Question

我有我的投资组合的时间序列数据。我想计算回报。

                 T2      T12        T5 
2012-08-16 102779.9 102779.9 -102779.9 
2012-08-17 106050.8 106050.8 -106050.8 
2012-08-20 107883.3 107883.3 -107883.3 
2012-08-21 107024.5 107024.5 -107024.5 
2012-08-22 108630.7 108630.7 -108630.7 
2012-08-23 109067.9 109067.9 -109067.9

我使用公式

rets <- diff(portfolio, arithmetic = FALSE) -1

这一行给出了NULL错误

修改
输出（头（投资组合））

structure(c(102779.89722, 106050.751092, 107883.320688, 107024.464404, 
108630.748512, 109067.890932, -102779.89722, -106050.751092, 
-107883.320688, -107024.464404, -108630.748512, -109067.890932, 
102779.89722, 106050.751092, 107883.320688, 107024.464404, 108630.748512, 
109067.890932), .Dim = c(6L, 3L), .Dimnames = list(NULL, c("T2", 
"T5", "T12")), index = structure(c(15568, 15569, 15572, 15573, 
15574, 15575), class = "Date"), class = "zoo")
                 T2        T5      T12
2012-08-16 102779.9 -102779.9 102779.9
2012-08-17 106050.8 -106050.8 106050.8
2012-08-20 107883.3 -107883.3 107883.3
2012-08-21 107024.5 -107024.5 107024.5
2012-08-22 108630.7 -108630.7 108630.7
2012-08-23 109067.9 -109067.9 109067.9

Answer 1

> library(zoo)
> diff(portfolio)
                  T2         T5       T12
2012-08-17 3270.8539 -3270.8539 3270.8539
2012-08-20 1832.5696 -1832.5696 1832.5696
2012-08-21 -858.8563   858.8563 -858.8563
2012-08-22 1606.2841 -1606.2841 1606.2841
2012-08-23  437.1424  -437.1424  437.1424

似乎按照我的预期工作，给出逐列的连续差异。（arithmetic参数是什么？）

但是，对于（百分比）返回（我假设这些是股票报价 - 但为什么是负数？）你真的不想要连续的差异，但比率，像这样：

> rollapply(portfolio,width=2,FUN=function(xx)xx[2]/xx[1])-1
                     T2           T5          T12
2012-08-16  0.031823868  0.031823868  0.031823868
2012-08-17  0.017280119  0.017280119  0.017280119
2012-08-20 -0.007960974 -0.007960974 -0.007960974
2012-08-21  0.015008569  0.015008569  0.015008569
2012-08-22  0.004024113  0.004024113  0.004024113

查看?rollapply。

投资组合使用时间序列数据的差异返回

1 个答案: