Technical Analysis alone will not give you an edge | Page 7 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

"You cannot make money (in the long run) from a price chart alone" 37 replies

An Edge for an Edge (System Development) 116 replies

Does an EA license Give You the Right to Give Signals? 2 replies

Give me premise or give me death.... 15 replies

How certain can you be that you have an edge? 79 replies

Attachments: Technical Analysis alone will not give you an edge
Exit Attachments

Technical Analysis alone will not give you an edge

  • #121
  • Nov 1, 2015 8:34pm Nov 1, 2015 8:34pm
Quoting the redlion
{quote} Thanks for the Link, since you and I have similar views on this subject this is more akin what I tend to use nowadays I look at the big fundamental picture starting with the macroeconomic environment, then I look at both Central banks and their policies as well as their hints, I usually read different opinion's from economist who know more about the subject than me and I analyse those views to help me make a more informed decision medium term strategy. then I ascertain the drift through statistical analysis {image} you see in the next month...
Thanks for details of your methodology. Interesting approach. :-)
  • #122
  • Nov 1, 2015 8:38pm Nov 1, 2015 8:38pm
[quote=FXEZ;8569771 A couple of quick points: 1) It isn't possible to learn enough just from comparing distributions (cdfs) to determine anything with regards to random/nonrandom. 2) You highlighted the obvious difference in kurtosis between the price return data and artificial data but didn't link this to trends in the market data that aren't present in the randomly generated data. Fat tails are evidence of trends. 3) When calculating acf on the whole period you are making an implicit assumption: that correlations are constant/persistent through time. What if this isn't true? [/quote]

Nice FXEZ. These points you raise are actually the fundamental assumptions applied to my methodology in generating alpha.
  • #123
  • Nov 1, 2015 8:44pm Nov 1, 2015 8:44pm
Quoting FXEZ
{quote} the redlion, A couple of quick points: 1) It isn't possible to learn enough just from comparing distributions (cdfs) to determine anything with regards to random/nonrandom. 2) You highlighted the obvious difference in kurtosis between the price return data and artificial data but didn't link this to trends in the market data that aren't present in the randomly generated data. Fat tails are evidence of trends. 3) When calculating acf on the whole period you are making an implicit assumption: that correlations are constant/persistent through...
wow great post!!!

you know when I took the hurst exponent of data going back to inception H>60 at the time although some other estimates were putting it at 80 and some at mid 50's i though perhaps it shows non randomness through persistence in the monthly data.

if you are doing a moving ACF at it is breaching confidence levels ...that is further evidence of persistence hence my pre supposition of drift as the year drift seems to be consistently working for me.

what is your sample size for the moving ACF? are you doing 22 points 264 points ?
AVT INVENIAM VIAM AVT FACIAM
  • #124
  • Nov 1, 2015 8:46pm Nov 1, 2015 8:46pm
Quoting the redlion
{quote} wow great post!!! you know when I took the hurst exponent of data going back to inception H>60 at the time although some other estimates were putting it at 80 and some at mid 50's i though perhaps it shows non randomness through persistence in the monthly data. if you are doing a moving ACF at it is breaching confidence levels ...that is further evidence of persistence hence my pre supposition of drift as the year drift seems to be consistently working for me. what is your sample size for the moving ACF? are you doing 22 points 264 points...
I replied too quickly I see you are using 1000 data points about 4 years of data with a moving acf of 10 .... I will try to replicate your results...

WAIT that is 10,000 data point 37 years unless you are using minute, hourly ?
AVT INVENIAM VIAM AVT FACIAM
  • #125
  • Edited 10:34pm Nov 1, 2015 9:01pm | Edited 10:34pm
Quoting the redlion
{quote} I replied too quickly I see you are using 1000 data points about 4 years of data with a moving acf of 10 .... I will try to replicate your results... WAIT that is 10,000 data point 37 years unless you are using minute, hourly ?
Hi, I used EURUSD 15 minute data so about 104 days worth with ~10K bars with a moving window of 100 bars for this example. I used the short period because computation time was a factor and I just wanted a quick test of the concept. Start and end dates for this test were:

