Menu
Sister Sites

Options

Bookmark Thread

First Page First Unread Last Page Last Post

Print Thread

Similar Threads

How to properly use Technical Analysis 137 replies

What indicators do Hedge Funds use or do they even use one? 12 replies

Do you backtest your systems PROPERLY? 0 replies

Custom Indicator not Drawing Properly? 0 replies

Attachments: How to properly use indicators
Exit Attachments

How to properly use indicators

Quoting AntiCre
{quote} Exactly. We can either interpret it as boundaries or we look at it in terms of momentum. You can see that recently prize didn't leave its upper prize range. Thus, prize went up due to your 151-momentum(*). Now use another momentum, compare it with the picture above and tell me where you could have entered and why. I think there are two possibilities: Enter after a new trend started when both (or more) momenta lines break out of a predefined level. Re-enter if long-term momentum is still high but low-term momentum became weaker and now strenghtens...
Didn't Turtles use 5 and 10 periods for Donchian Channels ?
This allowed them to see a shorter and longer momentum.
On breakouts the shorted and longer donchians merged together. On consolidations they diverged.

Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot_20171101-160737.jpg Size: 844 KB
The truth is hidden from you
Quoting MaxenshteinD
{quote} But where will you put the line and for how long?
Anywhere it doesn't matter. And you can move it whenever you like, once a day or after each trade place it at the current price and go from there. This is all stuff cp has said. If you want to try to "fix" the problem the stochastic or williams R has by averaging the range to smooth it out then you just end up i with an envelope. I've been through all this: identify some issue with an indicator so you modify it and you just end up with something that produces results basically the same as another simpler indicator.

EDIT: Also, another issue is these indicators that take range from the past for reference is that they lag. So you reduce the lag and you end up with something that looks exactly like what is on the chart. So what use is it?
Interesting thread.
I would make one (sorry if it is too obvious!) point about the price chart itself.
For visibility of this, it is good to use the line chart with added SMA(period=1, price=high), and SMA(period=1, price=low). Your moving averages then just mark High and Low of any particular candle.

1. Price moves in a uptrend typically such that its Close is near the High.
2. Price moves in a downtrend typically such that its Close is near the Low.
3. If it is the opposite, this means selling during an up move or buying during the down move. Expect the following move in the opposite direction to be stronger.
4. In order for price to change direction, a period of elevated spread (=High-Low) is necessary. The price acts like a slingshot in those areas: if the Low is more distant from the Close than the High (and the price is at the end of a significant down move), the "price slingshot" is shooting up which can only be stopped by an eventual price slingshot shooting down.

Cheers,
k

Attached Image (click to enlarge)
Click to Enlarge Name: Clipboard01.jpg Size: 201 KB
Quoting MaxenshteinD
{quote} Didn't Turtles use 5 and 10 periods for Donchian Channels ? This allowed them to see a shorter and longer momentum. On breakouts the shorted and longer donchians merged together. On consolidations they diverged.
Never looked at it. But it sounds interesting. Do you have a link?
CrucialPoint: "I'm alpha". AntiCre: "I'm your Omega - the end."
Quoting kprsa
I would make one (sorry if it is too obvious!) point about the price chart itself.
Donchian channel for example inherits this property of the price movement, notice the "spread widening" and "slingshots" at the end of every directional price move.

Cheers,
k

Attached Image (click to enlarge)
Click to Enlarge Name: Clipboard01.jpg Size: 202 KB
Quoting MaxenshteinD
{quote} Didn't Turtles use 5 and 10 periods for Donchian Channels ? This allowed them to see a shorter and longer momentum. On breakouts the shorted and longer donchians merged together. On consolidations they diverged. {image}
Yeah that's all great but what do you really think is happening when two different period Donchain channels converge or diverge? Because it's nothing magical its just exactly what it is. The range over the past 5 bars is the same/different than the range over the past 10 bars. I think if anyone is going to even have an iota of a chance to be in the 5% they need to just quickly strip away all the fantasy and identify what they are actually doing versus what they think they are doing.
I don't want to hijack your thread AntiCre so I apologize but it seems the regulars always converge in the same threads so why not throw a few ideas out there since we are all looking for the same thing: that real true edge, no fantasy.

