Closed codykallen closed 6 years ago
If there are no objections, I expect to merge this Thursday morning.
@codykallen, I admit I haven't had the time to follow in detail what you've been doing with BRC since our meeting in DC, but I think we all had a pretty clear understanding of where we were going. So, I'm confident that the changes you've been making are sensible and will make it easier to integrate BRC into Tax-Calculator. Thanks again for all the good work on this.
The new commit fixes the calculation of beta
in the AMTmodel()
function. The previous version rescaled both alpha
and beta
(the stochastic transition rates) based on the change in the fraction of firms on the AMT, but that is inconsistent with the underlying equations. The recent commit rescales alpha
and recalculates beta
based on the new values of alpha
and theta
.
This PR creates a new AMT model, based on the methodology in the document linked below. This new model uses stochastic transition probabilities to map shifts between AMT and non-AMT status, from which we can model use of the Prior Year Minimum Tax Credit. This change allows the eliminated of the
amt_repeal
special parameter, as the change in PYMTC availability is now mapped internally to theAMTmodel
function.New AMT & PYMTC.docx
This PR also changes the depreciation results very slightly. I discovered that I was using the wrong IRS corporate depreciation data. I previously was using depreciation for form 1120 corporations (C corporations), but I should have been using the series for 1120, 1120-A, 1120-F, 1120-L and 1120-PC. All of these forms are subject to the corporate income tax, but form 1120 claims almost all depreciation deductions; consequently, this only changes the depreciation results very slightly.
@andersonfrailey @hdoupe @martinholmer @Abraham-Leventhal @lucassz