search

fja05680 / sp500

Current and Historical Lists of S&P 500 components since 1996
MIT License
439 stars 104 forks source link

Info is inaccurate. #9

Open FrostBid opened 2 years ago

FrostBid commented 2 years ago

I feel like the data is somewhat inaccurate for certain years. The csv file listed AAL as a component in 1996, but according to Wikipedia, the company is only added to the index in 2015.

fja05680 commented 2 years ago

Hi FrostBid,

1996 data is part of the original data from I got from "Trading Evolved". I have no good way to check if his data is 100% correct. In the description of this project, I suggested that one could use the data from 2001 onward because of the low symbol count for the first 5 years. I'll try to take a look at AAL this weekend. I could deal with this specific case, but generally, I have no way to verify the correctness of the original list. I assume it must be fairly accurate because this is what Andreas Clenow uses for his professional trading.

bxclib2 commented 1 year ago

this AAL has been confusing me a lot. I even ask ChatGPT whether a stock symbol could be reused if a company is delisted.

Btw, may i know do you know the meaning of the numbers in his original CSV? such as ETS-200603, what 200603 means? I am so confused...

fja05680 commented 1 year ago

I believe that indicts when the symbol gets removed from the index.

bxclib2 commented 1 year ago

In fact this is same as what I initially thought but I ´ve checked for some symbols it is not correct so this is a wrong guess

fja05680 commented 1 year ago

Ok, thanks for letting me know. I filter those out as you know. I would suggest just selling when assemble disappears that you have a long position in. If the symbol has been removed from the index, then the position needs to be closed. If it was a mere symbol change, and the company is still in the index, you should still sell. Your algorithm should immediately buy it with a new symbol name.

On Thu, Sep 21, 2023, 2:50 AM bxclib2 @.***> wrote:

In fact this is same as what I initially thought but I ´ve checked for some symbols it is not correct so this is a wrong guess

— Reply to this email directly, view it on GitHub https://github.com/fja05680/sp500/issues/9#issuecomment-1728955356, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACD3KHWKQ2IVBZWIXPP7LFLX3PPUVANCNFSM54RJJJ6A . You are receiving this because you commented.Message ID: @.***>

swisstackle commented 6 months ago

@fja05680 First of all, thanks for your work. Did I understand it correctly that you filter out symbols that have weird names such as ETS-200603?

JohnSparks2024 commented 6 months ago

The 200603 is the last date for which that symbol is valid for that company. It is how Norgate deals with symbol changes (whether due to merger, delist or simple ticker change). This is separate from whether or not the stock is still a member of the S&P 500. Norgate maintains that information in a separate table.

fja05680 commented 6 months ago

@swisstackle Yes, they get filtered out. And John is correct.

NinoBach commented 1 month ago

Im looking at Reuters Refinitiv data and there is a similar problem. American Airlines has the RIC AAl, and before that, Aon Grp had the RIC AAl. To be fair, they have different suffixes, but it's still confusing, before previously I thought I can derive the ticker from the RIC by deleting the suffix, but that won't work.