Hi, I was wondering if somebody has experienced something similar: I haven't reached a new equity peak for about 6 months now. I had a good start last year, but then somehow my equity curve stopped rising. I can't say I am in a drawdown, my equity seems to stay flat below the peak. I trade 4 mechanical trading systems, with time frames ranging from intraday to several months. Any comments? Thanks, Pat
PatG, The premise for a systems tends to deteriorate over time. In particular, if you have too many parameters the systems tend to do worse and worse the farther away from the "release day" you get. I suggest the following: 1. do an equity curve for each system and see how it looks. Often systems make the majority of the profits in a relatively short period of time. Do R-squared analysis of your returns and make sure that the system doesn't produce profit by chance 2. review/optimize your systems again. If the "new" parameters of significantly different, you should question the validity of your system and/or the initial value set 3. has the equity curve for the four systems ever before had a flat 6 month period ? For longer term systems, being flat for 6 months happens all the time 4. do you have a system bias (ie trade only longs)? that could explain quite a bit.
May be it's based on our volatility. I don't know what you are trading but the vola in the S&P is on a very low level. Hope we get better times back
Six months flat for a systems trader is nothing unusual. What was longest drawdown in backtesting? In fact, being flat without a substantial drawdown may be a sign that your systems are robust.
I cannot offer any further comments regarding your 6 month flat equity curve other than the ones already mentioned. I am a discretionary trader and attach my 1 year equity curve for academic comparison. Annotations (1) The period is from January 1, 2001 to December 31, 2001 (2) This should not be misconstrued as a solicitation to manage OPM. (3) Note the change in 2nd and 3rd moments of equity following September 29 (4:1 margin)
I am having a similar (but different experience) right now too. I recently broached this question (gee what's going on here) to a trader i respect, who shall remain anonymous lest i be mauled by the blood hounds on this site, they (gender nuetral) suggested that: "a range is a range and must be traded like a range -- not like a trend, because a trend is a trend, and must be traded like a trend." to which i replied, well -- gee thanks, but i only know how to trade a trend "then stay out of the range or adapt and trade like it is a range" seems like sound advice and a good explanation to me. regards/greg
Pat and gh1, With numerous range contractions occurring volatility has lessened during the past year. I have some systems that have gone flat. Re-optimizing them for Net Profit does not change anything significantly. When you experience a period of 6 months where a system is not making money give it a rest for the next 6 months. Then re-examine it. Or you can try optimizing the system for something else, such as Percent Profitable. There are so many good systems to trade that there is no point in staying with one (or more) which has stopped working. In this situation you should write new systems quickly. You should have dozens of them ready to go, and choose the best ones that match your trading style. I build new systems all the time because I never know what I will run into and it is so much fun to try new things. Just this morning I built a very stable and solid system that generated a few hundred trades and has a Profit Factor of 5.65. That figure is amazing to say the least. The Equity Curve has a nice 45 degree angle, and no parameter has any unusual, extreme value. It would be interesting to read more about the 4 systems you are trading now. What type(s) are they? What are you using them on?
Thanks everybody for your replies and advice. I guess I will have to take a close look at my systems. In general, I don't want to over-optimze, and that's why I only re-evaluate and re-optimize when a system gets into a deep drawdown. I have done so about 3 month ago with 1 stock that I had to replace, since it went into a 30% drawdown. I only trade stocks sofar, 2 systems trade 2 stocks each, and the other two systems trade only QQQ. One system is long term trend following, one is based on a specific pattern, one is intraday trend following, and the last one is a mix of pattern and short term trend. I actually never analyzed the systems together as a portfolio, and probably that's what I should do, and see. I haven't changed the systems yet, since I thought one should be patient when trading, and not abandon the system right before it takes off again. But be patient for how long? 6 months seems too long. This is my first year as a strict system trader... I guess I will have to do something soon, if things don't turn around. Pat
Depending on the type of system, 6 months can be too long. It certainly would be too long for a volatility expansion system. About 2 - 3 consecutive months of bad performance is all it takes to recognize a system of that type is in trouble. A trend following system can chop for years. Support and resistance would be somewhere in between these two time intervals. An important point is all the system parameters must be in place and in agreement for a system to work reliably. David Stendahl has done excellent research in this area. Also, for trend following systems Charlie Wright has some great ideas. I was lucky enough to get his book and tapes free of charge a couple years ago when Omega Research was giving them away. They have not been available since then, and probably never will be available again. The 30% drawdown you mentioned would certainly challenge most traders. My systems have a very small percentage of drawdown. Some aggressive systems require a 70% or greater drawdown. I cannot imagine trading like that.
here is another way you could have optimized ... the selection of your two stocks. I generally want systems to work on most of the securities in a sector (e.g. SOX) to consider it viable/stable. It's easy for a system to work on one or two securities by chance. I generally like systems to work on at least SPY/QQQ to a limited extent before I consider it acceptable.