You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an alternative browser.
You should upgrade or use an alternative browser.
Backtesting Software & Autotrade
- Thread starter whoknows?
- Start date
Quote from Traveler:
Is Tradestation off the list due to lack of portfolio testing?
Yes. Also, I do not want to use them for execution so at $2,400 per year it's not cheap. They do have great charts and data, but backtesting is slow and in my opinion to in the league of some of the other vendors being discussed. Just my opinion.
Quote from whoknows?:
Yes. Also, I do not want to use them for execution so at $2,400 per year it's not cheap. They do have great charts and data, but backtesting is slow and in my opinion to in the league of some of the other vendors being discussed. Just my opinion.
Sorry for spell error..."not" in the league...
Quote from whoknows?:
Yes. Also, I do not want to use them for execution so at $2,400 per year it's not cheap. They do have great charts and data, but backtesting is slow and in my opinion to in the league of some of the other vendors being discussed. Just my opinion.
Anyone using Multicharts??? Not sure of where they stand on portfolio backtesting
Quote from taowave:
Anyone using Multicharts??? Not sure of where they stand on portfolio backtesting
I tend to agree about TradeStation - although I found a nice add-on that allows portfolio-level testing.
http://www.tsresearch.com/software/...view/portfolio/
It does require some work to integrate with an existing strategies but it is an option for those of you who want to use TS - and it is very affordable. Definitely worth checking out.
I just downloaded the beta of Multicharts to try out their new backtesting features - very simplistic - not really worth talking about at this point.
Seer looks interesting - I'm going to check it out.
Took a look at Seer - definitely interesting for those looking for backtesting software. Very similar, in approach, to QuantDeveloper by Smartquant, with a little bit of TradersStudio thrown in. Definitely could get expensive at $50-$100 a month, but if you're looking to get into backtesting on the cheap, it might be a good option.
Also, it has no neural or genetic algorithms if that matters to anyone. It seems to have optimization but I haven't found it yet.
Also, it has no neural or genetic algorithms if that matters to anyone. It seems to have optimization but I haven't found it yet.
Quote from droskill:
Also, it has no neural or genetic algorithms if that matters to anyone. It seems to have optimization but I haven't found it yet.
Neural software is a great help in backtesting of a trade strategy from my point of view.
How does it work?
Neural network is formed on the same indicators set as the strategy that should be backtested. In that case neural network represents trade strategy that was formed based on trade historical data and should correlate with traderâs strategy. The question is how close they correlate.
Next step. Trader compares conditions when TS is generated with his strategy with conditions of neural network.
Neural software can provide trader with explanation why it generates TS. For example, it can show indicator combinations with their values that neural software recognizes as prerequisite for TS generation and show frequency of such situations that it faced during training.
Trader should decide if his strategy is more efficient and if no then he can change some of conditions to make it more sensible.
More about this approach is at http://avis-c.gluk.org where Artificial Foreteller is described.
Pathfinder, it appears that the majority of posts on ET are about this site/product.Quote from Pathfinder:
More about this approach is at http://avis-c.gluk.org where Artificial Foreteller is described.
What is your affliliation, if any, with it?
Yes, I took part in development of this approach. Before I developed software for strategy backtesting and found that neural network is much better do this work.