Research @ UNSW

Introduction

The ESW is an experimental software system developed at UNSW. Its purpose is to facilitate conducting event studies for non-technical users. Its main feature is a user interface (see Figure 1), which offers a seamless experience to users and enables an event study to be conducted in just a few clicks. The user interface of ESW allows the users to focus on the mechanics of the event study as opposed to worrying about the computational knowledge required to conduct the analysis. ESW is only a front event for pre-processing data and relies on another software products to conduct the event study per se.

Research Team:
  • Prof. Fethi Rabhi
  • Lawrence Yao
  • Weisi Chen
  • Ali Behnaz
  • Islam Al-Qudah
  • Florian Guerin
  • Kim-San Dok
  • Christopher Tin-Loi
  • 1. Background

    1.1 What is an Event Study?

    The event study method is a popular form of financial analysis, which measures the impact of a specific event on the value of firms, by evaluating the abnormal returns associated with the event [1].

    Given rationality in the marketplace, stock prices immediately reflect the true value of firms, incorporating all relevant information including the discounted value of all future cash flows [2]. The effects of an economic event on the value of firms can hence be measured using stock prices observed over a relatively short time period.

    Event studies can be applied to both firm-specific events such as mergers and acquisitions or earnings announcements, as well as economy-wide events such as changes to tax law. Warren-Boulton and Dalkir [3] illustrated some examples of event studies on Australian companies.

    An event study consists of the following tasks [4]:
  • Defining the event of interest and identifying the event window – the period over which the stock prices of the firms associated with the event will be examined;
  • Determining the criteria for including a firm in the study;
  • Defining the estimation window, usually the period prior to the event window, over which the normal return of the firm will be modelled;
  • Calculating the abnormal return by taking the difference between the actual return of the stock over the event window and the normal return; and Making statistical inferences about the abnormal returns with the null hypothesis that the event had no impact.
  • 1.2 Software packages for conducting event studies

    There are many software packages designed to assist users in conducting an event study, e.g. SAS, SPSS (Statistical Package for the Social Sciences), Stata, Eventus, etc.

    Eventus is one of the most popular software used to conduct event studies. Normally, users can use SAS on their local PC and remote sign into the WRDS server to submit a query to run Eventus [5]. Eventus mainly uses CRSP (Center for Research in Security Prices, http://www.crsp.com/) US stock databases or user-collected data to perform event studies.

    For non-US data, users have to pre-process native data from other data sources to comply with Eventus format. Eventus requires the following to run an event study:
  • Stock return data, computed from stock price data and adjusted for corporate actions such as dividends, splits and spinoffs,
  • Index return data, computed from value-weighted index price data, including dividends, and
  • At least one stock identifier and corresponding event date.

  • Eventus also provides user control over estimation periods and event windows so that users are able to define their own event studies with their own parameters. The following options are available:
  • EstimationPeriodLength: specifies the estimation period length. By default, the estimation period is 255 trading days long.
  • EstimationPeriod: specifies the last day of the estimation period. By default, the last day of the estimation period is 46 trading days before the event date. Estimation periods before and following the event date are specified as negative and positive numbers respectively.
  • PreWindow: specifies the number of trading days before the event date to report abnormal returns and associated test statistics.
  • PostWindow: specifies the number of trading days following the event date to report abnormal returns and associated test statistics.


  • References

    [1] Fama EF, Fisher L, Jensen MC & Roll R 1969, 'The adjustment of stock prices to new information', International Economic Review, vol. 10, pp. 1-21.

    [2] McWilliams, A & Siegel, D 1997, 'Event studies in management research: theoretical and empirical issues', Academy of Management Journal, vol. 40, no. 3, pp. 626-657.

    [3] Warren B., Dalkir, F. & Dalkir, S. 2001. ‘Staples and Office Depot: An Event- Probability Case Study’. Review of Industrial Organisation, vol. 19, pp. 29-50.

    [4] MacKinlay, AC 1997, 'Event studies in economics and finance', Journal of Economic Literature, vol. 35, no. 1, pp. 13-39.

    [5] Cowan, AR 2007, Eventus 8.0 User's Guide, Cowan Research LC, Ames, Iowa.

    2. Conducting an Event Study with the ESW:

    You may follow the tutorial below or download a formatted version here.

