如何在HP滤波法中选择最优的lambda

目前在做一个HP滤波分析，教材上对HP滤波中lambda的选择，月度数据通常用14400.但是看一些国外的文献中采用的是129600.想请问如何在月度的时间序列中选择最优的lambda?求程序。谢谢！
clear
data=xlsread('D:\china.xlsx');

lambda = 100:10:100000;
N = length(lambda);
TS = zeros(length(data),N);
CS = TS;
for i = 1:N
    [TS(:,i),CS(:,i)] = hpfilter(data(:,2),lambda(i));
end
lambda_s = (std(diff(CS))./std(diff(TS))).^2;
[lambda_opt,maxi] = max(lambda./lambda_s);
[T,C] = hpfilter(data,lambda(maxi));



这样做下来，lambda_opt=87.0130，问题在哪里？求解答。谢谢！

lambda_s 是lambda./lambda_s的最大值，不是lambda的最大值，应改为

1. clear
   
2. data=xlsread('D:\china.xlsx');
   
3. 
   
4. lambda = 100:10:100000;
   
5. N = length(lambda);
   
6. TS = zeros(length(data),N);
   
7. CS = TS;
   
8. for i = 1:N
   
9.     [TS(:,i),CS(:,i)] = hpfilter(data(:,2),lambda(i));
   
10. end
    
11. lambda_s = (std(diff(CS))./std(diff(TS))).^2;
    
12. [lambda_opt,maxi] = max(lambda./lambda_s);
    
13. lambda_opt = lambda[maxi]
    
14. [T,C] = hpfilter(data,lambda(maxi));

复制代码

但是不必搞个优化值，follow literature就行，对月度数据取129600

并且，lambda设定的是100——100000之间，为什么得到的是87，不在这个范围。

λ=129600

Hodrick and Prescott suggest 1600 as a value for λ for quarterly data. Ravn and Uhlig (2002) state that λ should vary by the fourth power of the frequency observation ratio; thus, λ should equal 6.25 for annual data and 129,600 for monthly data

Ravn, Morten; Uhlig, Harald (2002). "On adjusting the Hodrick–Prescott filter for the frequency of observations". The Review of Economics and Statistics. 84 (2): 37.