Dear statalists,

I am running 12 different regressions of political risk factors on the amount of M&A deals, whereas the control variables are the same in each case. However, I was wondering how I can test, which risk factors is more important / significant, since each regressions shows significance.

That is my code
Code:
reg ln_Number_of_Deals_A Government_Stability_A win_FirmSize_A win_Market_to_Book win_ROA win_Cash_To_TotalAssets win_Book_Leverage win_TotalReturn52Week Acquirer_Age HIGH_TECH_DUMMY DIVERSIFYING_DUMMY HOSTILE_DUMMY PUBLIC_TARGET_DUMMY SameReligion SameLanguage Geographical_distance lag1_GDP_growth_annual_percent_A lag1_BCI_A i.timeFE i.AcquirorPrimarySICCode_ID21 i.AcquirercountryFE if CrossBorder =="Cross-Border" & AcquirorPublicStatus =="Public", robust
Also, is the command
robust
sufficient for clustering standard errors or would you rather commend to cluster by year and nation by using
vce(cluster year nation)
.

VIF shows values around 1-2.