Is a demo account waste of time?

[galvinlee888]

The biggest reason I went into quant-hft-algo land is because I hate the sweaty palms and blood pressure of manual trading. Software makes trades with ruthless cold blooded precision that I never could myself. It has no hope, no fear, no greed. Just math.

This only shows your still "unsure" if your manual system will work.

I never get excited when I am making or losing money in my manual trading, I just repeat the same process again and again and let the market tell if i win or loss each time. I know I will be in profitable if I keep on doing this and never have second guess in my trading approach. The confident level is build from experience for more than 9 years. I start will real money account and never use demo account, with small position before I go with the "normal" money.

 

[Risk619]

I was an options mm. Toward the end, I made markets electronically and automated. With that type of trading, you can't make money without risk. You can't get rid of all your Greeks at eod without a cost.

For sure, I don't think mm or any other HFT strategy is without risk by any means, but it's the way you handle the risk that's so different. I do research, back test, run in a paper-trading mode against live data, watch for curve fitting, etc. You're just employing a vastly different set of tools and strategies to handle the risk. All of which is done in a relatively calm and peaceful setting, none of which has to happen in a peaked blood pressure mode.

I would much rather spend a year honing a decent set of algo models then sitting there every day going through the market drama. If something goes wrong the algo (properly built) should shut down, it goes offline, it's fixed, and put back into service. To me it's the equivalent of any other piece of production software.

 

[Risk619]

The million dollar question is how your system can sense the market, namely is it trending, sideway, level of volatility and etc. It is easy to say this in hindsight (as most ET here did), not easy to predict it in real time. Hint: The usual BS education material such as 50MA cross 100MA and etc will not work, at least not have any statistical significant data to proof it works

The models (that make up an algo) are so different that it's hard to say what variables you use for what. But I would agree with you that anyone who thinks they have a risk-free trading system should be stripped of anyone else's money but their own. I'm of the mind that risk is the stuff you didn't model for, the correlations you didn't allow for, the black swan you just didn't anticipate.

I honestly used to be obsessed with alpha models and buy/sell signals, but what I've figured out is that they're really not that important, or at least only as important as the trx, risk, and execution side of things. You can data mine and figure out patterns, but that's a far cry from actually having code that executes properly, uses the right order types, provides/takes liquidity from the appropriate exchange, holds into winners and kills losers.

I'm not really saying anything that isn't widely known to anyone in the quant/algo/blackbox world.

So much depends on the time horizon as well. I don't really have any experience with longer time horizons (beyond intraday), although I think I'd like to learn that part.

 

[Buy1Sell2]

Demo account is largely a waste of time. Except for the ability to learn the orders and the logistics of the markets. You will learn far more by losing real money in the real market. --Must lose to learn how to win. Thank you for your time.

 

[Risk619]

This only shows your still "unsure" if your manual system will work.

I never get excited when I am making or losing money in my manual trading, I just repeat the same process again and again and let the market tell if i win or loss each time. I know I will be in profitable if I keep on doing this and never have second guess in my trading approach. The confident level is build from experience for more than 9 years. I start will real money account and never use demo account, with small position before I go with the "normal" money.

My "system" at this point is so fast that no one (you or me included) would be able to execute upon it. I can't talk specifics but I don't think I've ever algo-held a position longer than it would take to spread some butter over a slice of toast, and we're talking about thousands of securities.

It's akin to cellular biology in that unless you have the tools to see and conduct work at the microstructure level, it's all just going to seem really out of scope. Conversely, I feel like a drunk whale in a sea of jello when I manually trade.

 

[galvinlee888]

My "system" at this point is so fast that no one (you or me included) would be able to execute upon it. I can't talk specifics but I don't think I've ever algo-held a position longer than it would take to spread some butter over a slice of toast, and we're talking about thousands of securities.

It's akin to cellular biology in that unless you have the tools to see and conduct work at the microstructure level, it's all just going to seem really out of scope. Conversely, I feel like a drunk whale in a sea of jello when I manually trade.

Understood what you talk about. I worked in institute before and one of my best buddy is from a similar group as yours (I am in the equities/options but we both get well as we were from the same college). One bug in the code will cause his (and his manager) job.

I try to not get drunk when I am trading, a fat finger will wipe up many days of my profit which means I will need more drinks after that

 

[lawrence-lugar]

I'm not a fan of simulation accounts.

Trading in a simulation account, and in real life...are two...entirely...completely different beasts.

That's like playing Street Fighter on Super Nintendo...and saying you're ready to step in the squared circle with an opponent...

You just gotta study what you are planning to trade -- and just...Jump In...for better or worse...
The market is part art, part science...constantly changing. -- Your performance in a demo account doesn't necessarily correlate to real-world performance.

Consider it...tuition money -- if you're smart, or good, ...you will grow it. -- if not, then maybe this world is not for you.

 

[i am nobody]

The million dollar question is how your system can sense the market, namely is it trending, sideway, level of volatility and etc. It is easy to say this in hindsight (as most ET here did), not easy to predict it in real time. Hint: The usual BS education material such as 50MA cross 100MA and etc will not work, at least not have any statistical significant data to proof it works

Indeed that is the million dollar question. Indeed the usual material is not working. Most things that are common knowledge is like garbage, if it would work everybody would use it.
Some people are able to define the trend with very high probability, but most cannot. But those who know will not share.

 

[dbphoenix]

Indeed that is the million dollar question. Indeed the usual material is not working. Most things that are common knowledge is like garbage, if it would work everybody would use it.
Some people are able to define the trend with very high probability, but most cannot. But those who know will not share.

It's not particularly challenging. If one can connect two dots, he can define a trend.

 

[Turveyd]

Trading to me, isn't that hard, it can and is as simple as which was is the market going, jump on it, exit when it stops, I bet in a demo account you can with no real $$'s at stake do that all day long and make a profit.

Live account $$$'s at risk, you'll over think it, 2nd guess yourself, greed will make you get in too soon when it's not really going up or down, fear will scare you out.

Although it's good to trade demo and prove, above can work for a while, it does nothing for you risk wise, got to be learn to accept that risk with real $$'s and the total uncertainty of trading, at best it's always a educated guess, got to accept that and trade well anyway.

Demo is good for learning the way around your platform to, stop you from going the wrong way, adding more positions when your trying to close, especially if MT4 hate that POS!