Finding the p-value for symmetric binomial distribution

Question

I have essentially found whether a buy or sell signal was generated from 2007-2013 for a stock. I have 1827 observations. I then found the return the signal generated for each day. I then found the number of buy and sell signals, and then found whether or not each signal was profitable for each given day. In an efficient market, the profitable buy and sell signals would be 0.5, i.e. only half would be profitable. Lets say I have 470/950, 970 buy signals and 470 of those profitable. In this specific instance, I would assume they would not be statistically insignificant, but how do I find the respective p-value? I.e. probability/p-value that 470/950 differs significantly from 0.5. Am I possibly meant to be using the chi-squared test instead?