MATLAB金融工具箱的应用实验报告

MATLAB金融工具箱的应用实验报告该数据选取的是TCL自2004年1月7日起至今的股票收盘价数据：1.时间序列分析x=xlsread('tcl');tcl(1,1)ans=731979tcl(1,2)ans=8.3500plot(tcl(:,1),tcl(:,2));datetick('x',23);xlabel('Date');title('TCLComposite');2.绘制收益率序列Return=price2ret(tcl(:,2));Time=tcl(:,1);t=Time(2:end);plot(t,Return);datetick('x',23);3.自相关autocorr(Return)title('ACF');4.偏自相关parcorr(Return);title('PACF');5.AR模型m=ar(Return,2)m=Discrete-timeARmodel:A(z)y(t)=e(t)A(z)=1-0.03316z^-1+0.01952z^-2Sampletime:1secondsParameterization:Polynomialorders:na=2Numberoffreecoefficients:2Usepolydata,getpvec,getcovforparametersandtheiruncertainties.Status:EstimatedusingAR('fb/now')onReturn.Fittoestimationdata:0.08195%FPE:0.000963,MSE:0.00096146.ARMA模型formatlonggfora=1:10forc=1:10m=armax(Return,'na',a,'nc',c);FPE(a,c)=fpe(m);FPE(a,c)=fpe(m);endendnmin=min(FPE(:))[na,nc]=find(FPE==nmin)ormatlonggfora=1:10nmin=0.000962185144313305na=1nc=1结果分析：ARMA(1,1)为最合适的拟合模型。