(YYYY-MM-DD)
"2011-12-11 05:15:00 EST"
"2012-05-07 11:45:43 EDT"

It may not be convenient (or necessary) to replicate this time period exactly or this time-frame. I think this just shows that there may be some temporary dependencies in the data. Can they be exploited? I don't know (yet). By the way, you brought up Hurst. I tried the moving window approach with Hurst exponent and was less satisfied with the result than what is shown here with the acf with this data.
  • #126
  • Nov 1, 2015 9:04pm Nov 1, 2015 9:04pm
Quoting FXEZ
{quote} Hi, I used EURUSD 15 minute data so about 104 days worth with ~10K bars with a moving window of 100 bars for this example. I used the short period because computation time was a factor and I just wanted a quick test of the concept. Start and end dates for this test were: "2011-12-11 05:15:00 EST" "2012-05-07 11:45:43 EDT" It may not be convenient to replicate this time period exactly or this time-frame. I think this just shows that there may be some temporary dependencies in the data. Can they be exploited? I don't know (yet). By the way,...
are you using R cran for your stats?
mat lab?
AVT INVENIAM VIAM AVT FACIAM
  • #127
  • Nov 1, 2015 9:07pm Nov 1, 2015 9:07pm
Quoting dkrock
{quote} Seriously? Why are you interested in Forex? To validate it or to make money? You need to focus on trading forward, not backward. You are stuck in analysis paralysis and frankly I am starting to pity you. As I told you, precision kills. Your art projects are not making you any money. You are making it way too complicated, to the point of unsuccessful. It is not about statistical models, or support/resistance. It is about means and apexes from the mean. Up and down movement from the mean. Not what will the next candle be? LOL. How many guessing...
Those are some big assumptions you're making about me.

let's say I am picking a bollinger band strategy, make it simple right...... I could pick whatever I want...but am I justified in making the assumptions that such strategy actually yields results? I could grow equity through luck.... I am not interested in luck.

there are always dominated strategies and optimum strategies .... just because you believe something doesn't make it true. I cannot control your pity , that's fine with me, what is your suggestion? give me a concrete example on how I should be doing this?
AVT INVENIAM VIAM AVT FACIAM
  • #128
  • Nov 1, 2015 9:19pm Nov 1, 2015 9:19pm
Quoting dkrock
{quote} Pay attention hot shot. Bollinger bands are only effective if you know the market equilibrium. If you picking numbers out of the sky, they are worthless. Your lack of faith is disturbing to the force You seem to have all the data anyone could ever want. So find the market equilibrium and stop searching for highs and lows. Then use your vast math knowledge to track movement up and down from it. So much simpler.
correct me if I mis represent your views, you are saying there is an underlying secret formula that once I figure it out and compute the right settings...I will predict price.
because I will find the equilibrium, by that I take it is the mean of a non-stationary series (price series) regardless of macro economic events.

Forgive me if I don't buy it.

the market equilibrium of returns is 0 ---- the open price, can you exploit that?
AVT INVENIAM VIAM AVT FACIAM
  • #129
  • Nov 1, 2015 9:23pm Nov 1, 2015 9:23pm
Quoting the redlion
{quote} are you using R cran for your stats? mat lab?
Yes using R, acf function.
  • #130
  • Nov 1, 2015 10:54pm Nov 1, 2015 10:54pm
Quoting the redlion
{quote} correct me if I mis represent your views, you are saying there is an underlying secret formula that once I figure it out and compute the right settings...I will predict price. because I will find the equilibrium, by that I take it is the mean of a non-stationary series (price series) regardless of macro economic events. Forgive me if I don't buy it. the market equilibrium of returns is 0 ---- the open price, can you exploit that?
Appreciate you don't buy it, you shouldn't.....nobody should unless he/she is a dumbass and I believe we are not.

