Thanks for your help. Here is the project statement
- A simple approach to measure the effect of an event is to identify the size of its reaction. This can be done by comparing the volatility of the current day with the average volatility of the recent past, using the true range calculation:
Programmatically calculate the historical volatility ratios of the preferred market indices and identify the relative effect of the chosen news event in statistical context
- Determine the optimal values of Volatility Ratio that has greatest predictive power in back-tests
- Device a simple trading strategy that trades any of the broad market indices based on buy-sell signals generated from Volatility as an event driven indicator
---------------------------------------------
My questions, please help me find the answers
1. Where to find the economic data release dates for python access. I tried quandl but the dates are just 1st or last dates of a month and not the actual data release dates.
2. once I have the actual news dates, how can I statistically calculate the effect of news on the vol ratio during backtest? shall I calculate how vol ratio jumps up on the news release date?
3. what kind of trading strategy can we use for this vol ratio on new release date? for momentum, shall I use the return over past 2 days (of the news release date) and go long or short based on that? When do I unwind this position?
4. for the trading strategy, shall I use combination of miving average and volratio? I dont know how this can be used. please share some ideas. This is more for my brainstorming and actually showing that I tried different ideas
thanks a lot for the help