Huh? What's a historical dbase worth, a few thousand dollars?
looks like IBKR looking to acquire Tradestation
- Thread starter billyjoerob
- Start date
Huh? What's a historical dbase worth, a few thousand dollars?
If IB + TS == higher rates, on the theory they're providing added value by integrating the TS product, that'll rain on a lot of Etraders' parades.
Most charting software licenses are way cheaper... CQG comes to mind as "expensive" charting... but then it's expensive in part exactly because it comes with data.
Naturally, if you are prepared to buy data and charting platform separately (from the broker account and each other), the question of advantages of Tradestation is mute. There are cheaper brokers, better charting, better backtesting platforms and better data. They just don't come in one inexpensive package and don't work as seemlessly together.
Another issue is most providers of "live" datafeed don't offer long enough history of tick data. So, one ends up with 2 separare data sources: one for long history and one for recent history and live data. It is a lot of work to set up tick filters etc so that indicators and trading systems produce sufficiently close results on these different sets of data.
IB is not a good example of perfect datafeed. However, just to give an idea of what can happen when you don't have history form live datafeed... Normally IB data is filtered. However, this summer IB had a spike in prices of a number of futures. It affected even hourly bars in IB backfill. Such a spike hadn't appeared for a good number of months before. Without a history of "bad" dat one can't properly build a filter to take care of it.
Further, historical data would typically require conversion before it can be used by a charting/backtesting platform. Maintaining up-to-date history on all symbols of interest becomes a separate project etc...
Note, all of this vast amount of work and money adds nothing to your bottom line. It's just required to get started with what Tradestation does out of the box.
If you are just simulating fills at price levels, wouldn't any type of filtering of the data cause unrealistic market interaction? I just want my bars based on actual fill prices.
I was talking of filtering the trade (and bid/ask) ticks that don't correspond to any actual trades. Some of these appear as consequence of an attempt to provide most current data as fast as possible (aiming HFT algorithms). So, a trade proce may be sent down the data stream before the trade size is claculated in the order matching engine etc.
"Bad" ticks inevitably come from exchanges... but most data providers offer "filtered" data where such ticks are euristically identified as "bad" and removed. Because such ticks are removed on the "best guess" basis, naturaly, occasional valid ticks are removed too.
Removal of valid ticks, which matched to a "bad tick" pattern, is not not as bad a thing as it may seem. What occasionally happens is a few levels of bid or ask orders may be pulled simultaneously and then restored. As a result, a very small amount gets executed at a proice which is effectively far from the current "market" price. Any assumption that an order would be executed at such a price is unrealistic.
I consider Peterffy to be about the best in his line of business As such, he should know the potential and value of his purchase of a 6.3% stake in TRAD. I think the 6.3% purchase will work out OK for him. Peterffy probably could get Tradestation back on track if he took total ownership. It will be interesting to see what may develop with this.