I often read that institutional investors (mutual funds, banks, hedge funds, etc...) will build a position in a stock over time using relatively small orders so as not to move price and alert other market participants of their actions. In addition I also read of the use of "dark pools" by these institutions to hide their accumulation.
Is there a way to identify this accumulation with a greater than 50% accuracy using the price and volume data (end of day and/or intra-day)?
A couple ways, but its not a guarantee.
You have to have special charts. Instead of each tick being 1 min, 5 min, or whatever, it is a set number of trades. Say 2000. Since retail investors only buy and sell in small amounts, there will be small volume per tick. An institutional investor, however, would have a much much higher trade lot size, even if using an algo. Thus, large volume spikes in such a chart would signal institutional activity over retail. Similarly, daily charts showing average trade size can help you pick out when institutional activity is highest, as they have much larger trade sizes.
You could also learn how the algos work and look for evidence one is being used. ie every time price hits VWAP a large sell order goes through would indicate an institutional investor is selling, especially if it happens multiple times in a row.
You can have a pretty good guess by looking at price pattern and order flow (size of the trades) a) price should be traded in a range b) relatively large size orders, speed.