Having said that, few of us have found ways to exploit the price. 'How and show me' you ask if not you will not believe. You go through this mathematical process with this thread discussion to establish this for yourself, fully appreciate and understand. As to the 'how and show me' stuff it won't happen-stop dreaming, if you are on this side you won't too, I'm sure you agree with that. BUT the least we can tell you is YES it can be exploited and NO this is not the way to establish it. Spend time to find the 'how' and put that 'whatever' on your own charts then you will know for certain and not from some dumbos out there who try to convince you his 'believe'. I have given some, ask the 'right' questions and you will receive the answers you are looking for - no answer/response means not sharing.
  • #131
  • Nov 1, 2015 11:38pm Nov 1, 2015 11:38pm
Quoting the redlion
Here are some charts...do you think you can trade them? do you see the patterns? {image} {image} {image} {image} {image} {image} {image} {image}
The trouble I have found with this analogy over the years redlion is that price does not just move itself. People move price. So while the patterns are the same, the reasons for those patterns forming are different. No one knows when these patterns are forming, but when they are in progress, what the people moving price are doing is evident in price and volume, and occasionally in time.

That's the difference none of you random walkers ever want to address. It would be nice to start a discussion on that. Psychology is what drives markets. What people are thinking is why that mouse clicks buy or sell.
If you are good at something, never do it for free--Joker
  • #132
  • Nov 1, 2015 11:48pm Nov 1, 2015 11:48pm
How? Read this - http://www.forexfactory.com/showthre...76#post8569676

He has given some, it's so easy that I'm on forum typing this while trading live at the same time
  • #133
  • Nov 1, 2015 11:52pm Nov 1, 2015 11:52pm
Quoting jmn5611
{quote} The trouble I have found with this analogy over the years redlion is that price does not just move itself. People move price. So while the patterns are the same, the reasons for those patterns forming are different. No one knows when these patterns are forming, but when they are in progress, what the people moving price are doing is evident in price and volume, and occasionally in time. That's the difference none of you random walkers ever want to address. It would be nice to start a discussion on that. Psychology is what drives markets....
This is true... random, generated patterns are just that. Charts of price movement based on random price data... in forex there are none, because price is always driven with intent. It is driven by market makers, brokers and dealers. Retail traders do not drive price. Price is influenced by retail traders as they themselves gather at these "congestion" areas, ( common area of interest), and retail sentiment helps with the decision as to where market makers drive price.

Everybody says that no bank is going to stop hunt you or me... while true the majority of the time, they will hunt stops in an area of congestion where you, I and 50,000 other retail traders, have been drawn to by some lines on a chart... a chart that they created in the first place. It's like going to war and you have allowed your enemy to train you, your generals and your army, as well as lay out your battle plan against them.

These are the decisions that the various groups of market makers create during the "gap time" between trading sessions.... this is where the direction and the intent of price is determined. They are not flipping coins, they are plotting the course of price.

-CJ-
A man is not defeated when he is beaten, he is defeated when he quits.
  • #134
  • Nov 2, 2015 12:25am Nov 2, 2015 12:25am
CJ, do you refer to a similar pair for reference before taking a trade on all of your trades, or only your mapped pairs?
If you do check other pairs, could you share insight?
Thanks again

Quoting Captain Jack
{quote} These maps are based on pattern recognition alone... fundamentals take no part in it. What will occur is that the "news" this week, will drive price along that path, but in reality, the pattern drives the news.... Since I am trading it live I would request it be kept private for the time being. Once the bottoming structure is complete, the "W", you may do as you like with it. I could not prevent you from posting it now, but the charts I post are re-posted in many forums. I also know for a fact that my trades are mirrored in house at some...
  • #135
  • Nov 2, 2015 12:46am Nov 2, 2015 12:46am
Quoting forexdave1
CJ, do you refer to a similar pair for reference before taking a trade on all of your trades, or only your mapped pairs? If you do check other pairs, could you share insight? Thanks again {quote}
To tell you the truth, I very rarely look for them anymore. The one I'm trading is just a pattern I recognized from the past as it is distinct in nature and was easily spotted. Normally when I trade, I just try to trade MM patterns. Once you locate where price is in the market cycle for that pair, you can trade anywhere within the cycle. It's a pattern in itself, but much easier to find. I sent the map out to 2 people who showed interest in another thread so they could follow along to see if price would follow the map... the 2 people who received it are of opposite beliefs. As price travels the map, both will end up on the same side of belief or disbelief... it can be no other way.

