Munich Personal RePEc Archive

Synthetic data. A proposed method for applied risk management

Carbajal De Nova, Carolina (2017): Synthetic data. A proposed method for applied risk management.


Download (4MB) | Preview


The proposed method attempts to contribute towards the econometric and simulation applied risk management literature. It consists on an algorithm to construct synthetic data and risk simulation econometric models, supported by a set of behavioral assumptions. This algorithm has the advantage of replicating natural phenomena and uncertainty events in a short period of time. These features convey economically low costs besides computational efficiency. An application for wheat farmers is developed. The efficiency of this method is confirmed when its results and statistical inference converge with those generated from experimental data. Convergence is demonstrated specifically by means of information convergence and diminishing scaling variance. Modifications on the proposed algorithm regarding risk distribution parameters are not onerous. These modifications can generate diverse risk scenarios seeking to minimize and manage risk. Hence, risk sources could be anticipated, identified as well as quantified. The algorithm flexibility makes risk testing accessible to an ample variety of entrepreneurial problems i.e., public health systems, farmers associations, hedge funds, insurance companies; etcetera. This method could provide grounded criteria for decision-making in order to improve management practices.

Logo of the University Library LMU Munich
MPRA is a RePEc service hosted by
the University Library LMU Munich in Germany.