|
| Backtesting and Strategy contest results do not match |
|
Guest
|
| Post subject: Backtesting and Strategy contest results do not match |
Post rating: 0
|
Posted: Thu 26 Jan, 2012, 01:34
|
|
User rating: -
|
|
I tried out testing strategies of top 10 rank holders of January on tick data, and on same time frame as used by the strategies, still the results are very different. Am I doing it incorrectly? What is the way to get same results as strategy contest shows.
|
|
|
|
|
 |
|
marss
|
| Post subject: Re: Backtesting and Strategy contest results do not match |
Post rating: 0
|
Posted: Tue 31 Jan, 2012, 14:06
|
|
User rating: 0
Joined: Tue 27 Dec, 2011, 11:27
Posts: 2
Location: Hungary, Budapest
|
|
Hi,
does not make sense to compare, indeed the contest
other API's platform will run as the demo API's platform...
marss
|
|
|
|
|
|
|
Contest Support
|
| Post subject: Re: Backtesting and Strategy contest results do not match |
Post rating: 0
|
Posted: Wed 01 Feb, 2012, 18:19
|
|
User rating: 8
Joined: Wed 21 Apr, 2010, 10:42
Posts: 1167
|
|
Please note that there is a difference between running the strategy in the historical tester and the demo platform. The amount of ticks is always different. The platform will always use the last tick to run the strategy while the historical tester runs through all ticks from the queue. It might happen in the situation when several ticks occur almost at the same time.
It also depends from your strategy's algorithm. Some strategies are very sensitive, some are not. However, please note that the difference must not be crucial. And if it really huge, it means something might be wrong with the strategy itself.
|
|
|
|
|
|
 
|
Pages: [
1
]
|
|
|
|
|
