Investment efficiency of life insurance companies in Germany: application of a two - stage SBM

Thomas Krupa , Kirils Farbarzevics , Bassam Salame

Abstract

Purpose – To prove the robustness of the efficiency-measuring model against potentially system-relevant disturbances to company variables such as SIZE, ROA, solvency and organizational form. Methodology – In the first stage, the established model is applied using the SBM to measure insurance efficiency. The underlying data sets are from the twenty biggest life insurance companies (2008-2017) in Germany. In the second stage, the established model is examined for its robustness against disturbance variables. Several disturbance variables are introduced individually to the system and examined for their influence by three econometric methods, Tobit regression, OLS and the fixed-effect model. This approach allows a comparative analysis of the results with respect to the systemic relevance of every added variable. In the end, the accuracy of the second stage is compared through the Spearman test. Findings – The comparative analysis of all three econometric techniques brought ROA as an efficiency-influencing variable. Furthermore, both proved econometric models Tobit and OLS are SBM-suitable with cross-sectional data. Further evidence for SBM compatibility are found for Tobit and the fixed-effect model with panel data.
Author Thomas Krupa (FE)
Thomas Krupa,,
- Faculty of Economics
, Kirils Farbarzevics (FE)
Kirils Farbarzevics,,
- Faculty of Economics
, Bassam Salame (FE)
Bassam Salame,,
- Faculty of Economics
Journal seriesWspółczesna Gospodarka, ISSN 2082-677X, (B 9 pkt)
Issue year2019
Vol10
No1
Pages79-91
Publication size in sheets0.6
Keywords in EnglishDEA - SBM, efficiency, tobit, OLS , investment, insurance
DOIDOI:10.26881/wg.2019.1.08
URL https://czasopisma.bg.ug.edu.pl/index.php/wg/article/download/2965/2394/
Languageen angielski
Score (nominal)9
ScoreMinisterial score = 9.0, 11-04-2019, ArticleFromJournal
Ministerial score (2013-2016) = 9.0, 11-04-2019, ArticleFromJournal
Citation count*
Cite
Share Share

Get link to the record


* presented citation count is obtained through Internet information analysis and it is close to the number calculated by the Publish or Perish system.
Back