In retrospect, it was possible to see in advance there was something happening with Twitter. The stock was weak and getting weaker, so it shouldn't have attracted buyer interest, and judging by price alone, no one was interested. But things were happening "behind the scenes".
Here is an annotated chart showing Wyckoff-style accumulation. The main principle of Wyckoff analysis is, volume (meaning accumulation) precedes price.
I use Elder's Force index with custom parameters to uncover accumulation. It's similar to OBV (on balance volume) but does a better job of separating buy volume from sell volume in each bar.
I don't use standard Wyckoff terminology. It can be helpful but I don't think it's totally necessary for understanding what's going on.
1 - top panel shows Communications sector (XLC) RS (relative strength) weakening vs SPX since October,
so most investors would not be looking here
2 - second panel shows Internet industry losing RS vs its sector from December
3 - third panel shows TWTR losing RS vs its sector from November
4, 5 - Nov 29 huge volume bar totally fails, Force (a refined version of OBV) makes a new low
and price continues lower
6 - From Nov 2 forward, Force begins to IMPROVE as price falls, crossing its MA.
Only professionals consistently buy falling prices, so somebody is interested in this stock
7 - Jan 24 - in a down trend, big volume+new low+UP close shows strong buying. The pros have
tipped their hand. Several days of me-too buying follow.
8 - Feb 10 - huge volume+new leg high+DN close. Big volume in an up leg should close UP.
But the pros prevent price from taking off too soon (for them)
by heavily shorting the stock.The me-toos take this for overhead supply and gradually get out.
9 - By now, Force MA has turned up, Force stays above its MA. So accumulation is well established
10 - TWTR RS has also improved against its industry - its showing strength by not falling with its peers
11 - Sellers try to make a new low, but pros prevent it over next two weeks.
Mar 14-21 Three strong volume up bars show accumulation is over and mark up about to begin.
The chart ends before the big run up so the scales are right to make things visible.
I don't have a dog in this hunt but marked up a PnF chart looking for Dynamic CSS Patterns below.
I can see the advantage of the PnF approach being more objective. Although the TWTR bar chart is pretty clear, many others aren't - for instance, not all test the price lows, not all stay above the Force MA, etc. Of those, some work out and some don't. So what are the essential characteristics? Not always easy to tell. On PnF, on the other hand, three sets of circles plus a column of 7 x's is objective.
PnF charts can take a bit of getting used to but they are very clear and objective. And, if you can "see" the high and low of a column, or group of columns as being somewhat of a Price Channel, it can help a bit. The numbers and letters designate the month at the time the box printed.
A nice feature on PnF charts, and other noiseless charts like Renko, is that you can construct them using alternate price and volatility scales. 1 point box too big for you? Change to a 1/2 point. Or lower the reversal from 3 box (default) to 2 or 1. I prefer to use the % scale on the PnF chart for clarification. High/Low vs Close Only prices are also often useful to review.