Thoughts on trading bots?

[7.62x51]

Can a profitable trading bot exist in theory? If one did exist, I don't believe it would be sold.

However if trading could be broken down to an algorithm with guaranteed profits then why wouldn't everyone be doing it?

What I'm curious about is where do the bots fail? With recent advances in machine learning which allow for anomaly detection and pattern recognition, why has no one made a bot which is good enough to beat randomness?

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[AgainstAllOdds]

Can a profitable trading bot exist in theory? If one did exist, I don't believe it would be sold.

However if trading could be broken down to an algorithm with guaranteed profits then why wouldn't everyone be doing it?

They do. Read this book: http://www.amazon.com/gp/product/B00HVJB4VM/?tag=tff-amazonparser-20

 

[Leo Hendrix]

If it could recognize momentum, trade and time when to get out.

 

[illmasterj]

https://www.reddit.com/r/algotrading may be of interest.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[Sanj Modha]

I tend to stay clear of any automated tool that promises to make me money. That's across trading, investments, e-commerce too.

My argument has always been the same. If it was that easy - we'd all be millionaires.

 

[Sanj Modha]

I've also found that most bots are scams when you dig a little deeper.

 

[Raoul Duke]

They do. Read this book: http://www.amazon.com/gp/product/B00HVJB4VM/?tag=tff-amazonparser-20

https://www.nadex.com/learning-center/news-and-market-commentary/booms-bots-and-bogus-flash-boys
http://www.forbes.com/sites/russala...-investment-management-industry/#59af0bb92296

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[InspireHD]

It's possible to program a trading strategy, such as looking for certain patterns. Let's say you have a trading strategy that is profitable and the setup you watch for works 60% of the time. You create an automated bot that will trade those setups 100% of the time.

The problem with humans is that when you're looking at the "hard right edge" of the computer screen, you may second guess yourself, "what if the setup doesn't work this time?" (Trust me, I've done it a lot of times only to watch price do exactly what I was expecting).

As a human, you don't take 100% of the signals. Now, your trading strategy may only be profitable 50% of the time because you second guessed yourself on a couple of them. At 50% profitable trades, you'll actually be losing money because of commissions.

Further, it's difficult to backtest trading strategies because you may just be fitting the data to see results that look good.

Several years ago I had the opportunity to do some online day trading with the help of someone I met on another forum and it was one of the most eye opening things I've ever witnessed. I wasn't able to replicate the other trader's success because I let too many psychological blocks get in the way. But, I'll be the first person to tell you that day trading can work. I'd love to do it again someday.

 

[Efx]

At 50% profitable trades, you'll actually be losing money because of commissions.

Depends on what Risk-Reward youre aiming for, if youre right 50% of the time and your RR is 2:1 you'd be making money, however if your RR is 1:1 or less youre right.

 

[illmasterj]

I tend to stay clear of any automated tool that promises to make me money. That's across trading, investments, e-commerce too.

You have a website, right?

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[ApparentHorizon]

If you had a program automatically making money for you, how many people would you tell? How many licenses would you sell? How much competition could the market handle before the algo became ineffective? How long will the algo work before wallstreet reverse engineers it?

 

[samuraijack]

https://en.wikipedia.org/wiki/High-frequency_trading

 

[nitop]

its possible , as i do it everyday..
but its not easy , and certainly not static .. as a professional quant , we know that there is no such thing as a "perfect bot" ,, its more of the way pros use to design their portfolios consisting of noncorrelated strategies and trade them all at once to generate alpha..
traders need to be nimble , and "know when to throw off a certain strategy algo that just lost its edge".., as all algos lose their edge overtime ..

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[JokerCrazyBeatz]

I never thought of bots for stocks , sounds risky

 

[Sanj Modha]

If you had a program automatically making money for you, how many people would you tell? How many licenses would you sell? How much competition could the market handle before the algo became ineffective? How long will the algo work before wallstreet reverse engineers it?

THIS! If you had a system that made money on autopilot - why on Earth would you sell it?

Leave it running and go party. People who can't sell sell to others. It's the same in e-commerce.

 

[ApparentHorizon]

THIS! If you had a system that made money on autopilot - why on Earth would you sell it?

The questions weren't rhetorical, although in certain niches it certainly would be.

If the algo will last 3 years and you can sell 3 copies to companies, in the seven figure range and you all make money, then there's definitely a case.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.