Diversified Trend Trading Approach | Page 42 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Is Grid Trading (combined trend following and counter trend) Profitable? 3 replies

Making a good diversified portfolio 10 replies

Diversified EA Money Management Code 10 replies

HAMA PAD - A Simple Trading Approach 7 replies

Trading a (semi) automatic approach...but want more 11 replies

Attachments: Diversified Trend Trading Approach
Exit Attachments

Diversified Trend Trading Approach

  • #821
  • Edited 10:03pm Apr 10, 2016 9:05pm | Edited 10:03pm
Example 1A - Gold (XAUUSD)

Trade alert bringing me to the screen to ensure pending order activated and check pending order settings. Have a scratch and trawl through forum posts......pat the dog.........think about coffee and last night's roast after trade order executed. :-)

Pending Order Setup for Gold using line trader.

The red inclined dashed line represents a trailing stop mechanism with initial stop entry at 0.25% trade risk. The green line (superimposed on top of yellow predictive entry) represents the pending order entry.

Attached Image (click to enlarge)
Click to Enlarge Name: Capture.JPG Size: 121 KB


zzzzzzzzzzzzzzzzzzzzz
  • #822
  • Apr 10, 2016 10:00pm Apr 10, 2016 10:00pm
Quoting Copernicus
.........1. Putt a finger in the water with a trade at M30 using long breakout with low trade risk and be prepared to scale in by timeframe long if successful...............
wise
  • #823
  • Apr 10, 2016 10:03pm Apr 10, 2016 10:03pm
@copernicus you are real pro trader.ı will be your shadow and learn about statistics and other needful things :nerd:
  • #824
  • Apr 10, 2016 10:08pm Apr 10, 2016 10:08pm
Quoting Copernicus
PS I will follow this example through to the end for those who may be interested in this technique. It will demonstrate how I adjust pending order entries on higher time-frames to ensure that max break-even is achieved before further scaling into the trade. Stay tuned....if the M30 trade eventuates.
:nerd::nerd:
  • #825
  • Edited 10:28pm Apr 10, 2016 10:18pm | Edited 10:28pm
Example 1A - Gold (XAUUSD)


Quoting vasily
{quote} :nerd::nerd:
V. That red trailing stop line on M30 needs to intersect the order entry line (green) before I scale into my next trade. Worst case then on this trade if I opened up a new trade on say H1 would be break-even on M30. That's why my current predictive entry on H1 is about 1286.27 (but will be adjusted as time goes on to coincide with M30 positioning) and H4 predictive entry is far away at 1302.93. It allows me to catch fast momentum moves and also ensure that each timeframe has a substantive directional trend with momentum before entering.

Attached Image (click to enlarge)
Click to Enlarge Name: Capture.JPG Size: 275 KB
  • #826
  • Apr 10, 2016 10:21pm Apr 10, 2016 10:21pm
Quoting Copernicus
{quote} V. That red trailing stop line on M30 needs to intersect the order entry line (green) before I scale into my next trade. Worst case then on this trade if I opened up a new trade on say H1 would be break-even on M30. That's why my current predictive entry on H1 is about 1286.27 (but will be adjusted as time goes on to coincide with M30 positioning) and H4 predictive entry is far away at 1302.93. It allows me to catch fast momentum moves and also ensure that each timeframe has a substantive directional trend with...
hat off
  • #827
  • Apr 10, 2016 11:17pm Apr 10, 2016 11:17pm
Leave trade precision to the experts. :-)

Inserted Video
  • #828
  • Apr 10, 2016 11:36pm Apr 10, 2016 11:36pm
Example 1A - Gold (XAUUSD)

Pending Entry Update
Note how when I know a new high is being made I extend the SDC (1.0) immediately to the prior bar. This way I take as much trade history into the SDC sequence.

