Formalizing a strategy: from handtrading to cTrader cbot
Formalizing a strategy: from handtrading to cTrader cbot
06 Sep 2021, 18:58
Several years I’ve been working in a hand-trading mode and recently (over a year ago in fact AHAHHA) started moving in a direction of creating a ctrader cbot to delegate all this decision-making crap to a heartless machine and enjoy my pina colada on a deserted carribean island while it makes money. AHHAHAA again.
What I wish I knew before I started programming cbots?
- The trading patterns that we can perceive visually from the charts are VERY difficult to explain clearly to a machine. To be totally sincere, I found them difficult to explain to a human either. At certain moments I felt like punching the programmer in his face if he asked again “yeah, but what are the exact parameters that allow us to understand that the price has consolidated here?”
- You can learn programming yourself, but you can’t develop algorithmic mindset fast enough. When I started collaborating with professional developers, I realized that the algos and solutions I use and they offer differ like a lego spacecraft differs from Perseverance rover.
- Programming ctrader cbot takes time. I used to think that developing an algo would take 2 week at max and if I’m not finished by then, I will hire someone at upwork and some geek will do the job over the weekend otherwise. Huh! Automating a complicated trading system requires a lot of polishing. I wonder if it requires continuing polishing (I start suspecting it is so!) You really can make your bot function somehow in a little while but polishing is what makes the equity curve pretty. You know, it’s never pretty enough. The sky is the limit.
Did you ever try creating a ctrader cbot? What were the insights?
Replies
modimeena37
08 Sep 2021, 12:54
RE:
Well, that's an opinion that some people share and other people challenge :)
@modimeena37
Nikor
09 Sep 2021, 23:33
RE: Formalizing a strategy: from handtrading to cTrader cbot
modimeena37 said:
Several years I’ve been working in a hand-trading mode and recently (over a year ago in fact AHAHHA) started moving in a direction of creating a ctrader cbot to delegate all this decision-making crap to a heartless machine and enjoy my pina colada on a deserted carribean island while it makes money. AHHAHAA again.
What I wish I knew before I started programming cbots?
- The trading patterns that we can perceive visually from the charts are VERY difficult to explain clearly to a machine. To be totally sincere, I found them difficult to explain to a human either. At certain moments I felt like punching the programmer in his face if he asked again “yeah, but what are the exact parameters that allow us to understand that the price has consolidated here?”
- You can learn programming yourself, but you can’t develop algorithmic mindset fast enough. When I started collaborating with professional developers, I realized that the algos and solutions I use and they offer differ like a lego spacecraft differs from Perseverance rover.
- Programming ctrader cbot takes time. I used to think that developing an algo would take 2 week at max and if I’m not finished by then, I will hire someone at upwork and some geek will do the job over the weekend otherwise. Huh! Automating a complicated trading system requires a lot of polishing. I wonder if it requires continuing polishing (I start suspecting it is so!) You really can make your bot function somehow in a little while but polishing is what makes the equity curve pretty. You know, it’s never pretty enough. The sky is the limit.
Did you ever try creating a ctrader cbot? What were the insights?
That's a very useful post. I'd also add another point to this 'I wish I knew' before creating a Ctrader cbot list:
It's a very time-consuming task. Timewise you can't efford having a full-time job and seriously working on the strategyalong with continuing your hand-trading experiements. If you do so, the burnout is around the corner.
@Nikor
Gotfreeld
07 Sep 2021, 13:06
LOL, I didn't even try. I don't think that strategies based on visual patterns recognition and price action can be formalized clearly enough.
@Gotfreeld