So it appears to me that there are only a few things that can be derived from price:

Volatility
Momentum/Speed/Acceleration
Relative position
Trend

Okay so essentially a successful strategy must incorporate all of these. Be advanced in it's consideration but not be a contrived static approach. For example when one indicator is above x and another between y and z and you look out your window and see a bird that means buy. This is horse shit. Every x crosses y or something goes above a level system is garbage. The market is efficient enough not to allow systems like this to profit. The real issue is if you can measure these things in a logical and practical way how do you develop a method from them without creating arbitrary rules. Yes testing etc but back testing is essentially if I do this then how much money do I make? Ok didn't make money so tweak the rules. Now test again over and over. Maybe you will have good results but how will you be sure that it is because this rule set/method is resulting in a real edge and not appearing profitable because what occurred did occur. Which might not have occurred but chance would have it it did. Forward testing? same deal how long will you test for? a month a year? what happened in that period might not have happened. How to produce a method that is true?
Quoting AntiCre
A while ago I've studied many different indicators in order to understand what they calculate. This study revelaed that it is quite obvious why not every return from the so-called "overbought" or "oversold" indicator region is followed by a new trend. For instance, the stochastics oscillator with a given period %k just shows where the current prize is in relation to the lowest and highest prizes within the last %k candles...
You're thinking much too hard about this. Every indicator works, whether it's Stochastics or Bollinger Bands or the nearly-extinct PSAR. Aside from academic exercise it doesn't matter 'why' it works, just pick an indicator and study the historic relationship between it and price/prize until you think you know it well enough to use. There are times for any indicator when it works and when it doesn't, but it takes only some time -maybe a few months- to learn.
An indicator must show something concrete.

99.99% of all indicators fail here.

It's a waste of time to look at them.

The only one which i'm using is the %-change, because it's universal.

In my early days i watched the sessions (Tokyo, London and NY) a lot, but don't need that anymore.
Attached Image
Quoting VincentGallo
why not throw a few ideas out there
Ok, here you go (although I've already repeatedly mentioned it. While studying the charts with CPs influx indicator where he marked entries/exits, I made the this observations:

Whenever the longer-term momentum breaks out together with the short-term momentum and both momenta are above its average, then there seems to be a good chance to obtain some further prize movements in this direction.

If the short-term momentum retraces while the long-term momentum is still outside the "ranging" boundaries, there is a good chance to see a continuation of a trend if the short-term momentum changes its direction.

