This statement sounds funny. According to your graph, your systems made money even on the decline of 2002. So, now you're afraid a "flat" equity curve? You never completely answered my questions as to how many stocks are being traded at the same time with this system of yours? 5? 100? What is the average hold time?
I appreciate the feedback very much. I'll have more to say about the systems later...as I get the time. It may be there are fatal flaws I'm missing. m
Believe me , there is nothing mysterious in systems building anymore IMO, so you are not likely sitting on something you should not ever reveal. remember other party brings experience and capital needed and you definitely need someone with more confidence and more capital then you are possessing right now. They are probably familiar with something similar to your system anyway. My guess is that your capital is very close to a maximum drawdown, which , understandably , brings indecisiveness to the play. Good luck and do not give up ! Walter
Exit on target or on open on day 4. No, they are quite straightforward and mechanical. The only issue is filtering when there are too many signals, but there are several ways to do this without impacting results. m
Yes! I'm afraid to trade them and that is why I posted this in the pscyhology area. One system scans all stocks with 50 day average volume > 50k and the other scans stocks with 50 day average volume > 250k. Some days there are no signals. At times there are too many to take with my small account. Max number of entries is about 30 in one day. When there are many signals, I filter with volume or % change. hold is 1 to 4 days (4th day on open for one system; 4th on close for the other). m
I agree with all you said above. Yes, I'm near max DD. That is from my discretionary trading which was disastrous over the last three years. Pscyhological factors that I can think of that are influencing me: 1) The idea for both systems is so simple that many people must use the same ideas, so I fear they will stop working at any time, just like many breakout systems and moving average systems stopped working. 2) My stock universe does not include symbols that no longer trade, so my testing includes survivorship bias. However, the last year seems to support the conclusion that they work. 3) I've become hypersensitive to losses (wish I had had this problem several years ago!). 4) The systems both sustain significant individual trade DDs at times and no stop system that I've looked at seems to fix this. Stops just shift the win/loss ratio around and generally degrade returns. 5) Even though they are mechanical they still take a lot of time and I still have my day job, so it is hard to devote enough time to do things right. I have a bad habit of taking the gains from the systems and pissing them away on ES trading attempts. Boredom is another problem. System trading can be drudge work. I'm sure I'm just scratching the surface on the psych issues. I do feel that if I had a means of automating all the steps and keeping my hands off, I'd do much better. m
You can automate to IB via :http://www.ninjatrader.com/features1.htm You can run this on the simulator for free to get a feel for it. http://www.linnsoft.com/autotrading/index.htm http://www.tickquest.com/NeoTicker/ Both of these work with ninja. Look into these programs. I dumped IB recently for other brokers so I don't use ninja any more, it was a great little program for me.
Ninja doesn't look like it would do what I want. I suspect something like TS would, but I think I'd be better off creating my own program...if I could find the time. m
1/ Stick with the profitable only , for now. You can play later when you well ahead . This should not be too hard, it is not like you are working in a bloody coalmine . 2/ Do not overanalyze yourself, we all are much simple then we think we are . 3/ fully automated systems need to be overruled sometimes so I would be cautious about keeping hands off too far . Can you use your method on big issues like QQQ, MSFT, SPY ? If yes, it might simplify things for you a bit. Walter
Thank you Walter for the comments. Both these systems require scanning a very large stock universe: 1500 to 3000+ stocks, in order to get sufficient signals. m