the value should be ≥ log(11/10) or ≤ log(10/11)

  14. How to define events (2/2) • The Event given News Point Process (ENPP) • The Event Without news Point Process (EWPP)

  15. The Ratio of Events Related to News to Events (RERNE) • The Ratio of Events Related to News to Events (RERNE) • The probability that the events for the given parameters are preceded by news, which would imply that news is responsible for these events.

  16. The Likelihood that Events are Related to News (LERN) • The Benchmark • A measure of the likelihood of news arriving within the specified Δτ time. • The Likelihood that Events are Related to News (LERN)

  17. The Event T-Test (ETT) • The Event T-Test (ETT) • Test the null hypothesis that the occurrence of events is not influenced by the occurrence of news

  18. The News T-Test (NTT) • The News T-Test (NTT) • Test the null hypothesis that the occurrence of news before events is the same as the occurrence of news normally

  19. Outline • Introduction • Data • Methodology • Results • Conclusions and Future Work

  20. Characteristics of Dataset

  21. The Choice of Parameters

  22. Average number of minutes between events - Return

  23. Average number of minutes between events-Volume

  24. Average number of minutes between events-Ticks

  25. Average number of minutes between events-Volatility

  26. RERNE and LERN- Return

  27. RERNE and LERN- Volume

  28. RERNE and LERN- Tick

  29. RERNE and LERN-Volatility

  30. ETT and NTT - Return

  31. ETT and NTT -Volume

  32. ETT and NTT -Ticks

  33. ETT and NTT -Volatility

  34. Outline • Introduction • Data • Methodology • Results • Conclusions and Future Work

  35. Conclusions • The stock market does react to real time news • Return and volatility appear to give the most compelling evidence • 5% threshold for all countries • 2% and 10% thresholds for the UK and Australian markets • There appears to be some weak evidence that news effects volume and tick events

  36. Future Work • To determine if volatility events occur differently when the market reaction period is changed • When events and news occur is necessary to establish if the market behaves in a uniform manner • The content of news which the market reacts

  37. Thank you very much – The End –