管理工程学报
管理工程學報
관리공정학보
Journal of Industrial Engineering and Engineering Management
2014年
4期
146~151
,共null页
套期保值效率 复杂模型 模型(误设)风险 估计风险 RCMRS模型
套期保值效率 複雜模型 模型(誤設)風險 估計風險 RCMRS模型
투기보치효솔 복잡모형 모형(오설)풍험 고계풍험 RCMRS모형
hedging efficiency; complex econometric models; model-misspecification risk; estimation risk; RCMRS model
本文从模型风险的角度,选取环境突变样本区间,对“模型的复杂性和期货套期保值效率”这一问题进行研究.采用随机系数马尔科夫体制转换(RCMRS)模型对期货最优套期保值比进行估计.并将RCMRS模型套期保值效率和OLS、VAR、VECM及GARCH等模型进行比较和分析.样本内比较发现,复杂的动态模型并未带来明显优于静态模型的套期保值表现.样本外比较则显示,动态模型的套期保值效率明显劣于静态模型.当环境突变,模型存在明显的误设时,由于复杂模型较简单模型涉及到更多变量和假定,模型(误设)风险较简单模型更大.再加上复杂模型较简单模型包含更多噪音,估计风险更大,综合来看,复杂模型的总风险明显大于简单模型,这直接导致复杂模型的套期保值效率劣于简单模型.
本文從模型風險的角度,選取環境突變樣本區間,對“模型的複雜性和期貨套期保值效率”這一問題進行研究.採用隨機繫數馬爾科伕體製轉換(RCMRS)模型對期貨最優套期保值比進行估計.併將RCMRS模型套期保值效率和OLS、VAR、VECM及GARCH等模型進行比較和分析.樣本內比較髮現,複雜的動態模型併未帶來明顯優于靜態模型的套期保值錶現.樣本外比較則顯示,動態模型的套期保值效率明顯劣于靜態模型.噹環境突變,模型存在明顯的誤設時,由于複雜模型較簡單模型涉及到更多變量和假定,模型(誤設)風險較簡單模型更大.再加上複雜模型較簡單模型包含更多譟音,估計風險更大,綜閤來看,複雜模型的總風險明顯大于簡單模型,這直接導緻複雜模型的套期保值效率劣于簡單模型.
본문종모형풍험적각도,선취배경돌변양본구간,대“모형적복잡성화기화투기보치효솔”저일문제진행연구.채용수궤계수마이과부체제전환(RCMRS)모형대기화최우투기보치비진행고계.병장RCMRS모형투기보치효솔화OLS、VAR、VECM급GARCH등모형진행비교화분석.양본내비교발현,복잡적동태모형병미대래명현우우정태모형적투기보치표현.양본외비교칙현시,동태모형적투기보치효솔명현렬우정태모형.당배경돌변,모형존재명현적오설시,유우복잡모형교간단모형섭급도경다변량화가정,모형(오설)풍험교간단모형경대.재가상복잡모형교간단모형포함경다조음,고계풍험경대,종합래간,복잡모형적총풍험명현대우간단모형,저직접도치복잡모형적투기보치효솔렬우간단모형.
The debate on econometric models for estimating the minimum-variance futures hedge ratios has run for long time.According to the econometric theory,it is generally thought that,when the econometric test is " satisfactory",the complicated econometric models should bring the better hedging efficiency than the simple econometric models.But considerable empirical studies find that the hedging performances of the complex econometric models are not sure better than the simple econometric models.Therefore,the relationship between the complexity of the econometric model and the efficiency of futures hedging is worthy of further study.From the perspective of risk of econometric models,this article study the questions about "the complexity of econometric models and the efficiency of futures hedging" over environment shift sample period.A strand of the literature on futures hedge find the fact that the dynamic relationship between spot and futures returns may be characterized by market regime shifts.So,in this paper,we propose random coefficient Markov regime switching model to estimate optimal hedge ratio of china's copper futures market.The hedging performance of RCMRS model is compared against B-GARCH,VECM,VAR and OLS model using the minimum variance approaches over both an ex post and ex ante hedge period.The structure of this paper is as follows.The first part is introduction,the second part is the model specification of random coefficient Markov regime switching model,the next third part is the empirical analysis,including the statistical description of the sample data,the unit root test,the Cointegration test and the optimal hedging ratio estimation of RCMRS,B-GARCH,VECM,VAR and OLS model.The fourth part is the hedging efficiency comparison and analysis.Conclusions are given in the last section.The in-sample test finds that complex dynamic models,such as RCMRS and B-GARCH model,have offered no discernable improvement on the simple static models futures hedge.Compared to the in-sample period,the out-of-sample hedge efficiency decrease and the decline of dynamic models are greater than the static models.In the out-of-sample comparison of hedge performance of five models,the dynamic models are significantly outperformed by the static models.The risk of econometric models includes model-misspecification risk and estimation risk.When the relevant econometric test is "satisfactory",the model-misspecification risk of complex econometric model may be smaller than simple model.Complex model introduce more noise than simple model and its estimation risk is greater.Therefore,the total risk is uncertain and it is not sure that complex models can bring about the better hedging effect than the simple models.When in the presence of environment shifts,the model-misspecification risk of complex model will be greater than the simple model because it involves more variables and assumptions.So,the total risk of complex model is more than the simple model and the futures hedging pedormance of complex models will be outperformed by the simple models.