It is an interesting question in news trading whether activity prior to the news release can offer any clues to the upcoming reaction. This to some extent gets to the core idea of technical analysis that the chart sufficiently contains all inferences that may be drawn from the market (and it is task of the trader to find them). So one might assume that it's close to release; there are inferences even here. The issue, though, is that these are not hidden in the sense that other market inferences are, as the time frame and event have been made concrete. The problem for a run-up to a major news release is that this is a time of volatility, at least until very close to release (but not too close, as this can be a time of intense volatility).
However, this does indicate that there is information to be hidden in volatility. The problem with volatility is that it, by definition, does not contain clear information about direction; it is by definition volatile. One would need to strip away reactive elements related to pricing-in on expectations that have preceded the release. However, the focus is on the more immediate conditions prior to release rather than on the longer run-up, which may have seen a pricing-in trend.
There may be some inferences based on expectations of what will happen close to release, contained in the way the market will react to the emergence of order within this volatility. Now, whether this is an inherent market sensibility, that is, something related to the way the market functions, is another matter. This said, traders do ride the market; they don't necessarily try and impose order on it (in terms of direction), thus that tendency in itself could allow emergent behavior in the market.
The very volatility of the run up time can though allow for trends and other structure to appear, but the possibility of very volatile movement remain. This time period can in some senses be preferable to trading into the news, which is really a matter of a number of factors falling into place, which may not necessarily be the case (e.g., the required direction may happen, yet the move sharply reverses within a minute or so, or even less).
The more complex a set of reactions can be and the greater the range, the more likely it can be that while it may work out one or more times, this may not be the case at later times. This is why it can be preferable to trade on structure based events, such as market open and closes, even though these can simply highlight how hard it is to find a repeating pattern even when factors such as time are constrained.
The time before the news release can be looked at this way: it is a volatile time but constrained to some extent by potential outcomes that are close in time. The time of release is both tightly constrained by the event being known and also wide open; anything can happen and does, and the time after is for the market to settle back into its rhythm based on the news events, with volatility. As to why anything can happen at release may be related to the potential outcomes which were expressed in the run to release, meaning that there may be a point to looking for inferences even here.