I have always said that on each pair, the patterns in play are born on the tick chart. From there, the same tick chart pattern will repeat across EVERY time period for that pair, at varying times, based on the period. It can be within minutes on the lower time periods as well and many months and years on the higher periods. They mature across the periods at different rates. So the lower time period charts are leading indicators for the higher time period charts.

There are patterns that repeat across non correlated pairs at times, like the one I trade now. The source pattern is a D1 from several years back that is now being presented on a current chart, on a lesser time period. It is the same pattern, but it appears on 2 non correlated pairs.

For many years now, I have turned my efforts and time away from what all the professionals say, stopped reading any books, websites, forums, use of indicators, systems, ea's ...you name. I had to "unlearn" everything I ever read about the market, EXCEPT how those on the other side of the glass work their trade. I began looking at the patterns that the market makers leave behind, reviewing untold numbers of charts, looking for those tell tale tracks. It was while I was spending all this time looking at charts, that I realized many looked "familiar". Then I actively begin to search for like patterns and I found them.

It's not always easy and they are always trying to distort our vision of what we see... like when you scroll your chart history back and they change the scaling... so I started taking screen captures and have quite the library of them. It's where I pull a lot of the charts that I use for examples from. I also go back, when I think I have recognized one of these patterns and flip through old screen caps.... I got to the point that I can do this quickly, and it's easy to spot the match if it's there.

The are traps along the way that will swallow you whole as they like to thrown multiple copies of the same pattern into a pairs chart... if you choose the wrong one as a match to the other... bye bye account!

-CJ-
A man is not defeated when he is beaten, he is defeated when he quits.
  • #136
  • Edited 2:25am Nov 2, 2015 2:05am | Edited 2:25am
I will give a challenge for the math wizards here. You want to find a pattern to test and proof that the market is not random? OK. I will give you JUST one pattern.(there are thousands like this)
Search for this conditions:
Any major currency pair.
Technical conditions:
1. Flat(ranging market) or slow trend for the last 10-24 hours.
2. Intraday Volatility < Average
3. Daily range < Average

Fundamental conditions
1. Neutral or positive overall fundamental context
2. Major economic release(number) on the calendar with defined upper and lower expectations.

How to trade this pattern?
1. Wait for the news release.
2. On the release, IF the number is NOT outside of expectations, BUT the volatility (news spike) is above the average range(ATR) on 60 min, then trade against the spike. The direction is irrelevant. If the spike is up - go short. if the spike is down - Long!
For the algorithm - No fixed stop loss. Target 20 pips or close after 60 min. (whichever comes first.)
These are the overall parameters. If you can reduce this to something that can be coded and executed by algorithm -you will make profit at least 70% of the time. and your average loss should be around 25 pips. Average profit around 15 pips.
Overall positive expectancy.

You can also try the same pattern, but with different technical context.
IF Uptrend = true => then ONLY trade against down spike.
IF Downtrend = true => then ONLY trade against UP spike.
let's see what you can do with this.
  • #137
  • Nov 2, 2015 2:51am Nov 2, 2015 2:51am
lets do a jigsaw ,but first remove lots of pieces ,

