That kind of mistake in coding shows wreckless abandon for the instruments he's trading. No strategy can ever be profitable losing 80% immediately, because that's not just slippage, it's carelessness.
Glad to see you eke out a small gain but it sounds like you could have had a 5% week if not for the automation error. Honestly automation scares the crap out of me for exactly this sort of thing. I kinda get the impression you prefer automation mode as a way of not taking responsibility for your trades at this point. Wouldn't it be better if your system sent you alerts on the condition and you opened positions on your own?
you stated openly how many times you've blown up without success. "Automation scares me" ha ha. Keep the expert hat out of this room please.
"The more educated, experienced, or successful any trader becomes, the more susceptible that trader is to trade attachment, especially if he has previously taken a lot of money from that particular market previously." -J. A. Jankovsky I thought of Neke just now when I read this. GL man...
Weekly Update for week 16/50 ended 5/5/2012 Moderately positive week, up 2.2K (1.4%). Made 4.4K via automation but lost 2.2K on discretionary. Took out 6K for the month ended 4/30/2012. Will still cut further on the discretionary as it has become a drain. Ideally discretionary trading should be only for HC scenarios that I couldn't have foreseen with my automation, but instead I have applied it to anything that comes by. For the three weeks 4/14/2012-5/5/2012, account is flattish (down $400) versus target of +6K. 39 automated trades made +10879, even with costly execution bugs, while 28 discretionary trades lost 11286. So the direction is pretty clear - hold off as much as possible from the discretionary trades. Will maintain my risk level at the same level (max 3% size per option position) while target is set at +6K for the next four weeks (5/6/2012 - 6/2/2012). Code: Opening Balance: 161,162 Net gain for the week 2,243 cash Withdrawal (6,000) ------------------------------------------------ Net Balance: 157,405 Since Inception of Thread 01/18/2012 - 05/05/2012 Opening Balance: 203,729 Net loss (23,324)(Down 11.4%) Cash Withdrawal (23,000) ------------------------------------------------ Net Balance 157,405
Dang guess I was wrong about the automation part but I'd still be concerned about errant position openings long or short with automation.
The "errant" position opening is more likely to happen when the trades are placed manually. My automated system has placed hundreds of trades, and there was never a case when a trade was "unwanted". It did take a while to shake out the system bugs, though. But that's what a simulated account is for. By neke's own admission, risk control and position size control are not his strength, and that's where the automation comes in, with the machine discipline, patience, speed of entry/exit, and rigorous rules and parameters to maximize the reward/risk ratios. What's left in favor of the manual trading? The alleged ability of a trader to outsmart the computer? Maybe, but my own system outsmarts my trading intuition consistently, and by what I sometimes consider a disturbingly large margin. The markets change, so what worked for neke for a number of years clearly doesn't work anymore in his discretionary trading. His automated record, on the other hand, appears to be much more impressive. So, it's rather obvious what should be done -- instead of fighting the market with the manual system that has lost money over the last 2.5 years, focus the energy on perfecting the automated system.
Just one problem: most funds use automated systems and trading is a zero sum game. Now what does that mean?
It means that your automated system is competing against their automated systems, and yes, the fittest ones will survive while the weak ones will perish. However, this is hardly a good argument against automation. That's because if you trade manually, you are still competing against the auto systems, and by definition, they have an edge against you.
I am glad you agree that most of the systems will lose money. But it is still a question whether a system has an edge against a human. Systems are designed by humans, a system cannot do 100% of what humans want to do. By definition, human is better than a system.