Download This PaperBuilding the Bayesian Network Model of Digital Images Portfolio
Proceedings of the 32nd International Business Information Management Association (IBIMA), ISBN: 978-0-9998551-1-9, 15-16 November 2018, Seville, Spain, pp. 4279-4284.
6 Pages Posted: 28 May 2020
Date Written: October 1, 2018
Abstract
The main idea of this work is to rebuilding the intellectual assets portfolio model taking into account strong heterogeneity of its structural components. Our research shows that more than 40% of assets may be not saleable at all and thus not influence to overall earnings and portfolio risk metrics (EaR). There are asset groups doing very small financial returns also. Together with not saleable group they may occupy up to 90% of total portfolio. At last, the sales leader’s groups, doing main return, may present just 6-10% of total assets number, but bring more than 80% of total portfolio return. Thus the structure of portfolio may be very heterogeneous, which probably is explaining sufficient asymmetry of earnings random variable distribution about normal in daily horizon. We offer more sophisticated probabilistic portfolio model, which allow us to take into account not only sufficient quantitative heterogeneity, but also causal relations between asset groups, mapped on the principle of comparable demand. We prove that there is the process of permanent overflowing of assets in the direction of sales leader’s group, and with this flow the risk parameters of portfolio are changing.
Keywords: bayesian network, intellectual assets portfolio, conditional probability
JEL Classification: O34, C11
Suggested Citation: Suggested Citation