Quoting the redlion
Here are some charts...do you think you can trade them? do you see the patterns? {image} {image} {image} {image} {image} {image} {image} {image}
On-site guru & also FF member's' psychiatrist, when not drinking tea
  • #138
  • Nov 2, 2015 3:22am Nov 2, 2015 3:22am
Wow nice thread guys I just spent an hour reading it from the start.... It seems most traders here are profitable... Can technical analysis only give an edge ... it appears that most of you have proven that it does.... However the thread leader appears to want to creak the big one.. hats off is worth a try... there are 3 ways to make profit 1 the big banks and market movers .. commit to a trend based on underlying fundamentals .... and the big money is made there.... but the banks do not win all the time and when they lose it is billions ... yes they move the market .. but only to a critical area stop hunt ... the reality is they stop hunt each other .... the retail trader is inconsequential .. if we are hit our stops are too small for the time frame. So the number 1 way is the holy grail .... as my broker says I love it when it goes sideways.... the accumulation phase.... we all know ..like a spring being wound up price will move sharply up or down ... can it be done volume spread analysis .. so they can .. I do not think so .. I think these guys are inside traders have a birdy giving them info ... follow futures ... there are only bids ... can be take of the table in a heart beat..... and may I said it looking at the pending orders .. order flow is very useful but not reliable .. this is not any tangible assets ... it's currency what moves this market is NOT trading decisions we are seeing the movement of countries stakes in other countries .... or big assets deals between the seller and the buyer.... if BHP want to take over some company say 20 billion they will not be hitting transfer funds 20 billion at a monthly high of the owner currency .. a payment plan if followed .. and of course the big multi national companies have a basket of assets and currencies as do the nations governments .... if you have an equal amount spread in all the majors it does not matter what happen to the home currency we are hedged.... so the man talking about equilibrium is correct ... at the end of the day who can said eg.. EUR/USD is over valued .. we all know there is no such thing the price NOW is the price.... question is do you HAVE transfer NOW or can you negotiate the purchase latter? the 2nd way is trending close to the 1st one market sentiment ... 3rd if over extension the pattern trader IS making a living out of this .... Fibonacci WORKS... but were you pull those fibs is very subjective .... if we use a microscope on the edge of a leaf maybe the pattern is not obvious looking at the leaf it is easy to see a maple from a spinach... the charts are price/time patterns and trend lines are time projections and they are accurately respected more than 50% of the time .. YES do you all agree .. that is an edge ... we are all seeking consistency.. we gotta get off the small time frames.... support and resistance is important the big boys run each others stops ... get a little MT4 indicator called periodcon it is free I am not sure were you get it my mentor gave it to me ... it multiplies the time frame .. put on a 1 hour chart you can make 3,6,8,12h charts as off line charts ... this enables me to draw S/R Areas on each and a quantitative approach is very useful and for looking for indecision bars... make you stops larger than your S/R areas add to that at the extension past 2 discernible dips ... so you are most probably in a 5 wave... so putting it all together .. impossible to predict when a major transfer of currency will happen ..I have seen a spike in the dead of night ....but greed and fear en-shore swings into over extension... equilibrium ... you can a free ABCD pattern indicator from PZ Trader.... sorry so long winded and thanks guys most excellent discussion ... I would love to know were the accumulation phase is heading .. that would be great...
  • #139
  • Nov 2, 2015 3:27am Nov 2, 2015 3:27am
Quoting the redlion
{quote} here is a simple analysis of a random chart.......you could clearly see support and resistance levels as well as price rejections around those levels. It is completely random, that means that the next candle has nothing to do with the previous ones no matter what pattern or levels it shows on the chart {image}
Morning Red. Of course, you would never trade these charts. There is no structure to them. Without structure, all TA is useless. Now, even with structure, there are times when PA behaves in a random manner, such as periods when price is near or at equilibrium. We go to great lengths to stay out of the market at those price levels, specifically because the narrative of the market has paused, or perhaps is repeating itself. So what your random charts lack is narrative, purpose, reason, and intent. Add an intelligent agent to your random charts, and they no longer are random, right?
  • #140
  • Nov 2, 2015 3:35am Nov 2, 2015 3:35am
I'm guessing you using previous patterns somehow



Quoting Captain Jack
{quote} An alternative question might be "Can you tell the predicted price movement from the real price movement?" The original chart was posted in 2012, here at FF. The results follow the predicted path of price movement proving that price can indeed be predicted, and not only predicted , predicted with great accuracy. -CJ- {image} {image} {image}
Attached Image (click to enlarge)
Click to Enlarge Name: eu USD (DFB).png Size: 22 KB
On-site guru & also FF member's' psychiatrist, when not drinking tea
1 trader viewing now