• Visit the 2024 QuantNet ranking of the Best UK Quant MSc Programs.

  • C++ Programming for Financial Engineering
    Highly recommended by thousands of MFE students. Covers essential C++ topics with applications to financial engineering. Learn more Join!
    Python for Finance with Intro to Data Science
    Gain practical understanding of Python to read, understand, and write professional Python code for your first day on the job. Learn more Join!
    An Intuition-Based Options Primer for FE
    Ideal for entry level positions interviews and graduate studies, specializing in options trading arbitrage and options valuation models. Learn more Join!

Using Social Media to Make Investments

Ari

Ari

Joined
5/15/15
Messages
22
Points
13
Hey everyone,
Very curious if anyone is informed on how people/companies/entities are using social media to assist in analyzing the stock market and make investment decisions, or perhaps more generally the synthesis between big-data-social-media and finance?

E.g. off the top of my head:
Perhaps someone notices that when scanning facebook (a hard job I know, but this is hypothetical for now), the stock price of coca-cola rises shortly after there is a rise in the number of occurrences of the word 'coke' on facebook. Perhaps there is a delay between the vocal demand of the consumer on facebook and the reflection of that demand in the stock price. If one could notice that early and be a middle man that would be $weeet.

The above example isn't at all the only kind of thing I'm looking for. I feel like companies should be trying to incorporate social media into their investment strategies somehow. Anyone hear of anything interesting?

Ari
 
I did this in my job last year. was fun... In terms of implementation:

https://developers.facebook.com/docs/keyword_insights/
https://gnip.com/realtime/ (Twitter)

I found that other sources were largely a waste of time / money but I wasn't doing for finance industry specifically and data was sparse.

If you want to mess about with data without paying anything:
  • Use python-twython script to parse out historic twitter news items - with some tweaks you get pretty good number of JSON items and dodge the throttles
  • Automate the above over a period of time until some big news breaks
  • correlate to returns and see if it does anything interesting.
that's what I did anyway. Turns out it does do interesting stuff & weirdly sometimes seems that the market anticipates interesting events before they even occur by small time - does this indicate that HFTs have direct (faster) access to news data? probs.

So if you can't beat the market in pure reaction speed. Can you beat the market in accuracy of prediction for highly complex non-linear relationships - this motivated my post on deep machine learning.
 
Alex,
Thanks very much for your post. Can't wait to get a chance to dig through your links that you shared.
 
Thanks for sharing your links , Good article about social media marketing,i'm too much intrested in this.
 
Back
Top