Frosty's auto-trading bot goes live with REAL money

[brownsfan019]

Quote from jonnyman:

While that's true, I think the timeframes depend on different situations. For example, our system took about 3 months of dedicated work once we had developed our algorithm, which took about 10 months of on again off again testing and research. However, right now the system is running completely independent of outside human interference. If you're in the trenches every day emotion affects trading. Even after years of experience I still find myself getting angry etc in some trades.

The key is to get as quickly as possible get to the point at which you have the confidence and the mathematical research done to allow you to leave the bot on all the time and simply watch the results. You are right - it is not for everyone, but I believe that it is at least worth trying. If you give up before you have started then you have truly failed.

jon - I think you may be in the minority with the amount of research, time and money you have in your bot. That's purely a guess, but I would imagine most people do not have the dedication that you did.

If you can get to the point where you turn the bot on and walk away and are confident in the bot, then that may be a great setup.

It sounds like it comes down to:
1) Having the utmost faith in your bot (which a human programmed).
2) Being able to turn it on and walk away. And by walk away, I mean for days, weeks, months, etc.

After reading some of Frost's journal here, I am not sure those components were there.

This is a great discussion, at least someone is challenging you programming guys!

I think any newbies reading this now or in the future just need to at least examine how this stuff works before programming a stochastics crossing system and then wondering why it doesn't work. If you can trust your bot 100% AND trust that you will not interfere with it, have at it!

 

[jonnyman]

Quote from brownsfan019:

jon - I think you may be in the minority with the amount of research, time and money you have in your bot. That's purely a guess, but I would imagine most people do not have the dedication that you did.

If you can get to the point where you turn the bot on and walk away and are confident in the bot, then that may be a great setup.

It sounds like it comes down to:
1) Having the utmost faith in your bot (which a human programmed).
2) Being able to turn it on and walk away. And by walk away, I mean for days, weeks, months, etc.

After reading some of Frost's journal here, I am not sure those components were there.

This is a great discussion, at least someone is challenging you programming guys!

I think any newbies reading this now or in the future just need to at least examine how this stuff works before programming a stochastics crossing system and then wondering why it doesn't work. If you can trust your bot 100% AND trust that you will not interfere with it, have at it!

Hmm I honestly thought that we were about on par with the kind of work the amateurs (like us) are putting into things. Though we have a great system that makes good money I still consider our little group fresh blood.

I can leave the bot for hours on end and not be concerned. I'll often go to lunch and/or the gym from 12-2 and not be worried at all. I only go back to make sure the machine is flat at the end of the day (an irrational worry as it hasn't happened yet but still...)

It would be interesting to know how much time and money and research people DO put into these things. I've been watching the automated forums for a while and I don't really see that discussed (maybe it's like asking people what their profits are - like pulling teeth) but perhaps I missed it.

I suppose it may be different when trading derivatives, but for equities live trading is easy and low risk as you can trade 1 share at a time, which is what we did for a few months. I highly recommend it as you get the live element in there, but the risk is almost nonexistent. If your bot does terribly you'll know and won't take a huge hit to your capital.

I can see where you're coming from but only if you make the assumption that it's a wild west where people are putting in little time and expecting big returns. In any industry that's a fools gambit.

I'm also going to go out on a limb here and say going solo is almost impossible. I know for a fact that I could not have done this alone. Period. If you can find even 1 person (preferably a friend) that can match your weaknesses with strength you have a much higher probability of success.

 

[frostengine]

-$175 today

 

[August]

I just spent until 5:50 am reading this thread and learning a tremendous amount from everyone, thank you.

While I'm a newbie trader by any starndard (a student really) - I've done well in business management and from that perspective I was compelled to echo some other people's mentioning that seeking out a partner is a wise choice.

Tradebot is reported to be the brainchild of one man - but this is a man who had extreme experience in the market. And, furthermore, I think they have like 30 people in their company now and I betcha that at least one other person there was involved from the start who may have provided some bit of perspective that a singular mind might not have covered. Synergy is often overlooked.

Just a suggestion. Much respect for you exposing yourself to the community and your openness. And mostly I thank you for making me realize how unqualified I would be to do something like this.

It all does leave me with a question though.

Everywhere I look I read "trade emotionless - stick to your system and write it down." It seems like if this is really the case - an ATS is the only way to trade. Yet, not everyone trades that way. And some people trade with ATS unsuccessfully. So, what is the X factor of successful traders?

Anyone who is trading an automated system successfully must have harnessed that X factor to some degree.

Is it simply a matter of taking more conditions into account? (Meaning more variables considered for trades?) Or is it a matter of quantifying and simulating "gutt" response such as "This is a bad day, moment, time - I'll take an early lunch."

 

[Trader273]

Quote from August:


Everywhere I look I read "trade emotionless - stick to your system and write it down." It seems like if this is really the case - an ATS is the only way to trade.

A lot of people believe this. However, the markets are moved by emotions( at least in the short-term). Take a look at the attached article from Brett Steenbarger.

 

[trendtech]

Quote from jonnyman:


I suppose it may be different when trading derivatives, but for equities live trading is easy and low risk as you can trade 1 share at a time, which is what we did for a few months. I highly recommend it as you get the live element in there, but the risk is almost nonexistent. If your bot does terribly you'll know and won't take a huge hit to your capital.


Hi,

Sorry, I don't trade equities so forgive my ignorance. Is it possible to trade 1 share? I thought that the minimum quantity for stocks was 100shares.

Good trading

trendtech

 

[August]

Quote from Trader273:

Take a look at the attached article from Brett Steenbarger.

Thanks for that. This article is a perfect extension to my question Here is a perfect example of a "gutt feeling" - but could these factors not be programmed into a system? There was a stimulus which triggered this response.

I felt a brief stab of fear--the feeling of "this isn't right!--and quickly took my three-tick profit. Almost immediately after, the trade was a full point under water.

and an even more ambigious yet perfect example

Frankly, in my experience as a psychologist to traders, a lot more traders have perished from overconfidence than from fear.

If you accept this as true, then could you not research what things cause fear in a trader?

Perhaps if one were to conduct interviews with a successful trader (or more), one could profile things such as this?

Would that be the X-factor? which is basically gutt feelings based on knowing when something doesn't seem right beyond basic indications?

 

[jonnyman]

Quote from trendtech:


Hi,

Sorry, I don't trade equities so forgive my ignorance. Is it possible to trade 1 share? I thought that the minimum quantity for stocks was 100shares.

Good trading

trendtech [/B]

We did it, so yes it is possible. It may depend on your broker - perhaps they may limit trades to a minimum of 100 but I don't think that is popular. We did get charged steep fees and commissions, but on 1 share they are still tiny (only a couple cents per trade).

 

[frostengine]

The bot has just closed out for the day.... today was an EXCELLENT day... this is the type of market where the bot really excels..... Profit today is +$1280...

I wish I would not have stoped the real money tradding the other day... because past few days have been great and MORE than made up for that one bad day that blew me out..... but you live and you learn...... I wont be going back to real money anytime too soon.. but glad to see the bot catching a rhymn again... I hope it continues...

 

[nonlinear5]

Quote from frostengine:

glad to see the bot catching a rhymn again... I hope it continues...

I assume the bot is non-adaptive. That is, it doesn't change its parameters during the session or even in between the sessions. If that's the correct assumption on my part, then it's more like the market is catching the bot's rhythm, not the other way around. You may be using that reference as a form of expression or a play of words, but it could be misleading, too. For example, one way to interpret what you said is to think that the bot runs a neural net or some other "machine learning" mechanism to tune itself regularly to the current market. Again, I assume it's not the case with your bot.