| Welcome Guest, please sign in to participate in a discussion. | Search | Active Topics | |
Registered User
Joined: 12/31/2004
Posts: 163
|
Anyone,
I know that Time Frames for Indicators, such as a moving average, are subjective.
I've been wondering if most stocks may have a natural 20 day cycle.
What do you think?
Perhaps my question should be, "When you select a time Frame for an Indicator, what goes through your mind in making the decision to use a particular Time Frame?"
Bill
|
|
Registered User
Joined: 9/2/2006
Posts: 10
|
monthly or weekly I use a 200 and 50 ema.
any shorter timeframe I use a 20,10,or 5
Hopefully you are looking at all time frames to see how the stock is acting. If the stock is in downtrend on a daily but you then look at a weekly and the stock is trending upward. I would hesitate to short it. It is probably just pulling back and might be a good time to go long. You need to figure out what time frame you are going to trade in and use those charts to go in and out of a stock. Use longer and shorter times to figure out if you want to be in and if you do when to enter.
Hope this help you out.
Brady
|
|
Registered User
Joined: 9/2/2006
Posts: 10
|
Also for swing or intermediate trades I have found that the TRIX indicator along with MS and TSV works great.
You can find more information about TRIX in the support section. It is basicly 3 moving averages with a signal line.
Brady
|
|
Registered User
Joined: 1/28/2005
Posts: 6,049
|
I've heard that there was thought to be a 28 day cycle in the market.
(thats why all "standard" indicator values start at 14)
In general you want to find "robust" indicator values.
(values that tend to work in most situations not
necessarily perfect)
The indicator value you use has a lot to do with your trading
style. In the case of a dual moving average cross:
Short-Term trader: 3 and 11
Intermediate-Term: 10 and 40
Long-Term: 21 and 100
The rule of thumb is the shorter the period. The more trading you
will do.
thanks
diceman
|
|
Registered User
Joined: 12/31/2004
Posts: 163
|
Excellent suggestions!
Thak you both.
Bill
|
|
Registered User
Joined: 9/14/2006
Posts: 11
|
On the subject of indicators for trend following in the book "Bollinger on Bollinger Bands", he uses "an old rule" that says an indicator length should be approximatly one half of the calculation period for the bands. He doesn't know the orign of the rule but suggests "...it is likely an adaptation of a rule from cycle analysis that suggests using moving averages a quarter the length of the dominant cycle." But he also indicates that the proper length may vary depending upon the characteristics of the vehicle being traded.
|
|
Registered User
Joined: 12/8/2004
Posts: 213
|
Dr. Alexander Elder in his "Elder Ray" uses a 13 day moving average.
|
|
Registered User
Joined: 12/31/2004
Posts: 163
|
Hi,
I asked the question about time frames, earlier, in an attempt to get ideas from others about how they measure a fluctuations in the price of a stock. I know there is a wide divergence in opinions regarding which length is best.
I had been thinking that there was a 20 period cycle, which I referred to as natural when I really meant an imposed cycle by participants (brokers, analysts, investors, etc.) of the markets. There is a cycle associated with a 3 month earnings reporting period. I thought there may be a sub cycle of this of 20 days, the approximate number of trading days in a month. Perhaps there is, but it isn't easy to use it to plot, analyze or project behavior. Nor, for that matter, is a 3 month earnings cycle usable.
I've also been thinking about the natural cycle illuminated by fibonacci.
However, none of these views is wholly correct.
The cyclical nature of a stock is much more complicated that just 3, 8, 13, 21, 34 .... ?
A cycle of a stock is partially identified by Fibonacci observed natural numeric progression, reporting requirements, analysts statements, investor expectations, lender requirements, and much more.
A multi-variable formula, like those used to predict weather, would be required to come close to understanding a stocks behavior. I suspect that very very very few people on this planet have an understanding of how to define the relationships needed to create such a predictive formula.
What the majority of us are left with is using arbitrary analysis criteria, such as 3, 8, 13, 21, 34 ....? time frames.
I want to thank each of you who have responded to my question. You've helped me think about analysis more clearly.
Bill
PS. If you have more inputs for me, please post them!
|
|
|
Guest-1 |