    Background info:

    Consider the following telecommunications company: Vodafone Qatar (VFQS.QA). In a recent Thomson Reuters Middle East Market Update on Feb 4th 2015, Vodafone Qatar was said to have surged by 7%, while its trading volume jumped to a seventh month high. The firm announced the same week that it had become fully Shari'a-compliant, opening the stock up to Islamic funds.

    This occurred despite a Jan 29th 2015 announcement by the company reporting a quarterly loss of almost 4% - equivalent to almost 68.8 million riyals ($18.9 million). This announcement was undoubtedly related to Vodafone Qatar scrapping its bid for the Qatar National Broadband Network (QNBN) on Nov 29th 2014 as well as several other factors.

    The following tutorial details an event study surrounding the aforementioned events and their impact on the Qatar Exchange All Share Index (.QEAS), as well on the Qatar Exchange Telecommunication Index (.QETLC).

    Tutorial:

    1. Click New Study + in the Current Event Studies pane. This will open a new window on the right hand side.

    2. Enter a relevant:
  • Event Study Name
  • Event Study Description
  • Company/Date list: formatted as [ric],[date] where [date] is of the format dd-mmm-yyyy, e.g. BHP.AX,25-Dec-2014

  • You can either create a text file with ric/date and Upload Existing List, or Construct New List and enter the ric/date directly into the text box. The date will correlate directly to your “day 0”. In this case there were several news announcements over a period of 3-4 months. So as long as our day 0 lies within this region we’re satisfied. For this tutorial we will use the following ric/date pair: VFQS.QA,01-Feb-2015

    3. Concentrate on the Stage 1 pane. This stage relates to a Thomson Reuters Tick History (TRTH) import workflow of daily returns data.
    a. Enter TRTH Credentials: Hopefully your supervisor has provided you with credentials, or has enabled the auto-credential feature.
    b. Define Import Window: This will define the range of the import window. The defaults (200,200) set the TRTH import to get a 400 day window of data (200 days before and after your specified day 0).The defaults are fine as is.
    Now click the START button to the right of the Stage 1 pane. When the progress bar has finished, you may download the data and then go to the next step.

    NOTE: If you set your window too large it may take a long time to download the data.

    4. The next stage focus on is the Stage 2 pane. This stage relates to a Compute Adjusted Returns workflow. This step is simple - if the previous stage completed successfully (green bar) all you need to click is the START button to the right of the Stage 2 pane . When the progress bar has finished, you may download the data and then go to the next step.

    5. In order to run Stage 3 of the process (the Event Study Workflow) you will need to import data from an Indices. This will allow the workflow to derive a relationship between the firm’s stock and a reference index. Click on New Index + in the Current Indices pane. This will open a new window on the right hand side.

    Enter a relevant Index Name and Index Description.

    6. Focus on Stage 1 of your created index. This stage also relates to a Thomson Reuters Tick History (TRTH) import workflow of daily returns data. a. Enter TRTH Credentials: Use the same credentials as before b. Benchmark Code: Enter the code of the Index. In this case you may either use .QEAS or .QETLC Now click the START button to the right of the Stage 1 pane. When the progress bar has finished, you may download the data and then go to the next step.

    NOTE: This stage will take a long time due to TRTH importing all historical data of that index. Indices are often reused between multiple event studies. If an index has already been created omit steps 6-7.

    7. Focus on Stage 2 of your created index. This stage also relates to a Compute Adjusted Returns workflow. Click the START button to the right of the Stage 2 pane and wait for stage to complete.

    8. Go back to Stage 3 of your previous Event Study.
    a. Select Adj Return Source: From Compute Adj. Returns
    b. Select Index Source: Select your newly created index from the list (the index will only show up of both stages of the index have been successfully completed)
    c. Define Event Window: These parameters will define your Event Window
    i. Before Event: Specifies min range of the event day. Leave as default: 1
    ii. After Event: Specifies max range of the event day. Leave as default: 1
    iii. Before Window: Specifies min range of the event window. Leave as default: 30
    iv. After Window: Specifies max range of the event window. Leave as default: 30

    Start Stage 3.

    After Stage 3 succeeds you may download the data and then finally Visualise the results.