Using the books of retailers to trade against the herd

[kmgilroy89]

Has anybody built a system where the main strategy is looking in the books of retail firms to see how many people were long and short, and then going against that position? The idea is if, lets say the book is 80% long and 20% short, if things go bad, and they need to sell all the selling will drive the stock down harder.

 

[Kevin Schmit]

Quote from kmgilroy89:

Has anybody built a system where the main strategy is looking in the books of retail firms to see how many people were long and short, and then going against that position?

I looked into that strat using Oanda data, and found nothing useful,
at least not at the time-horizon I was looking at.

Oanda is largely a retail firm, and they put their aggregate customer positions online per pair in close to real time. Nothing predictive that I could find in either levels nor innovations in the web-scraped series.

This was forex, not equities. If you know of a retail equities brokerage that puts its aggregate customer book online in a scrapable or downloadable format, let me know and I will have another look.

 

[1a2b3cppp]

I buy S&P500 when it is going down. I buy more in big downtrends.

That's kind of doing what you're saying, right?