The entry order is about to be triggered and the trailing stop and 0.25% trade risk is perfectly aligned with the SDC lower channel. Retracement entry traders would be screaming to wait for the retracement....however breakout traders drown out the screams and plough on regardless. When put into context on the MN and W1 timeframes price is in an oversold zone so from a breakout perspective, we possibly still have lot's of upside. It may never retrace for an entry in these conditions so you are caught in a conundrum, do I wait....and possibly miss the trade altogether....or do I go? Do you trade or do you go now.......*there has gotta be a song in that* :-)



Attached Image (click to enlarge)
Click to Enlarge Name: Capture2.JPG Size: 110 KB



Inserted Video
  • #829
  • Edited 1:31am Apr 11, 2016 12:49am | Edited 1:31am
Example 1A - Gold (XAUUSD)

Still waiting for entry but we may find we have just reached a pivot on a lower order trend. The pending order remains in place and we simply wait to see what unfolds. We may find that we naturally enter at a later time when price moves to a higher high within the lower SDC (oversold sold)....who knows.

Note how we do not continue to extend the SDC until a higher high is made. The reason for this may become apparent later so we will wait and see what transpires.

I have drawn lower order trends within the dominant trend on M30 to display the fractal nature of wave dynamics using the SDC (zig-zags) with 1.0 sigma setting. This is an area I will look at more closely in the future in accordance with some of Bill Dreiss's rules (*hint*....have a listen if you haven't already).

Attached Image (click to enlarge)
Click to Enlarge Name: Waves within waves - Lower order trends within trends.JPG Size: 155 KB


Notes: What causes this fractal market structure? When breaking down moves between pivot points, there appears to be repeatable structure...... What we don't see in the chart above is the price action on lower timeframes which if you have a squiz is far less regular in nature....however the superposition of participant behaviour of the lower timeframes sees price action 'interference' which suppresses the irregularity of destructive interference and magnifies the correlated movements (momentum persistence) to result in higher order co-ordinated 'wave-like' momentum moves which give clues to coordinated participant behaviour over time and how longer term trends unfold.

What we are observing are the thousands of participant interactions with the market over time, which starts off as 'noise' and progressively become ordered in structure as a slight bias in the net price activity takes shape on the chart. The noise is shaped over time into an ordered structured price move by virtue of destructive and constructive interference. Throw pebbles into a pool at different frequencies to see how this might unfold.
  • #830
  • Apr 11, 2016 3:09am Apr 11, 2016 3:09am
Example 1A - Gold (XAUUSD)


We're long with 0.15 Lots at 0.25% risk to initial stop. Let's see how this plays out. I will keep you updated. Note how every bar, the SDC trailing stop reduces the total possible loss in accordance with current momentum.

Attached Image (click to enlarge)
Click to Enlarge Name: Capture.JPG Size: 154 KB


Multi-timeframe Context of Trade on M30
Attached Image (click to enlarge)
Click to Enlarge Name: Multitimeframe context.JPG Size: 272 KB
  • #831
  • Apr 11, 2016 3:26am Apr 11, 2016 3:26am
Quoting Copernicus
Example 1A - Gold (XAUUSD) We're long with 0.15 Lots at 0.25% risk to initial stop. Let's see how this plays out. I will keep you updated. Note how every bar, the SDC trailing stop reduces the total possible loss in accordance with current momentum. {image} Multi-timeframe Context of Trade on M30 {image}
advatage of trading slope.slope is difference betwen gambling and trading
  • #832
  • Edited 4:58am Apr 11, 2016 3:31am | Edited 4:58am
Example 1A - Gold (XAUUSD)

Quoting vasily
{quote} advatage of trading slope.slope is difference betwen gambling and trading
The benefit of using the SDC or your linear regression tool is that it detects momentum changes and changes slope accordingly. If we get a nice push up in gold, the SDC will ratchet upwards thereby getting us to breakeven sooner. If momentum becomes more subdued, provided we reach new highs the slope will decrease in accordance with the data and allow for more room to breath in-trend. The good thing about this approach is that you are managing the trade at all times using these statistical tools.
  • #833
  • Apr 11, 2016 3:34am Apr 11, 2016 3:34am
Quoting Copernicus
{quote} The benefit of using the SDC or your linear regression tool is that it detects momentum changes and changes slope accordingly. If we get a nice push up in gold, the SDC will ratchet upwards thereby getting us to breakeven sooner. If momentum becomes more subdued, provided we reach new highs the slope will decrease in accordance with the data and allow for more room to breath in-trend. The good thing about this approach is that you are managing the trade at all times using these statistical tools.
this is what ı am willing to learn from you.incredible statistic work you have
  • #834
  • Apr 11, 2016 5:15am Apr 11, 2016 5:15am
Example 1A - Gold (XAUUSD)

