. If you try predict it, it's double prediction, first you have to predict the data, then you have to predict the markets reaction to the data. Very few have the insight for that.
You don't have to predict anything bud.
I trade news spikes as one of my methods using my deviation figures which I have accumulated over 4 years.
I have traded news events for 3 of those 4 years. I have collected data over the past 4 years for the GBP- US-NZD-AUD-CAD over 48 news releases that print every month. With this data my records show that 33 of these have devaiton/forecast triggers, these triggers are figures that the release must distance its self away from forecast, I have also averaged the amount of pips each news event moves price based on the deviation if hit! I use a live news feed 'trade the news' that gives me the release figure as it is printed, no lag. I already have the tradable triggers in place (deviation number) from forecast. If the release prints out side the forecast range its a straight BUY/SELL depending on which side of the range.
Take the NZD retail sales that printed just 20 mins ago. The forecast was 1.1 the devaiton I need to trade this is 0.5 this is based on 4 years of reasearch, the release printed 1.6 giving me the figure I needed to trade. I took 28 pips in around 9 seconds.
I didnt need to predict anything, all I needed to do is wait to see if the release printed outside the forecast range and then bu/sell depending on what side the surprise factor hit, in this case it was positive (buy).
If the release printed inside the forecast range then no trade. Nothing gained nothing lost.