Your method only uses price and doesn't account for time (speed) and volume (speed too). Adapting the trail adjustment to incorporate these, would do more justice to the actual market movement. You're now trading of a line-chart, use timed- or volumebars instead. Ursa..
I do have tradestation, but this system was developed on excel and it uses too much data anyway to be backtested, at least as far as I'm capable of. I don't know how to create a trailing stop that's a function of volume and other variables. I think we're talking about too many variables here. I already have like 15 variables in the system, all things considered. I can't add any more.
Travis, I think the ideal system would incorporate different options for taking profits. A hard target based on support/resistance, or a hard target based on a known high probability profit amount, or an exit based on change of momentum signalling a reversal, or a trailing stop, or a time stop. The ideal system might incorporate all of these. As for the trailing stops, obviously you want a stop that allows for the normal noise in the market and lets you know when the price reversal is no longer "random noise" but a reversal that is likely to continue. I like to use a trailing stop system developed by Cynthia Kase, an energy trader who has written a book called "Trading with the Odds." She based the stop amounts on the volatility of the market as measured by the Average True Range--although she actually uses the ATR of the last 2 bars, not 1 bar. Using standard deviation calculations she determines different levels of stops (see attached article.) I use the different levels of stops based on my risk and time frame of a particular trade. Anyway, its better than any other trailing stop system I've seen, in my opinion.
article didn't attach. Trying again. BTW you can find the article and others for free on her web site. http://www.kaseco.com/support/articles.htm this one is called "Setting Stop Losses Using Price Volatility"
Thanks for the article and the advice, riskette. I know what you're talking about, when you say that an exit strategy has to incorporate as many things as possible, just like I would add an ideal system should be able to read the news and watch tv, and take everything into account. But unfortunately there's only so much an automated system can do without losing track of what it's doing and what is happening and why. So after having tried your approach, I decided that the most variables in my system exits I can afford are a stoploss, a take profit and a trailing stop, the last two being alternative to one another. Concerning the method for determing the right trailing amount I used Cynthia Kase's approach without even knowing it was her approach, or who she was. Yet, it's also too complicated, so I see many more advantages to keeping things simple and knowing what's happening, rather than having things change all the time because they are a function of something else, even though that would at the start seem the most scientific approach. In other words, 3 ticks are always 3 ticks. They are a very simple concept, and I have seen them in many situations, and I am totally familiar with 3 ticks, and I know what to expect of them. But a trailing stop based on ATR, I would have to spend a month just discovering at all its implications for my system, and I didn't find it to be worth it, actually I find the 3 ticks better.
Riskette, I read the Kase article, and found it very clear except for one point -- one that's probably clear to people stronger in math than I am. I didn't understand precisely why she focused on 1.,2, and 3 standard deviations. When would one be focused on which deviation the security was closest to? Is that an indirect proxy for trending power? Or what? Appreciate you contributing the article. 0steve
You might consider dumping about 12 of those & working on incorporating MajorUrsa's suggestions. With so many variables, you are highly unlikely to gain a real understanding of how your system actually works. Oh...you're screwed then. Nevermind. :eek: (However, if you keep working on this, you might work yourself out of the state of being screwed.)
Osteve, Different multiples of the standard deviation of volatility will give you different levels of confidence in how likely the stop level is signaling a real reversal. In my simplistic mind I think of it this way--if I use a stop based on 2 standard deviation multiple of the 2-bar ATR, then the price will stay within the stop 95% of the time if the trend is intact. If the price crosses the line then I am "95%" sure that the trend is ending. In a 3 standard deviation stop it would be 99.7%. It's easier to see on a chart (see attached). I just eyeball the chart of the market I am trading, and decide which stop to use based on my risk parameters and time frame of the trade. I will try to attach a chart of the Emini S&P from the start of the rally in July. You can see that using Dev3 kept you in until the end, but with higher risk than a closer stop.
"seems" is the correct terminology. Your trailing stops need to be set outside the recent noise even if it means giving the whole profit back.