Is anyone using the SOBI method as described in this paper? Basically, it is comparing VWAPs of both the buy and sell order books with the last price to see whether the buy-side support or sell-side pressure is larger. I am using eSignal but I can't get the values of the entire order book, only the highest bid and lowest ask are available...
To some extent, OBI and their derivatives are being used. Probably mostly because the algorithms are so simple to implement (heh). I have several variants, similar to WMA, EMA, you can create different order book analytics. Purely my own opinion, but OBI based trades are not good, but that's from experience (and the markets I am involved it). OBI is interesting as a guideline, if OBI is severely against the direction of the order, then filter out the order. Depend on what book you look at, there are different characteristics, NASDAQ shows "age" as well, where as most derivative markets don't. So this creates opportunity for additional SOBI-like analytics, Survival model with Age / Depth / position. Only problem with those more complex analytics is that the computation required gets out of hand very quickly. Note: I use OBI (order book imbalance) to denote SOBI-like family of analytics, SOBI (standard order book imbalance) is the simplest form.