@cunparis, Insightful comments regarding backtesting. I paid that tuition, too. Trying to get that ideal parameter optimization That was an expensive lesson for me (more in time than in P&L). Now, I take different approach to backtesting. I try to gauge the parameter sensitivity and then find a plausible value within the range where the parameter works. I don't care if it the version I run live doesn't backtest as well as another version. I'm more concerned that it's robust enough to hold up out of sample. Personally, I think the backtesting tools given to retail are there to give them hope and keep them trading. Without backtesting many rational people would figure out that they have no edge and drop out sooner but backtesting always gives them hope that they've come up with some magic rule/indicator set. I don't agree that all systems fail. I think that all rule-based and indicator-based systems WILL fail though. Also, I think that good systems CAN fail b/c the trader fails to account for parameter evolution over time. That's actually my biggest concern right now, getting a firm grasp on the parameter evolution over many years. Where's your blog. I'll check it out. -bluelou
I totally agree on the first part there. Ninjatrader is very smart. It's free for simulator & backtesting. So once I thought I had an edge, I paid for it and started trading. Edge lasted a month and fell apart. But by then I had some variations of my system and those lasted a few weeks. Soon I was down over $10k and couldn't figure it out. Meanwhile they pocketed $1k. By the time people realize it doesn't work they've already paid and lost money. If only I had run my systems on simulator for 3 months instead of 3 weeks. my blog is http://www.tradewiththeflow.com
I read it. Sounds honest and painfully familiar. At least you have a realistic idea of what you don't know. That can help you from flushing your money down the trading toilet. I'd offer ideas if I could but I don't trade discretionary.
That was part of my idea for the blog to discuss my trading. And in the process I realized what I was doing (with the cycles & timeframes) was all wrong. So many people have told me that their story is almost identical, so I think it's great to share. And will save newbies a lot of time & money. Man the money I wasted to learn what I know know.. Now my trading is going very well. I'm averaging $250/day consistently with 1-2 contracts. So my next challenge is to get my stops tighter and slowly increase size. Trading is really simple but that doesn't make it easy. When we make it complicated that makes trading impossible.
I'm glad it's working out for you now. I doubt too many newbies will be saved though. It's just human nature to repeat the same mistakes over and over. That is, unless you've found a cure for the rashness of youth and inexperience.
It seems like a lot of people go back to trading their ideas by human hand on account of it's more challenging to get a code a non-linear vs linear system or dynamic vs non-dynamic system. Would you think this to be the case?
I really don't understand the question. Can you elaborate? I trade "by hand" because I react to what's happening. I try to dance with the market. If my trade doesn't move my way I get out. I trade off intuition and that's not possible to code. I had two nice swing trades lately. First short from 1207 got 20 pts on last contract. Then long from 1174 still holding 1/4 of that one. Both were purely intuition based. Sure I'd love to automate and relax on the beach but it's a pipe dream. It doesn't work. When people accept & embrace this trading gets easier.