Look at where SIMO share have been in the last 2 months, it is clear that institutions are selling during a market expansion period. Their purchase power is needed for SIMO to go up. Absent of it, we have the current low share price.
They did it because their analysts learned, from tracking QCOM and carriers, that SIMO would not be in Samsung's LTE baseband this year. A couple of years ago, the same analysts thought SIMO was the best alternative to play the smartphone trend, but changed their mind in February 2012 after QCOM released its then new solutions. The analysts claimed that the opportunity was shut forever. They were wrong as SIMO captured design wins and chieved record revenue. But it kept share price under pressured for most of 2012.
They did it again this year, and it appears they are correct for now. One of the analyst even asked SIMO whether to consider divesting the transceiver division since most of the analysts had taken it off in their valuations. Now this is key. This is why there are still buyers of this stock.
They are wrong again because they do not understand the strategic necessity for Samsung to diversify supply. Samsung can only dominate if it can supply itself all critical components. Baseband (with the transceivers) is a critical component because QCOM is the only true game in town.
More importantly, they forgot why they liked SIMO in the first place. It has the highest margin in a thin margin memory industry and now dominates mobile memory controller with strategic partnerships with Samsung, Hynix, and adding one more. It is also getting into SSD controller, a potentially huge market with PC, servers, etc.