PortfoliosLab logoPortfoliosLab logo

Incorrect Correlation Table Calculations?

KR
Karl Restall31 мая 25 г. | Опубликовано в Общая

Hi! Could be me...but I think the correlation table calculations are incorrect for at least some of my portfolios. For example, I have a portfolio that is 70% SPY (S&P 500) and 30% BIL (T-Bills) and the correlation table is showing me a portfolio-to-market correlation of 1.00. Given that almost a third of the portolio is BIL (with a -0.11 correlation to the market), I can't see how the portfolio's correlation to the market could be 1.00.

I was going to send a screenshot showing all the parameters but I can't see any way to upload it for you to look at. Anyway...thoughts?

I'M WRONG...IT DOES APPEAR TO BE CORRECT. MY MISUNDERSTANDING. WISH I COULD REMOVE THIS...(smile).

1 комментарий
1 ответ

Сортировать по

DS
Dmitry Shevchenko04 июня 25 г.

I just took a look and agree, it does seem pretty counterintuitive that a portfolio of stocks and bonds would show such a strong correlation with the stock market. So I wanted to share a few thoughts that might be helpful for others who come across the same question.

First of all, a correlation of 1.0 doesn't mean the portfolio's returns are equal to those of SPY, it simply means the returns tend to move in the same direction as the stock market, though to a lesser degree. Since BIL's daily price movements are usually minimal, they aren't enough to offset the large swings seen in SPY. That means even with a sizable bond allocation, like 30%, the portfolio's overall volatility may be reduced, but its directional movement will still largely mirror the stock market. If the market goes down, the portfolio will likely go down too, and vice versa.

I experimented with different allocations, and it looks like only portfolios with more than 90-95% in bonds begin to noticeably reduce correlation with the stock market.


KR
Karl Restall

-10 мес.

Awesome explanation, Dmitry. I had BETA in my head for some reason when I wrote the post - confused it with correlation. Everything is as you say. I just couldn't remove the post, so I edited it and added the "I'm wrong..." sentence. My fault. But, thanks for posting an explanation as it helps!


Категория
Общая

Просмотры
240