Negative Externality Generated By Risk Behavior Of Institutions With Diversification

As international financial market becomes interconnected,fintech speeds up fi-nancial transactions,and blockchain breaks the centralized management pattern,the underlying assets in the financial market are enriched increasingly.However,diversifi-cation does not efficiently reduce the risk of financial institutions,and financial crises erupt occasionally.Considering the increasing financial complexity,Basel III adds a new category of credit risk exposure,specifies the risk exposure of companies,real estate and retail,and adjusts the risk reserve of financial institutions.Besides,Basel III develops a combination of macro and micro prudential regulation.It is crucial to understand thoroughly the transmission mechanism of risk,and the important research approach is to predict the risk value via dynamically fitting the complex patterns of empirical data.This work derives that the first domino institution is available in the market under the assumption of heterogeneous risk.Through tail index and the large position in the investment,this institution dominates the risk of market.In this work,the negative externality is defined as the reduction of market value.We construct a bankruptcy index to observe the risk of institutions and market creatively,whose efficiency is also verified by empirical study.We propose the marked point process model including historical information to capture the extreme downside risk,and the Gaussian mixture model to simulate the other data.The innovations of this work are as the followings:First,it completes the effect mode and the transmission mechanism of the heterogeneous heavy-tailed risk within the system consist of the underlying assets,financial institutions and financial markets,and explains that there still exists the systematic risk under diversification from the points of view of both the behavioral portfolio theory and microstructure theory.Sec-ond,it adopts the dynamically Marked Self-exciting Point Process model to explicitly measure the extreme downside risk and to characterize the durations between extreme events,and simultaneously adopts the Gaussian mixture model to softly characterize non-extreme data.Third,our empirical study implies that,the mixed model not only recognizes the risk behavior of the institutions,but also forecasts explicitly the Value at Risk under different confidence coefficients.The main content is as following:In chapter one,we introduce the background of this research about diversification disasters and the relevant research in recent years.In addition,there is an introduction to the research and innovations of this work.In chapter two,we illustrate preliminaries that is necessary for the following chap-ters,including the point process characterization with extremes and the asymptotic property of heavy-tailed risk.In chapter three,we study the transmission mechanism of the risk in the financial system.First,a three-level value model is built for the underlying assets with cor-relations,financial institutions with sustained operation,and financial market within economic cycle.Then,while optimizing the financial institutions behavior,we find the strength of the correlation between underlying assets can be expressed as the index of heavy-tailed distribution of these institutions risk.Under the assumption of homoge-neous risk,institutions and the market share the same tail index,and severe heavy-tailed risk would hazard the value of institutions and market,and then influences the prefer-ence of sharing risk of them.Under the assumption of heterogeneous risk,we derive that the first domino institution is available in the market,and give the influence pattern of the institution's risk spillover to other institutions as well as financial market value.The work studies,for the first time,the fluctuation of the value of institutions and mar-ket via the bankruptcy index constructed from heavy-tailed distribution.We find that,the first domino attracts institutions with similar tail index to follow and imitate its risk behavior,which conducts the investment aggregating on the smallest tail index,and generates the negative externality of market by increasing bankruptcy index.In chapter four,based on the conclusion of chapter three,we include the time in the analysis and study the dynamical risk behavior of institutions,which would provide more accurate suggestions for risk management.We construct the marked point process with history information to capture the dynamic risk behavior that evolves randomly and is of volatility clustering.The current time interval is modelled upon the expectation of the historical time interval,and adjusted by the generalized Gamma distributed random error.The magnitudes is modelled to follow a generalized Pareto distribution.And we propose a flexible Gaussian mixture model to fit the non-extreme data.After training the whole model by financial data sets,we adopt Monte Carlo simulation to predict Values at Risk.The empirical results show that the diversified institution have more risk because of heavier tail and sensitivity to the intensity of time interval.In chapter five,we summarize the mechanism of negative externality rase from the correlation of underlying assets,and the approach of Marked Self-exciting Point Pro-cess model that dynamically captures the risk behavior of financial institutions.The limitations and unsolvable problems of this paper are also stated in this section.Fur-thermore,we put forward some respects and research plans about the risks of financial ecological environment.