Have you ever plugged non-snapshot data into your charting program to see the difference for the exact same time period for IB data and say Esignal (true tick-by-tick data)? BTW, very nice chart....I "get it"!
No not yet. I have a 'gut feeling' that IB data is a lot better than it is sometimes given credit for. I have compared some of my stuff to some of the charts on the MarketDelta web site, which presumably are not generated from IB data feed, and the correspondance, though not perfect was not bad at all. These comments apply to futures - I've seen some stuff that really doesn't look right for stocks on IB, but I havn't followed up at this time. I'm certainly interested in another data feed. I've 'half done' an adapter for opentick - http://www.opentick.com. Unfortunately it was very unreliable - certainly unusable for trading last time I looked. I really hope they get their act together as I like their approach. I also like the fact that they allow 1000 concurrent instruments including Level 2 for each of the 1000. It could be very interesting to look at the book for a large number of stocks. I'm running on Linux, so there's not too many options. I could write some sort of gateway/proxy to run on a windows box and interface with one of the feeds (probably DTN) but I'd rather not waste the time. It's an ugly solution and ugly and reliable don't often go together.
Over the last decade, I have run every data feed you could name. IB surprised me ---it is not bad at all. I have had my stock account at IB for the last year and use them for all kinds of trading these days. I have been pleasantly surprised. The data appears to have even gotten better. The big complaints were when the exchanges were giving out every tick. Now the exchanges (like CME) are not doing this anymore. It was impossible to keep up with them for the vendors. Eurex has done this since day one. None of this has ever effected my trading using IB. But lets remember that IB was started off of Timberhills information backbone. My office in CME was near the Timberhill people and they are first class all the way. They understand trading and data. Timberhill does not fool around. They make their money by trading.
As I understand it (and I may be wrong) Timberhill use the same infrastructure, including TWS, as IB customers. As you say Timberhill does not fool around. Common sense says that they are not going to put themself at a disadvantage with inadequate datafeed. For this reason, I always view the discussions of IB's feed with a little skepticism. One thing I would like though, is more than 3 concurrent L2 feeds.
Attached is a one minute chart of Kospi. Notice that the order book delta (ACV ratio) cannot get above 0.5 during the down move. Strangly enough 174.9 is about last week's low. K200 order book is a lot less flightly than DAX. Probably because it is a lot more liquid.
Simple answer - I don't know. I've spent all my effort at this stage at looking at stock index futures. My reflexive response is to say that I think it's going to be different because there are a lot of games going on in SIFs that aren't going on in the same way in individual stocks eg arbing, hedging etc. I also don't think there is an "ACV method" as such. I think it's a valuable tool, but you need S/R, market profile or whatever else you deem to be valuable. I know it's said that scalpers just trade off the book and order flow, but to me it just seems like folly to not avail oneself of all sources of information. Providing of course that it doesn't lead to "paralysis by analysis". However putting it all together in a coherent plan is not easy.
To illustrate the point about S/R, on the attached 5 min chart of kospi, we see K200 bang it's head against S1, LVA and VWAP on declining volume and divergent (falling) ACV. I really like the way K200 trades.