How much time do you need to paper trade your strategies before going into production ?
I’m curious if there are traders here who think that backtesting is enough and go directly into production or the opposite if they directly paper trade and skip backtesting.
I would go directly into trading BUT with very limited size and increase the size if strategy shows expected results.
check you system back at least 10 years, and start trading on small amount for a year – after that you ready to go
THERE ARE A NUMBER OF COMPANIES THAT LET YOU PAPER TRADE BEFORE PUTTING UP MONEY. TRY WITH THEM FIRST
Everyone is going to be different. I would personally recommend trading live, once you feel comfortable. If you need a time frame…I would say a few days. It’s ridiculous to do it for weeks on end without making a move. And quite honestly, I have found that a good strategy develops and gets fine tuned from losses you incur in the beginning.
speaking of back testing: have you ever tried back testing as a cloud service in a high performance computing environment such as amazons ec2 cloud?
I’ve been researching the stock market for over 35 years. I know this does not give me any merits for being obsessed for that many years, but then again, it does give me a certain degree of perspective on things.
If you do not back test, then welcome to the club of the also ran: your portfolio has chosen oblivion as the ultimate outcome. New traders take less than 18 months to realize that their account has crossed the zero line and are never heard of again.
If you think that making a few baskets over the weekend entitles you to apply to the NBA and then on the next Friday apply to the NBA Hall of Fame, then think again. This stock market game requires a lot more, and if you want to win, you better determine your own rules of engagement.
Back testing, back testing a lot, is the only way that you can secure an edge. At least you will have proven to yourself that it is feasible to trade profitably. However, should you cheat on your back testing meaning curve fitting or over optimizing your trading program then be assure that the market will curve fit your portfolio to reality.
The other side of your trade will not hesitate to cheat, misrepresent, deceive or contrive what ever explanation is required to keep you long on the declining stock. And should you adopt a dip-buyer scenario on the next Enron (or the numerous others) you will simply have shorten your expected duration as a wannabe trader.
Your objective is not to blow up your account. On the contrary, all you want is doing your thing and then collecting your justified rewards. However, remember that long or short on a rising stock, the other side of your trade has just the opposite view.
I do think, paper theory won’t close the gap for you to be a trader on the winning side. So it doesn’t matter if it’s on a paper or on run from an automate. When you can’t win money manually, no magic will happen.
Thank you for all your interesting answers. For the moment, I don’t have the ambition to become trader. I’m just curious about usage and custom of traders.
I agree with you that backtesting can at least give you confidence in yourself. But I also think that now you have to trust your automated trading system. So maybe a paper trading phase before going to live can be necessary to aquire this trust. I’m developper, so as far as I’m concerned I will trust my work but what happen for users ? (I guess that all systematic traders wasn’t software engineer at first )NOTE I now post my TRADING ALERTS into my personal FACEBOOK ACCOUNT and TWITTER. Don't worry as I don't post stupid cat videos or what I eat!