我有一个令人费解的问题:
我有一个时间序列数据,我试图通过拟合gjr模型来进行iid。
首先,我尝试使用Econometric Modeler应用程序并为残留样品自相关函数获取以下图形:
但是,当我尝试自行复制此图时,会得到完全不同的东西。为什么会这样?
function GJR_Returns33 = modelTimeSeries(Returns)
% GJR_Returns33 = modelTimeSeries(Returns)
% returns an estimated model. This code recreates the estimated model
% produced in the Econometric Modeler app. Use the code below to estimate
% the same model, or estimate a model with the same parameters on a new set
% of data.
%
% Input:
% Returns: A numeric matrix with the same number of columns as the data imported into the app
%
% Output:
% GJR_Returns33: The model containing estimated parameters
%
% Auto-generated by MATLAB Version 9.4 (R2018a) and Econometrics Toolbox Version 5.0 on 27-Jul-2018 21:05:12
Returns3 = Returns(:,3);
%% Estimate a GJR Model of time series Returns3
GJR_Returns33 = gjr('Offset',NaN,'GARCHLags',1:2,'ARCHLags',1:2,'LeverageLags',1:2,'Distribution','t');
GJR_Returns33 = estimate(GJR_Returns33,Returns3,'Display','off');
[residuals2, variances2] = infer(GJR_Returns33, Returns3 );
figure(1)
autocorr(residuals2.^2)
title('Sample ACF of Squared Standardized Residuals')
figure(2)
autocorr(residuals2)
title('Sample ACF of Standardized Residuals')
end
我的代码有什么问题? 预先感谢