 |
umair
2013-03-24 19:29:53
|
|
hi, i want the see the performance of my logic for all the symbols that I have added. i have already done individual testing via your SymbolIndex script posted on http://www.quantshare.com/sa-502-how-to-backtest-each-stock-or-asset-individually
\ The problem with that script is that if I'm testing 900 symbols and it's 1 min data, the backtest can take hours to finish.
Alternatively, I thought that If I do a regular simulation WITHOUT using the SymbolIndex script and set my max positions to 1000 and capital to 1000000000, theoretically, the system should be able to execute every single signal. But it does not. Any idea why? The alternative method would be much quicker
|
|
|
|
|
umair
2013-03-24 19:43:33
0
|
|
how to ''theoretically'' to get results of individual securities without using the symbolindex script
set positions to 1000 if your symbol count is 900
set you capital to a ridiculously large number, 1000000000
export the data to excel
create a pivot table for the symbols
then you can play around with the pivot table to get individual securities results
BUT the above simulation seems to be not working.. any idea why??
|
|
|
|
|
umair
2013-03-24 19:49:50
0
|
|
wait nvm, it does work!!!!!!
conclusion of the story: instead of using the symbolindex script (cuz simulation literally took me 11 hours to complete) use the above method. the above method took my 10 MINUTES....... you can export your data to excel and play around with all your signals
QS can you please confirm / give your thoughts? is the above method same as testing via symbolindex and will the signals be the same/
|
|
|
 |
clonex
2013-03-24 19:59:32
0
|
|
Best Answer
It would be good to have directly in report performance per symbol. And Options for performance per day/hours( idea for intra day traders)
|
|
|
 |
QuantShare
2013-03-25 10:23:07
0
|
|
Yes, you can do that. You can also compare one or two results of using the pivot table versus the "symbolindex" method.
|
|
|
|
|
0
