Congratulate me! Or not...

from my own experience, 2.24 per trade will definitely not cover slippage and commissions. i've done sim modes with scalping strategies where you do everything but send the order and got great results (actually, better than yours). but when i trade it with real money you'll see that slippage and ECN routing fees hurt you. you'll even get slippage on a thick stock like C which does 40+ million a day.

so lets say .005 slippage per side (but more like .01+), plus .003 for taking liquidity per side, plus commissions. thats could be more than 2.24 per 100 shares. and if youre using limit orders i'm sure they will be plently of 'just touched' orders too

thats unless youre scalping stocks that are at a stand still. but judging by the timeframe of your scalps it seems most likely your doing something while theyre moving.

anyways, the only way to really know is to send out the orders. just do 100 shares. hopefully at worst you'll just break even.
 
Quote from kanonka:

Subj!

"Dry run" of the black box that I wrote makes 16% (on average) A DAY. "Dry run" means full production mode, except orders are logged instead of being sent to execution (as I didn't code that piece yet).
More...

You better inform the fed reserve before you start your program.
With that return compounded, you will soon controll all the M1, M2 and M3 money supply.
 
Quote from kanonka:

Subj!

"Dry run" of the black box that I wrote makes 16% (on average) A DAY. "Dry run" means full production mode, except orders are logged instead of being sent to execution (as I didn't code that piece yet).

OMFG.

I hoped to 1-2%. But real outcome drives me nuts. I don't believe my own eyes. Now it's time to verify, verify, verify...

Just wanted to share, because I feel like I will explode if I wouldn't tell someone:)
More...


Just wondering, what software did you use to backtest this system?

And what broker are you planning to use?
 
I'll try to answer most of questions:

1. $2.24 is after fees. Fees are calculated based on MBTrade schedule ($0.01 per share first 500 shares, then $0.005 per share for over 500 shares). And of course, this is average. Minimum was ($-50.xx) and max ($+180.xx).

2. Slippage is also calculated into this gain (or better to say, projected). Cause I don't send real orders, I can only "try to project" what sale price will be. Once I issue the sell order, I take lowest bid price or trade price withing this and next second. Same with buy order (except I take max of ask/trade).

3. Software is written by me from scratch (C++). I spent last 6 months :( writing it.

4. The biggest assumption is that once I sell shares, broker will make those money again available for me within 1 minute. This might be the weakest point of the algo. Only few brokers calculate balance by open positions only, and I don't know how fast they restore it.


And yes, only practice will give real answers. Right now this is just dry run. I just wanted to share my excitement because I can't believe my eyes :).

-------------

And by the way, "cold", please don't behave like 13-y old calling other people names. I probably made a mistake somewhere in a logic/calculations, which I intend to research, but it has nothing to do with you using terms "fool" and such. Is this clear? This is last time I'm polite with you if you continue this way.
 
You must log in or register to reply here.
Back
Top