Frosty's trading bot goes live part 2

[malaka56]

Thanks for the tech info on your ATS. Interesting. Not to get too offtopic here, but about how long did it take you to code all that up, and any thoughts about using C# for ATS design?

Some of the systems I have seen in Java that track a large number of stocks are annoyingly slow.

Good luck frosty.

 

[frostengine]

malaka,

I have been through so many iterations that its hard to say how long it took to code it up. The code is always evolving anytime I have a new feature or idea I want to add.

For just a bare bones trade execution engine you can code that up in a matter of a few weeks. Same with a bare bones backtesting engine, can code that up in a few weeks. It all depends how advanced you want to get with the system.... the part that takes the longest is developing strategies. Getting some software up and running can be quick, especially if you don't need a lot of features. Such as if you know your going to be trading 1 contract on x bars.. that is a lot quicker to get that trading/backtesting engine up and running than it is if your going to sometimes be trading on x bars, sometimse on volume bars, sometimes on range bars..... mutliple contracts at the same time.. all of those features take more time to implement.

Concerning C#.... It wont be as fast as C/C++.... but you can write a good system in C# and be just fine.

 

[frostengine]

I guess I jinxed it PRETTY bad... ended up only making +$35.60....... Had it not reversed and went short RIGHT after my message it would have made well over 1k today... oh well.... another so-so day, at least its green...

Today +$35.60
Week to Date: -$352.8
Since going live: -$1675.6
Total days: 8
Avg per day: -$209.45


In other news, I am thinking of going live with my EOD bot stock picking strategy soon as well. Will make a new journal for recording the results of that one when I go live. That one I will probably supply the next day picks generated by the strategy before I even take them if anyone is interested. If not i'll just do like I do in this journal and chronical the results.

 

[lrm]

frosty,

Awesome thread, keep it up. I'd love to watch the EOD strat as well, and you'd definitely get a gold star IMO if you posted the day before.

Very cool!

Louis

 

[frostengine]

Another bad day, what a surprise.... when its running on the sim account I never notice the day to day losses.... all I notice is wow overtime that sim account is growing and growing..... different story when you see day in and day out what is happening....

Today -$369.60
Week to Date: -$722.4
Since going live: -$2045.2
Total days: 9
Avg per day: -$227.4

Up Days: 4
Down Days: 5

 

[GTS]

Quote from frostengine:

Another bad day, what a surprise.... when its running on the sim account I never notice the day to day losses.... all I notice is wow overtime that sim account is growing and growing..... different story when you see day in and day out what is happening....

Hang in there Frosty!

 

[nkhoi]

Quote from frostengine:

Another bad day, what a surprise..

that's a surprise, bot should be able to handle fed day well since all fed days have the same characteristics.

 

[frostengine]

Fed day, or not fed day doesn't matter. The bot doesn't notice the difference. Just another day..... was doing fairly well until the afternoon when it started bouncing up and down in a range

 

[nkhoi]

maybe a sub-routine to handle fed day is called for since it does behave diff than the non fed day.

 

[nonlinear5]

Quote from nkhoi:

maybe a sub-routine to handle fed day is called for since it does behave diff than the non fed day.

Yeah, but then you'd also have to code around PPI days, CPI days, GDP days, ISM days, Home Sales days, and maybe even Truck Sales days.

Regarding the current drawdown of 2K, I'd note that it's larger than the largest one on the equity curve posted earlier as "currentmeth.png", so maybe something is broken. On the other hand, maybe it's just the manifestation of the "your worst drawdown is yet to come" principle.