0
$\begingroup$

I do have a theoretical question about a difference-in-difference (DiD) model I constructed in Stata.

Construction: Over a monthly time span of 7 years I try to analyze the impact of investment-volume (IV) on the stock return of some companies (DV). Therefore, I use the Fama–French Five-Factor Model including the IV.

Now, I want to see if there is a difference between companies who have financial advisors compared with those who do not. Therefore, I added a dummy variable (=1 if they have a financial advisory and =0 if they do not) and created the treatment group. However, I am not sure if I understand the methodology behind my method. Do I compare them over the whole sample period which includes multiple time periods, not just two.

Therefore, my method does not observe outcomes for two groups for two time periods, whereby one group (i.e., treatment group) is affected by a treatment in the second period and the second group (i.e., control group) is not affected by the treatment at all.

Or did I misunderstand something?

Improve this question
$\endgroup$
8
  • $\begingroup$ Welcome. Is your treatment “investment volume” or the placement of a financial advisor? Also, does your treatment start at the same time for all companies? Please clarify. $\endgroup$ – Thomas Bilach Jan 22 at 0:43
  • $\begingroup$ Hi, the treatment is the investment Volume. In general I want to investigate if there is a difference in the stock return for companies when the Investment Volume changes. Additionally, I want to look if the change of the stock return is different when Investment Volume changes plus the company has an financial advisor. The treatment starts at the same time. $\endgroup$ – Shorty20 Jan 22 at 7:24
  • $\begingroup$ So you have two treatments? I though the introduction of financial advisors to a subset of companies was the treatment of interest? $\endgroup$ – Thomas Bilach Jan 22 at 7:38
  • $\begingroup$ @ThomasBilach Thats my understanding issue. Maybe to illustrate it a bit better: Investment Volume is my first explanatory variable. I want to see if my stock return changes when the Investment Volume changes. After that I want to see which influence my Financial Advisor has on the stock return (second explanatory variable). The final step is to see what is the influence of the Investment Volume to the stock return of companies with an Financial Advisor. Edit: So I think you are right. my Financial Advisor is the treatment of interest $\endgroup$ – Shorty20 Jan 22 at 8:43
  • $\begingroup$ But don’t all companies have some dosage of investment volume? The goal of a DiD design is to assess some shock (i.e., treatment) that affects a subset of companies, while others remain unaffected. And please explain what you mean by “non-time dummy variable” in your question’s title? A time dummy is required; you need some way of delineating pre- and post-treatment. $\endgroup$ – Thomas Bilach Jan 22 at 19:03

Your Answer

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

Browse other questions tagged or ask your own question.