Started Algo Trading - Daily Paper Trade updates

#41
Hi Cubt,
Since your algo is already coded, why don't you back test on atleast 10 years data and get the results fast. Global data feeds provided tick by tick data for all nse stocks for the past 10 years . I dont know the price , but since you have already spent so much , you can do that and know your results faster than paper trading
 

onlinegtrash

Well-Known Member
#42

they use simple excel file which receives RT data while it runs in background:thumb:incidentally they have one system which trades directly from excel...
you guys have the proof already with you with below quote
so they trade from Excel RT data!

aaah...I thought algo trading involves:
Fix protocol or streaming sockets (like Interactive brokers have)...
real time complex event processing (such as ESPER etc)...
and databases to handle tickdata and derive candles for different time window... etc.
That's why I have seen algo traders and quants with awe !

hmmm... anyways excel or high tech... its good as long it makes money!
 

bpr

Well-Known Member
#43

they use simple excel file which receives RT data while it runs in background:thumb:incidentally they have one system which trades directly from excel...
you guys have the proof already with you with below quote
My statement which you are quoting is an assumption and not to be taken as a fact. I am trying to understand the setup based on the inputs given by Cubt.
 
Last edited:
#44
Hi Cubt,
Since your algo is already coded, why don't you back test on atleast 10 years data and get the results fast. Global data feeds provided tick by tick data for all nse stocks for the past 10 years . I dont know the price , but since you have already spent so much , you can do that and know your results faster than paper trading
Buddy, I think he is paper trading to get the feel of the algo, get familiar with the platform and to see if there are any glitches which need to be sorted.
He is not doing paper trading to validate the strategy and to get any sort of results.

Hope u got it.
 

TradeOptions

Well-Known Member
#45
is there a possibility of symphony also being able to use ur algo for their trading without ur knowledge ?
maybe this area can be explored more, i dunno much about it but if they use same strategy then you 2 will be directly competing for both entry and exits, it can have significant impact on overall profit..
Vertigo, I too think that it is an important point to consider.
 

Cubt

Algo Trader
#46

they use simple excel file which receives RT data while it runs in background:thumb:incidentally they have one system which trades directly from excel...
you guys have the proof already with you with below quote
They have invested huge amount of money for Real time data, its not something like they get the feed from yahoo or google and put them into their server.

They have their own database, multiple servers. In fact, their simulation environment which is used for paper trading matches 99% with live environment.

Find the below the screenshot of NEST Terminal live quotes and Symphony's Paper Trading environment.



Dude, I wouldn't have spent so much money on developing an algo by signing up with Symphony, if the company was as cheap as you explained..
 

Cubt

Algo Trader
#47
Hi Cubt,
Since your algo is already coded, why don't you back test on atleast 10 years data and get the results fast. Global data feeds provided tick by tick data for all nse stocks for the past 10 years . I dont know the price , but since you have already spent so much , you can do that and know your results faster than paper trading
Am doing paper trading just to check the execution part, how fast the orders are getting placed and getting filled. Checking if the algo is working as per the requirement etc. Without backtesting the strategy no one would try to develop an algo.
 

Cubt

Algo Trader
#48
Buddy, I think he is paper trading to get the feel of the algo, get familiar with the platform and to see if there are any glitches which need to be sorted.
He is not doing paper trading to validate the strategy and to get any sort of results.

Hope u got it.
Exactly!! :thumb:

Did you face any issues with your execution? Is the orders getting filled at correct price? What about slippages?

On testing their algo, I faced the below problems.


  1. Orders are not getting filled in if there is a sudden abnormal spike in prices. I have given my requirement as Stop Loss - Market Buy/Sell orders, but they informed me due to NSE approval issues, they will not be able to use SL-M orders, instead they use SL-Limit orders with higher bid price for Long and Lower ask price for Short. So that it would get filled in immediately. But still I have seen few orders not getting filled in, development team is working on it now to fix it

  • Open orders and Positions are not getting cancelled/squared off at the specified end time. Though MIS orders will be squared off by brokers automatically, this is not a high priority issue, but still need to be fixed.

  • Their EOD reports are not user friendly, difficult to calculate profit/loss details.
 
#49
Exactly!! :thumb:

Did you face any issues with your execution? Is the orders getting filled at correct price? What about slippages?

On testing their algo, I faced the below problems.


  1. Orders are not getting filled in if there is a sudden abnormal spike in prices. I have given my requirement as Stop Loss - Market Buy/Sell orders, but they informed me due to NSE approval issues, they will not be able to use SL-M orders, instead they use SL-Limit orders with higher bid price for Long and Lower ask price for Short. So that it would get filled in immediately. But still I have seen few orders not getting filled in, development team is working on it now to fix it

  • Open orders and Positions are not getting cancelled/squared off at the specified end time. Though MIS orders will be squared off by brokers automatically, this is not a high priority issue, but still need to be fixed.

  • Their EOD reports are not user friendly, difficult to calculate profit/loss details.
As far as the problems are concerned.. I am still having issues with my execution part, as my strategy has a custom execution logic which is implemented by them. I am still testing the execution only, have not moved on to the full simulation yet.

I will let you know about the above said issues once i reach there.
 

Similar threads