I've marked both opportunities in the image below with blue circles. However, the red circle is not an opportunity because the long-term momentum is below its average.
Attached Image
CrucialPoint: "I'm alpha". AntiCre: "I'm your Omega - the end."
Quoting Rob Mondave
{quote}just pick an indicator and study the historic relationship between it and price/prize until you think you know it well enough to use. There are times for any indicator when it works and when it doesn't, but it takes only some time -maybe a few months- to learn.
Maybe you could give an example ...
CrucialPoint: "I'm alpha". AntiCre: "I'm your Omega - the end."
Quoting The-Flipper
An indicator must show something concrete. 99.99% of all indicators fail here. It's a waste of time to look at them. The only one which i'm using is the %-change, because it's universal. In my early days i watched the sessions (Tokyo, London and NY) a lot, but don't need that anymore. {image}
I'm sure that there are other ways to succeed. But I think you will also agree that there are some traders out there that are successful using indicators.
CrucialPoint: "I'm alpha". AntiCre: "I'm your Omega - the end."
Quoting AntiCre
{quote} I'm sure that there are other ways to succeed. But I think you will also agree that there are some traders out there that are successful using indicators.
Yes, it's possible.
Quoting AntiCre
{quote} Maybe you could give an example ...
Example of what, when an indicator works and when it doesn't? Just put an indicator on your chart and study the times it gave a good signal and a bad signal, you'll find consistencies. OK, for example, Stochastics signals are generally not reliable in consolidation and for counter-trends, but are more reliable at extremes in ranging markets and in trend resumptions after pullbacks. It can be more nuanced which is why it's helpful to study.
  • Post #35
  • Quote
  • Edited at 12:38pm Nov 1, 2017 12:27pm | Edited at 12:38pm
Quoting AntiCre
{quote} Ok, here you go (although I've already repeatedly mentioned it. While studying the charts with CPs influx indicator where he marked entries/exits, I made the this observations: Whenever the longer-term momentum breaks out together with the short-term momentum and both momenta are above its average, then there seems to be a good chance to obtain some further prize movements in this direction. If the short-term momentum retraces while the long-term momentum is still outside the "ranging" boundaries, there is a good chance to see a continuation...
Yeah combining short and long term momentum is key but I think the approach of "Whenever the longer-term momentum breaks out together with the short-term momentum and both momenta are above its average" is too static. Also I know the screen shots which you are referencing that CP posted and I think it's worth noting that in some he gave examples of cross buy cross sell etc. but that's not what all of his trades look like. I know he has those indicator levels of 7, 16, 50, 75 etc but they are in my opinion not used always used in this fashion. What is clear is that his influx indicator shows volatility, momentum and trend all in one. Also a tip if you are focused on influx, his screen shots from more recent years the influx indicator is different. I don't just mean the colors or levels it is more advanced and gives more information. And if you ever have tried to produce something like his influx indicator you will know there are a bunch of calculations all with different settings that can look very very similar but are not the same. The zero lag MACD is close but it is not correct. And then if you do get it you will realize CP is not kidding when he says nobody can reverse engineer his system. Especially not thinking in a way that people approach any other system x crosses y buy/sell etc.
Quoting VincentGallo
This is why something based on a moving average is probably more practical than something that uses total range over N bars. Because as the window of reference shifts you can have a significantly different reference range from the window shifting along a single bar, which can have an effect on the current output. Not really because of what is happening now but what happened in the past. This decreases the value of such indicators in my mind. Though something that doesn't slide along and keeps a static reference point does not have this problem....
reduced to a line ok but how to go about trading it ?you mentioned 2 options in another thread. Tight sl and placing a line where there is lot of volatility.
tight sl could still fail due to being whipsawed to death by price and the problem with volatility is being able to predict a place where there will be lot of volatility in the first place which aint easy. Do you trade this way?of course if you dont mind me asking. Also instead of a single line, what do you think of a grid? thanks.
I think the logic is not that bad ... However - just my interpretation.
Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot1.png Size: 21 KB
CrucialPoint: "I'm alpha". AntiCre: "I'm your Omega - the end."
How to properly use indicators - don't. Delete them from your trading screen because they will drain your account. Indicators are derived from prices, so they are always, by definition, lagging. My experience is that Forex is a losing game, one might as well go to Vegas - at least Vegas offers drinks and shows as you lose your money. But if you really insist on trading, look at recent highs and lows and target the stops that will be just above/below . The "big boys" hunt stops - might as well join them.
how to use indicators? Study math. Know the formula behind the indicator and what it measures or calculate.
Pecuniae obediunt omnia
Quoting VEEFX
{quote} Show us this 'formula'... or any formula which is dependent on two variables Price and Time that will result in a constant outcome i.e. positive pips. It is not possible ! It is ONLY possible if price and time are treated as constants within your own interpretation of market structure.
Thats the misconception about indicators. Its is not made to give you a positive pip, it was made to calculate/ simplify data faster. You are then using it worng
Pecuniae obediunt omnia
0 traders viewing now