***Data underlying the research on daily currency exchange rate using event study methodology***
Author:
Hon Chung, H. (Hui)||orcid:0000-0003-0738-6535
Contact information:
huihonchung@gmail.com / hc.hui@help.edu.my
Department of University of London Programmes, HELP Academy
Level 3, Block E, Kompleks Pejabat Damansara,
Jalan Dungun, Damansara Heights,
50490, Kuala Lumpur
***General Introduction***
The dataset is collected as part of a project on understanding the effects of political events on the behaviour/response of foreign exchange markets. The data has yielded a working paper that is currently being subjected to peer review. It is possible to use the data for other projects that require studying exchange rate trends.
The data was collected in HELP University, Kuala Lumpur in December 2018. The original source of the data was mostly from Bank Negara Malaysia (Central Bank of Malaysia) Monthly Statistical Bulletin.
***Data specific information***
The dataset contains the time series of 6 exchange rate variables, namely Ringgit/USD, Ringgit/EUR, Ringgit/HKD, Ringgit/SGD, Ringgit/Yen and Ringgit/GBP. Each series runs from 1 August 2005 to 30 August 2018 (daily data). Also included are domestic and foreign interest rates – the former is represented by conventional interbank rates (taken from http://www.bnm.gov.my/index.php?csrf=c6ae51093dbc4f24bafcc00e428eee8441b17e4e&ch=statistic&pg=stats_convinterbkrates&lang=en&tpl=&smonth=6&fyear=2006&emonth=6&eyear=2015 ) whilst the latter is Effective Federal Funds Rate taken from https://fred.stlouisfed.org/series/DFF
Each exchange rate is transformed into logarithms, yielding another 6 variables. After this, we take the first difference of the logarithms, yielding another 6 more variables.
The first difference of the logarithms of exchange rates are used to estimate a market model for exchange rates. The estimation window can be changed, but must be set prior to the event being studied. Concomitantly, the event window must also be determined.
***Methodological information***
Estimation window is from 29 March 2012 to 4 April 2013. The dataset contains 3 event windows i.e. 5 April 2013-30 July 2013, 4 June 2015-30 September 2015 and 9 April 2018-6 August 2018.
Taking the difference between the observed logarithmic first difference of the Ringgit/USD exchange rate and the predicted value from the estimated market model in a particular event window, gives us the abnormal returns (ARs). Summing up all the abnormal returns in an event window gives us the cumulative abnormal returns (CARs). Do this exercise for the 3 event windows.
The t-statistics for abnormal returns are calculated as (AR/standard deviation of AR) and CAR/standard deviation of CAR). Both the standard deviations are calculated from the regression output. Do this exercise for the 3 event windows.
The critical values for the t-statistics are bootstrapped. Take the t-statistics over an event window and randomly draw the t-statistics for 1000 times. Arranging them in ascending order, the 50th value represents the critical left tail value at 5% significance. The 950th value represents the right tail value at 5% significance. Do this exercise for the 3 event windows.