Proceedings of the Second International Workshop on Data Science for Macro-Modeling 2016
DOI: 10.1145/2951894.2951896
|View full text |Cite
|
Sign up to set email alerts
|

Linking Deutsche Bundesbank Company Data using Machine-Learning-Based Classification

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1

Citation Types

0
26
0

Year Published

2018
2018
2021
2021

Publication Types

Select...
5
1
1

Relationship

0
7

Authors

Journals

citations
Cited by 22 publications
(27 citation statements)
references
References 7 publications
0
26
0
Order By: Relevance
“…The second source of employment data is the Microdatabase Direct Investment (MiDi, see Lipponer, 2011;Schild and Walter, 2017), which contains information on all German firms with inward or outward stocks of Foreign Direct Investment (FDI) above a reporting threshold. This threshold has been unchanged since 2002 at a minimum of 10% shares or voting rights in an affiliate with a balance sheet total exceeding e 3 million.…”
Section: Data Sourcesmentioning
confidence: 99%
See 1 more Smart Citation
“…The second source of employment data is the Microdatabase Direct Investment (MiDi, see Lipponer, 2011;Schild and Walter, 2017), which contains information on all German firms with inward or outward stocks of Foreign Direct Investment (FDI) above a reporting threshold. This threshold has been unchanged since 2002 at a minimum of 10% shares or voting rights in an affiliate with a balance sheet total exceeding e 3 million.…”
Section: Data Sourcesmentioning
confidence: 99%
“…The three micro-level datasets are combined via firm identification numbers, which are identical in SITS and MiDi. They are matched to USTAN via a correspondence table provided by the Deutsche Bundesbank (see Schild et al, 2017), resulting in the new combined dataset. Since USTAN does not cover the full population of German firms, and since by far not all firms are involved in service trade or FDI, the three data sources overlap only imperfectly.…”
Section: Data Preparationmentioning
confidence: 99%
“…In case more than one new affiliate was established in a given year, we compute sums or weighted averages across newly established affiliates using fixed-asset weights.14 Using gross investment avoids problems related to reporting the book value of fixed assets and depreciations, which may in part be related to tax considerations.15 For more details regarding the matching methodology and the quality of the match, seeSchild and Schultz (2016).16 Note that T A i,t−1 (Affiliates) and M N E i,t−1 are set equal to zero in case no such activity is observed, which is naturally the case when firm i is a domestic one.17 Note that for illustration purposes we do not use all information from MiDi inFigure 1, since for reasons of confidentiality, regions with fewer than three parent companies cannot be displayed.…”
mentioning
confidence: 99%
“…From 2002 until 2006, FDI had to be reported if the participation was 10% or more and the balance-sheet total exceeded 3 million Euros. Since 2007 the threshold of 10% applies only for direct shares, whereas for indirect shares or a mixture of direct and indirect shares the threshold has been raised to 50% (for details see Schild and Walter (2016) Whereas GDP and GDP per capita may increase the probability of holdings in the same country since they indicate the market size and purchasing power (and hence maybe profits which could be offset by interest payments) they may reduce the probability of holdings in an intermediary firms with large sales) it is assumed to be easier to receive external debt financing (see, e.g., Graham and Harvey (2001)). On the other hand, large firms may typically be more mature and hence might rather finance themselves with retained earnings (Ruf, 2011).…”
Section: Datamentioning
confidence: 99%