Update.
Note the status of the trade below.

Attached Image (click to enlarge)
Click to Enlarge Name: Capture3.JPG Size: 134 KB


The SDC has not been updated since the last high reached. What this does is force a time-based stop on the trade in accordance with historic price data over the trend sequence. Note the area highlighted in blue. Progressively the trailing stop moves towards breakeven reducing our total loss. The rate at which this occurs is in accordance with historic momentum of the price series.

We know that most of the time we lose in our trade as our historical trades tell us that about 60% of the time using this method we lose.....so our default position before we ever enter a trade is that it will be a loser. What we really care about in the back of our mind is our risk management and not whether we win or lose. We have confidence in our historic stats and recognise that when the trends come....we clean up.

The SDC has shown us what historic momentum exists in the price series by virtue of its slope and the width of the SDC has also told us what the volatility over the price series has been within 1.0 sigma......so we take heed to what the stats say and do not attempt to override them with our gut judgement. We therefore use a trailing stop that reflects historic momentum and volatility in trend, and this applies a time duration for events to unfold in. If we get stopped out, this is saying that the historic stats are no longer in force. It is not a subjective judgement but a statistical one. It is subtle, but it is important.

Already our max loss on this trade is <0.25% risk. It is a pin prick in the context of our historic trade performance. Furthermore we know that we already have built up significant unrealised profits in our other trades across the portfolio (eg. below) as we are diversified that reduce any anxiety associated with a single trade.

Attached Image (click to enlarge)
Click to Enlarge Name: Capture4.JPG Size: 284 KB


Keep singing the mantra of Ed Seykota to remember what this game is all about...."one good trade...pays for em all".

C
  • #835
  • Apr 11, 2016 6:50am Apr 11, 2016 6:50am
A brief excerpt from a Kevin Davey Email..quite incredible..


" Quantopian is a backtesting platform, where users can upload strategies. So far, over 3,000,000 strategies have been uploaded. Of that large number, Quantopian has determined that only 8 are worth live trading! Wow!! My Strategy Factory process has a lot of fallout, but not that poor. I suspect many "professional" strategies were developed incorrectly..."
Memories caught in time but never forgotten
  • #836
  • Edited 8:29am Apr 11, 2016 7:39am | Edited 8:29am
Quoting j4d
A brief excerpt from a Kevin Davey Email..quite incredible.. " Quantopian is a backtesting platform, where users can upload strategies. So far, over 3,000,000 strategies have been uploaded. Of that large number, Quantopian has determined that only 8 are worth live trading! Wow!! My Strategy Factory process has a lot of fallout, but not that poor. I suspect many "professional" strategies were developed incorrectly..."
"Wow" is the word Ja. Comprehensive robustness testing seems to be the order of the day.

When you hear an utterance on this forum that 'every strategy works', you can see how that leads to trouble. The bottom line is that very few strategies work as traders simply do not know what is required to determine if their system actually has an edge or not. So many are caught up in the noise and the simple 'apparent' truth of a few consecutive wins, or few months-years of consecutive wins. An unsinkable ship....now where have I heard that before :-)

The chart below is an example of the return profiles of just 4 traders results each with a 50% win rate and a win$ to Loss$ size of $1. Two different traders can trade the exact same system in the exact same markets at different time periods and achieve these widely dispersed results outcome....simply due to uncertainty that exists between a finite number of trades. Lesson.....learn to love losses and back-test like crazy to ensure your system has legs to cope with crisis events (consecutive strings of losses) over an extended timeframe as 'uncertainty' alone can nail you and leads to unpredictable return profiles.

