Hi Statalists!
When replicate the Table 2 of "Banks’ Motivations for Designating Securities as Held to Maturity", I have difficulty in running the spline regression and interpreting the results. 1) I want to make sure whether my model specification is correct. There are three categories of banks (indicator: AA, Opt-out and non-AA) and four subperiods (period count: Pre-Basel III QC, Basel III QC, Tailoring rule QC and Interest Rate Rise QC), this paper uses the interaction between bank category and subperiod as independent variables as shown below. Does that mean I can use "xtreg y AA*Pre-Basel III QC ......non-AA*Interest Rate Risk QC CV, fe vce(cluster banktype)"?
2)The interpretation of results: I am confused why the note says that "In columns (1) – (4), the coefficients on the interaction terms are the time sensitivities of the dependent variables for non-advanced approaches (non-AA) banks changed over the 2012−2022 sample period." (highlighted in red column above). The change in the coefficient for the same category shows the change in sensitivity within the category, however, how can I compare the sensitivity across different bank categories? Do I need to set different interactions for each bank category?
Reference
Kim, Sehwa and Kim, Seil and Ryan, Stephen G., Banks’ Motivations for Designating Securities as Held to Maturity (October 29, 2023). Columbia Business School Research Paper No. 4452667, Available at SSRN: https://ssrn.com/abstract=4452667 or http://dx.doi.org/10.2139/ssrn.4452667
When replicate the Table 2 of "Banks’ Motivations for Designating Securities as Held to Maturity", I have difficulty in running the spline regression and interpreting the results. 1) I want to make sure whether my model specification is correct. There are three categories of banks (indicator: AA, Opt-out and non-AA) and four subperiods (period count: Pre-Basel III QC, Basel III QC, Tailoring rule QC and Interest Rate Rise QC), this paper uses the interaction between bank category and subperiod as independent variables as shown below. Does that mean I can use "xtreg y AA*Pre-Basel III QC ......non-AA*Interest Rate Risk QC CV, fe vce(cluster banktype)"?
2)The interpretation of results: I am confused why the note says that "In columns (1) – (4), the coefficients on the interaction terms are the time sensitivities of the dependent variables for non-advanced approaches (non-AA) banks changed over the 2012−2022 sample period." (highlighted in red column above). The change in the coefficient for the same category shows the change in sensitivity within the category, however, how can I compare the sensitivity across different bank categories? Do I need to set different interactions for each bank category?
Reference
Kim, Sehwa and Kim, Seil and Ryan, Stephen G., Banks’ Motivations for Designating Securities as Held to Maturity (October 29, 2023). Columbia Business School Research Paper No. 4452667, Available at SSRN: https://ssrn.com/abstract=4452667 or http://dx.doi.org/10.2139/ssrn.4452667
Any thought
Comment