Quantopian's community platform is shutting down. Please read this post for more information and download your code.
Back to Community
TOO GOOD TO BE TRUE???

Sentiment analysis based predictions

2 responses

Two things I noticed:

set_slippage(slippage.FixedSlippage(spread=0))  

That doesn't tend to lead to realistic backtests unless you plan on trading on Robinhood.

context.security_list = symbols('AAPL', 'GOOG', 'BA', 'MRK', 'WMT', 'INTC', 'PG', 'JPM')  

By selecting these stocks you're exposing yourself to some look-back bias, although all these companies have had a good time recently, you didn't know back in 2016 that that would be the case, I would suggest picking stocks from a universe, or perhaps randomly selecting a few.

@Sanna
I cloned your algo and run it. But the RETURN is ZERO. No any transaction happens.