Hey I am doing an analysis,

I have 12 securities with weekly observations regarding spread and volume. There was a certain law which I want to test if it significantly influenced liquidity once it got implemented.(liquidity independent variable) Hence, I am testing whether the spread increased or the volume declined after the implementation compared to the period before. Consequently, I am doing an event window period for the time before the law and the time after.

My regression looks like this:

Liqi,t = a + b1 *spreadi,t +b2 * volumei,t + b3 * Dt * spreadi,t * b4 * Dt * volumei,t

i = security t = time

b1 would be interpreted like if the spread goes up by 1 than liquidity should also go up by b1 (same for volume right?) Dt is a time dummy which is 1 for everything from law and later, 0 for everything for the period before. So how can I interpret b3 and b4 regarding liqudity? Also if the signs of the coefficents b3 and b4 are different than from the signs of b1 and b2 what does that mean?

I have observations for the period before and for the period after. The number of firms is constant for all observations and there is no change in firms for the whole data set. For my data set I only have firms that were subject to the law. That is why I created the time dummy variable to measure if the event lead to a significant change in liquidity. Since I don't have observations regarding firms not subject to the law the difference-in-differences estimate can not be applied.

All in all:

I try to write a regression that gives me the answer to the question whether liquidity changed after law by observing spreads and volume

It got pretty long. Thank you for your help!

If it helps you I attached a sample of my Data set.

Liquidty is the dependent variable NOC = volume Pread = spread DummyNOC 0 or 1 * NOC for i on time t (same for spread)

is it wrong that I already included in the excel the result (ie. 1 * 1.62)? Or do ihave to keep the dummies only 1 and 0 and tell stata to multiply it? enter image description here


You could create groups from the data, one group for before the change and one for after the change.

Then you could use an "F test about the variance" to test whether the variance between the two groups are significantly different.

Likewise, with the same two groups, you could use ANOVA to test whether the mean of the "volume" variable has changed significantly.

| cite | improve this answer | |
  • $\begingroup$ thank you for your help! using the F-Test would not give me information whether liquidity has gone up or down right? since I only have two variables is an independent-samples t-test better? and what about my suggested regression? are the dummies bad or is the approach wrong? $\endgroup$ – Diether Busch Apr 20 '18 at 14:48
  • $\begingroup$ The F test alone would not, but if the test did show a significant difference you can assume the group with the greater mean is greater. Or you could perform a tukey post hoc if you really wanted to. Though with only two variables I don't see why this would be necessary. I think the dummy vars Dt would work (it's hard for me to thing through it without data), but those binary variables are essentially creating a factorial ANOVA, which I use more and am more comfortable with. $\endgroup$ – Michael Parent Apr 21 '18 at 15:21
  • $\begingroup$ I added the dataset so you can have a better understanding of my problem $\endgroup$ – Diether Busch Apr 28 '18 at 10:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.