Wealth-lab versus TradeStation 2000i

A

andrasnm

Eventually I likely buy both... but why is the big difference in price and why is the
testing multiple contracts(symbols) a such a big plus in favor of Wealth-lab?
Can't you just run multiple backtesting
on each contracts when using TradeStation?
Which is easier to program?

What is the difference in TradeStation 2000i and TradeStation 7? I know both can be programmed using EasyLanguage
and TradeStation 7 is also a brokerage.
Did Omega just license the software to a different entity?
Also Wealth-lab 3.0 and TS2000i can also interface brokers so I fail to see why anyone would pay the hefty monthly charge TS 7.0 want...
Thanks...
 
Quote from andrasnm:

Eventually I likely buy both... but why is the big difference in price and why is the
testing multiple contracts(symbols) a such a big plus in favor of Wealth-lab?
Can't you just run multiple backtesting
on each contracts when using TradeStation?
Which is easier to program?

What is the difference in TradeStation 2000i and TradeStation 7? I know both can be programmed using EasyLanguage
and TradeStation 7 is also a brokerage.
Did Omega just license the software to a different entity?
Also Wealth-lab 3.0 and TS2000i can also interface brokers so I fail to see why anyone would pay the hefty monthly charge TS 7.0 want...
Thanks...
More...

Does wealth-lab or TS7 support intraday, tick-by-tick back testing on a portfolio of symbols over a large date range?
 
>>Does wealth-lab or TS7 support intraday, tick-by-tick back testing on a portfolio of symbols over a large date range?<<

Wealth-Lab supports intraday, bar-by-bar (not tick-by-tick) testing on a portfolio of symbols over a large date range. I'm not exactly sure what you mean by tick-by-tick backtesting, can you provide any more details?
 
Quote from Glitch:

>>Does wealth-lab or TS7 support intraday, tick-by-tick back testing on a portfolio of symbols over a large date range?<<

Wealth-Lab supports intraday, bar-by-bar (not tick-by-tick) testing on a portfolio of symbols over a large date range. I'm not exactly sure what you mean by tick-by-tick backtesting, can you provide any more details?
More...

That's where you don't wait until a bar is over to generate an order. For example, if you have the following prices over the course of a minute:
50.25, 50.24, 50.19, 50.20, 50.18, 50.10, 50.05, 49.95, 49.96, 50.10, 50.15, 50.20, 50.25, 50.35

And you had a sell at 50.20 followed with a buy at 50.00, then how would a system using one-minute bars handle that not knowing all the ticks that occurred in-between?
 
Thanks for clarifying. We currently don't have plans to implement tick by tick backtesting, there are many other things we're currently working on. If there is a doubt, the system uses the most pessimistic possiblity for back-testing results.

But this brings up a question. I've heard repeatedly that TS does not have tick by tick back-testing, but a few years ago I remember a full page add for TS that indicated that the product did indeed have this feature. The ad showed a price chart with 2 smaller charts zoomed into a single bar. It said something like, "Did prices in the bar move this way, or that way."

I'm curious about this now, anyone have a clear answer?
 
Quote from Glitch:

Thanks for clarifying. We currently don't have plans to implement tick by tick backtesting, there are many other things we're currently working on. If there is a doubt, the system uses the most pessimistic possiblity for back-testing results.
More...

That's a shame because I want to test some strategies that are TRULY tick dependent and while most pessimistic is good, it would obscure the results and invalidate any testing.


But this brings up a question. I've heard repeatedly that TS does not have tick by tick back-testing, but a few years ago I remember a full page add for TS that indicated that the product did indeed have this feature. The ad showed a price chart with 2 smaller charts zoomed into a single bar. It said something like, "Did prices in the bar move this way, or that way."

I'm curious about this now, anyone have a clear answer? [/QUOTE]

TS has a mode (TS2000) where you can test at the "tick level". This obviously takes much longer to run, but for some strategies, can be very useful.
 
The most important thing to be able to test against is the B/A. An even more sophisticated program would look at the size at each as well as the trades that occured around it and it's B/A and would make an even more accurate guesstimate as to the probability of a fill.

I know of no retail software that goes throught this amount of pain to assess accuracy, and therefore, are nearly worthless for this kind of testing. In some sense, the "problem" stems from a kind of Hisenberg Uncertainty Pricicple of high frequency data.

nitro
 
I believe the problem is that historical data for this is hard to come by. If the historical data were available it would be possible to use this in back-testing.
 
Wealth-Lab Version 2.1 (current release) can optimize stops. Version 3.0 (coming this summer) can also optimize position size, across a portfolio optimization.
 
You must log in or register to reply here.
Back
Top