Large funds cannot move without leaving heavy footprints. One good relevant trace that we can find is in volume analysis, since we can assume that professional players need to place larger orders than retail players.
Even if hedge funds account for the lion's share of trading on the NYSE, we could still say that the majority of the trades (mainly small trades) originate from retail players. The simple difference is that hedge funds place larger orders than retail players (or a succession of mid-size orders). The consequence is that retail players make everything rather foggy when you look at all the data. What we therefore need is a reliable method to filter out the noise generated by retail players; we can then focus only on large players.
Using the Effective Volume Flow analysis model and separating the minute-by-minute volume by size allows filtering out such a noise. The filtering calculation is shown on Table 5 (separation size of 10,000 shares). The plotting of the LVS_VF (Large Volume Size Volume Flow) and the SVS_VF (Small Volume Size Volume Flow) results allows separating small from large players.
The separation of the effective volume flow analysis by volume size allows us to study possible divergences between price and volume patterns.
This method can be used in several instances:
A flat trading range will probably break in the direction of the LEVF.
A price uptrend with a negative LEVF indicates that problems lie ahead.
A price downtrend with a positive LEVF indicates that the downtrend is not sustainable.
Check page 34 (50 via the pdf reader) of the book Value In Time that I sent you.
According to Pascal, the best and most obvious way of to separate the volume between large and small is using the average separation method, as explained there.