It can be noted in news trading that market reactions can be based on the extent to which the news is a surprise. But what constitutes a surprise. The data generally needs to be significantly different from what is expected, for example higher or lower then the consensus estimate, or of a different direction from what is expected. This will typically generate a reaction, because of this: this kind of deviation is seen as an effect which will generate a reaction in the market. In effect it is a pattern to be traded on.
Given the pair it can also be seen which way the news will tend to affect the pair. Depending on the pair, this can be more straightforward in Non Farm Payrolls, but more complicated in FOMC. However this can also be more complicated in NFP, if a good jobs number could be seen as indicating that interest rates may rise. However there may be an argument that a strong surprise in either market will have a significant reaction which can delay the complications (even if only briefly).
Arguably, one could see the pause which typically happens after a strong news reaction as one which reflects the extent to which more complex outcomes may be possible, perhaps resulting in a reverse, rather than a continuation. The thing about really unexpected news events, is that they present such a clear path to valuation that all complications are washed away. But more subtle news reactions may not do this, resulting in complex moves, mirroring this. A strong reaction can push a pair into a new valuation range, where it can work out and through the nuances of the news, away from the news event.
The problem for trading is that these nuances can be part of the news reaction, resulting is oscillations. So the take out from this is that a strong surprise can provide a directional move, but the capacity for complexity to emerge is there including during the initial reaction, rather than simply being priced back in after the event (resulting in a more gradual move back towards the starting point).
