求助：计算债券协方差时出现的问题

是 否 +2 论坛币

k人 参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群

赵安豆老师微信：zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

立即领取

感谢您参与论坛问题回答

经管之家送您两个论坛币！

+2 论坛币

小弟想算一个bond portfolio的covariance，结果发现几个bond的交易日期不同

例如，Bond A在2011年，7月只有1个交易日，而其他bond7月都是22个交易日，这在算协方差矩阵时，如何解决？是按0处理补齐，还是怎样？


小弟是初学者，请见谅
先谢谢各位了

扫码加我 拉你入群

请注明：姓名-公司-职位

以便审核进群资格，未注明则拒绝

分享0 收藏0 回帖

关键词：协方差 covariance Portfolio variance Portfoli 计算 债券

Several ways to deal with this problem.

1. ignore it :-)
2. find better data source
3. use liner or exponential interpolation (something like average) if missing data is a small part of the whole.
4. state space model.

for state space model this is very good but very complex technique. The general idea is that you assume the other bond's price is driven by the hidden state variable, the bond with missing data, you use the observed other bonds' data to filter out the unobserved bond price

Because you are a beginner it is a huge topic for the fourth method and doing it is another big project probably much larger than the whole project you are doing with. So I recommend you to follow 2 and 3