I am a little bit stuck with my analysis regarding: the effect of liquidity on performance of carry trading during the recent financial crisis. Carry trading is an FOREX trading whereby someone borrows in a country with low interest rate and invest this money in a high interest rate country, gaining the interest rate difference for a certain time t.
My variables: I have 3 variables that measure liquidity let’s call them variable A,B and C (these are the independent variables). And off course I have the dependent variable called R (return/performance of the carry trade). All are measured in percentages.
My time series data: I have monthly data from 2002 till 2011. Let’s assume the crisis took place during 2008.
What I know and did: 1st regression: I want to know what the overall effect was of liquidity on the carry trade performance (variable R) during the entire sample period. This regression is just a basic multiple regression.
What I want to know: I want to know per variable (a, b and c) how it affected the performance (variable R) before, during and after the financial crisis, . So I need to make a dummy for these time period crisis. But these dummies (in my opinion, correct me if I am wrong) will indeed look at the needed time periods, but take all the variables in once shot. As said for each variable I want to compare its effect during the crisis and non-crisis period (before and after)
Attached is an example of my dataset in excel format. For simplicity I have assumed the following periods: 2007 is before crisis, 2008 is crisis and 2009 is after crisis.
How should I do this? Do I run several regressions? I am just not able to figure it out. So I ask the humble people of this forum for help.
Thanks,
Krishan
Bookmarks