With the increased sample, we now find that LBOs are associated with reduced likelihood of patenting overall. However, this results is driven by the LBOs in period 1 from 1980 to 1995. In period 2, from 1995 to 2015, there is no statistically significant association. We also tried using decades but this wasn't instructive.
The matching scripts, LBOmatchingscript.jl and LBOmatching.jl, were rewritten by Ed to run with Julia 1.1.1. They use the LBOregmarker and regfilter -- we are matching at t-1 (or t-2). Note that we only export if regfilter==1 to MatchInput2V1.txt. All 179 patenting LBOs are in there, however, not all of them have a logitpreg: .count if hadlbo==1 & regfilter==1 179 .count if hadlbo==1 & regfilter==1 & n2 <. 179 . count if hadlbo==1 & regfilter==1 & logitpreg <. & n2 <. 173 This is because variables just aren't available to calculate them. This was explored and is the best it reasonably be after applying some manual fixes (see below). su tobinql if regfilter & hadlbo //175 su revgrowthl if regfilter & hadlbo //176 su ebitdatoassets if regfilter & hadlbo //179 ...all other vars have full coverage
To run the matching script:
julia -i LBOmatchingscript.jl
dumps you to an interactive shell afterwards to do diagnostics:
println("Number of LBOs with complete data available for matching", size(LBOs,1))
diagdf
173 in, 1 unmatched: 1×9 DataFrame │ Row │ gvkey │ year │ lboregmarker │ patentstock │ indu3 │ logitpreg │ hadlbo │ regfilter │ matchpair │ │ │ Int64 │ Int64 │ Int64 │ Int64 │ Int64 │ Float64 println("Size before drop ", size(LBOs,│ Int64 │ Int64 │ Float64 │ ├─────┼───────┼───────┼──────────────┼─────────────┼───────┼────────────┼────────┼───────────┼───────────┤ │ 1 │ 8367 │ 1983 │ 1 │ 27 │ 3 │ 0.00564947 │ 1 │ 1 │ -0.1)) │ ====Notes on manual fixes====
First run matches 162 (163 but no match 66). Although the 179 are going in, some are missing indu3 (because they are missing n2) and some are missing logitpreg.