Use this formula then double it to account for fat tailed distributions to ensure you stand the test of time with respect to consecutive losses and adverse events.

(Ln(50000)/-Ln(1-Pwin)

Example. Pwin of 49.3% = 16 consecutive losses in a row for 50,000 trades = 32 consecutive losses for conservative measure to account for non-linear auto-correlated power laws. What's 32 consecutive losses for a max trade risk of 0.25%..... = 32 x 0.25 = 8%. Well we can handle that. What's 32 consecutive losses for a trade risk of 2% per trade = 32x 2 = 64%......Ouch!!!!!!!!...or about half ouch if you base your trade risk on available capital. Not a good hole to be in.

Attached Image


Inserted Video
  • #837
  • Apr 11, 2016 9:13am Apr 11, 2016 9:13am
Example 1A - Gold (XAUUSD)

Update - The Momentum Ratchet

Momentum has started to pick up long. Notice how we redraw the SDC to the highest high immediately prior to the current bar. Note also how the entire SDC is ratcheting upwards to reflect the increased momentum bias long as new data is added to the SDC data set. We are accelerating now towards breakeven. You gotta love this statistical tool. This is a principle reason why you only draw your SDC to higher highs or lower lows. The only time you change your trailing stop is when momentum confirms the ability to accelerate or decelerate the trail. Use statistics to achieve this outcome, not a token previous bar low or fictitious support that in the scheme of the distribution of a data set has no intrinsic meaning and could just be 'illusory noise'.

Attached Image (click to enlarge)
Click to Enlarge Name: Momentum ratchet.JPG Size: 155 KB
  • #838
  • Apr 11, 2016 11:48am Apr 11, 2016 11:48am
Quoting Copernicus
{quote} "Wow" is the word Ja. Comprehensive robustness testing seems to be the order of the day.
A quote I read recently from a good trader about back-testing a decent strategy comes to mind " I try everything I can to break it" ..

Even if the numbers are a bit off-side at Quantopian with professional strategies being developed incorrectly, it pretty much spells it out, if your not willing to put in the effort and time and grind out the results using the correct procedure to form a positive outcome, moving forward will be almost impossible..

Nice reply C..

Edit: cool tunes..
Memories caught in time but never forgotten
  • #839
  • Apr 11, 2016 5:46pm Apr 11, 2016 5:46pm
Quoting j4d
{quote} A quote I read recently from a good trader about back-testing a decent strategy comes to mind " I try everything I can to break it" .. Even if the numbers are a bit off-side at Quantopian with professional strategies being developed incorrectly, it pretty much spells it out, if your not willing to put in the effort and time and grind out the results using the correct procedure to form a positive outcome, moving forward will be almost impossible.. Nice reply C.. Edit: cool tunes..
j4d, since you mentioned Quantopian I wanted to cite an interesting paper just published by some of its principals that approaches the large number of strategies backtested by users and the subsequent out of sample performance of those strategies. The paper essentially asks whether it is possible to determine in advance which strategies will perform well and which will not, and which performance features may be predictive and which may not have much predictive power. It takes a bit of a different angle on the subject but is an interesting read.

All that Glitters Is Not Gold: Comparing Backtest and Out-of-Sample Performance on a Large Cohort of Trading Algorithms
  • #840
  • Apr 11, 2016 8:02pm Apr 11, 2016 8:02pm
Quoting FXEZ
{quote} j4d, since you mentioned Quantopian I wanted to cite an interesting paper just published by some of its principals that approaches the large number of strategies backtested by users and the subsequent out of sample performance of those strategies. The paper essentially asks whether it is possible to determine in advance which strategies will perform well and which will not, and which performance features may be predictive and which may not have much predictive power. It takes a bit of a different angle on the subject but is an interesting...
Cheers FXEZ. Looking forward to getting into this in more detail. :-)
0 traders viewing now