Work detail

Bidding patterns in public procurement auctions

Author: Bc. Jakub Kožíšek
Year: 2015 - summer
Leaders: PhDr. Ing. Jiří Skuhrovec PhD.
Work type: Bachelors
Language: English
Pages: 71
Awards and prizes:
Abstract: Public procurement market and auctions by which contractors are selected is
difficult to analyze using standard econometric models, for their heterogeneity
and the lack of complete data. This thesis expands on econometric models,
using simulation of real life events in an experimental environment – an agent
based model. The chosen ground-up approach allows for better understanding
of market behaviors and patterns and to predict these behaviors, depending
on the given conditions. With the analysis of how particular agents react to
their changes, this represents a tool for evaluating different legislative proposals
with respect to market efficiency. As the main objective of the thesis, factors
affecting bidding patterns and therefore auction outcomes are studied using real
data from the Czech Republic in an econometric model and then four possible
legislative changes are tested using a custom made simulation model. Out of
four proposals, the minimal requirement on the number of auction participants
is considered a viable option in reducing procurement cost and is recommended
for further evaluation. The complex model itself is released under permissive
license to encourage broader analyses of the public procurement market.




Patria Finance