Diversified Trend Trading Approach | Page 169 | 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

  • #3,361
  • Edited 9:59pm May 3, 2018 8:44pm | Edited 9:59pm
Systematic Trading with Nick Radge

It's always good to get a good dose of reality from Nick. Note how he refers to the 'holy grail' being diversification across systems. A key takeout from this podcast is the need to understand when your system does and doesn't perform. As your skills develop in this area....that's where you can then deploy a diversified suite of fully automated systems that can stand the test of time (market uncertainty).
  • #3,362
  • May 3, 2018 10:06pm May 3, 2018 10:06pm
“The information we consume matters just as much as the food we put in our body. It affects our thinking, our behavior, how we understand our place in the world and how we understand others.” .....so it begs the question....what skills can be deployed to access the right information in a world suffering from data overload? Michael Simmons has a few ideas.

Attached Image
  • #3,363
  • May 3, 2018 10:40pm May 3, 2018 10:40pm
Quoting Copernicus
{quote} You are a legend and a scholar Mr B. I watched it through to the end. It took off when the instrument started to trend. Very interesting approach. Thanks very much for the lead. I have tracked it through to a Mr Steve Walker on twitter, Linked In and a blog. I can't seem to find an article on his approach yet but have lots of examples to interrogate......but it seems right down my alley. I will investigate further. :-) Cheers C EDIT: I have been able to get a bit more info from his approach and it is very interesting with lots of good ideas...
"In looking at this it has given me a great idea in relation to defining divergent and convergent zones which could be fully automated in an 'all-in-one' solution that spans the timeframes. A possible ideal solution under diversification to provide system diversification and timeframe diversification on the same instrument. I am going to get my teeth stuck into this. It also solves the problem under automation of how to define the swing highs and swing lows to commence the SDC from as the multi-timeframe approach solves this dilemma. Cheers Mr B"

Hey C,
If divergent=diversion from mean(breakouts) and convergent=reversion to mean(bounce then this is what my area of focus has been for past 3 yrs. I have a comprehensive strategy developed which (I think and hope) will prove the test of time for years to come in any combination of the 6 portfolios I have developed. Plan is to trim/consolidate it further but what I have works great if you only consider ROMAD as the most important trading stat.

Just this past weekend, I was thinking about reaching out to you (Mr. Guru of Diversification!) but found a bug in my drawdown control that had to be addressed.

Timeframe diversification or what I call Universal Time Action Strategy (different from Price Action or MTF approach) is far far superior imo if coupled or aligned with the underlying types of market conditions instead of types of systems or trading styles. "define the swing highs and swing lows" is not (or should not) be considered at all to avoid limitations. They are just imaginary obstacles to me as some will break and some will bounce resulting in a 50/50 spread. Unfortunately, I need a partner overseas (or eventually relocate) as this strategy fails with US brokers who enforce FIFO and hedging restrictions.

Quoting Copernicus
{quote} It's cool Mr B. This strategy is very short term with an average hold of about 3-4 days. The SWAPS on this strategy are not a biggie for me. This is good old mean reversion coming out to play that is chipping away at the results. The volatility we had been experiencing earlier in the week across markets was turned off last night....and C is suffering from the symptoms. R maybe a bit safer because of his use of a dynamic portfolio that has it's pluses and minuses....and in this instance probably a plus. It's all part of the strategy....you...
This happens every NFP week. Monday/tuesday is okay then comes wed/thurs and market pauses waiting for NFP to come out. Every month this year, I have experienced drawdown going into NFP kicking off my damn drawdown manager function and chipping away realized profits from me. There is no solution but to ride it out as I don't believe (or want to) trade based on market conditions or avoid news releases. Looks like you also need volatility. Me2 and I have a way of tackling it by dissecting the 24x5 clockwork function of the global currency market sessions. Some Win Some lose resulting in 50/50 :-
Staying in my lane...
  • #3,364
  • Edited May 4, 2018 5:13am May 3, 2018 11:47pm | Edited May 4, 2018 5:13am
Quoting VEEFX
Hey C, If divergent=diversion from mean(breakouts) and convergent=reversion to mean(bounce then this is what my area of focus has been for past 3 yrs. I have a comprehensive strategy developed which (I think and hope) will prove the test of time for years to come in any combination of the 6 portfolios I have developed. Plan is to trim/consolidate it further but what I have works great if you only consider ROMAD as the most important trading stat.
Yep mate. That is what I am meaning with divergent versus convergent. Sounds great V.....and I am doing the same thing.....however I am always biasing my result with divergence as it tends to be more sustainable long term. As a result....my primary strategy is always divergent in nature.....but I do get a little bit pregnant with convergence when it assists. The result is that return distributions of my portfolios are necessarily more volatile than pure convergent plays....however that is the nature of the game in the divergent space. What is clear to me is that volatile return profiles does not necessarily equate to risk. Each strategy possesses their own unique profile based on how they play the game. Some more volatile than others. The only way to reduce the volatility of the return series is through diversification....which you seem to have a handle on :-)

The problem with most convergent styles that attempt to catch the falling knife is that market risk is not totally released via the strategy and intrinsic risk builds in the market until that inevitable nasty and fast drawdown when markets diverge. I think Alphaomega is experiencing that symptom in his great convergent strategy......but I have yet to find a solution to this inherent risk that lies in the design of convergent systems. You can define some convergent styles with stops to manage the tail risk....however this almost always totally compromises the mean revergent nature of the strategy.

Quoting VEEFX
Just this past weekend, I was thinking about reaching out to you (Mr. Guru of Diversification!) but found a bug in my drawdown control that had to be addressed.
Always up for a chat V particularly in aspects of diversification. :-)


Quoting VEEFX
Timeframe diversification or what I call Universal Time Action Strategy (different from Price Action or MTF approach) is far far superior imo if coupled or aligned with the underlying types of market conditions instead of types of systems or trading styles. "define the swing highs and swing lows" is not (or should not) be considered at all to avoid limitations. They are just imaginary obstacles to me as some will break and some will bounce resulting in a 50/50 spread. Unfortunately, I need a partner overseas (or eventually relocate) as this strategy...
Nice mate. You obviously have thought about this a lot. There is definitely merit in avoiding the temptation of assuming that the mere selection of a range of arbitrary timeframes leads to low correlation........ however if you trade the short term timeframe but apply a filter to avoid the noise....some of the market structure features that traders and investors in longer term timeframes would act on seems to have more weight than others.....well that's the logic. It needs to be tested with the application of the correct assumptions. A test simply operating on a random selection of market features might not pick this nuance up. Yes the market is fractal like but you can bias your strategy in the short term to act on signals from the long term that might have more relevance than others in terms of short term momentum signatures. For example what is the level of signal to noise at short time segments of price data around these thresholds versus the signal to noise ratio of the same short time segments at other random selections?


Quoting VEEFX
This happens every NFP week. Monday/tuesday is okay then comes wed/thurs and market pauses waiting for NFP to come out. Every month this year, I have experienced drawdown going into NFP kicking off my damn drawdown manager function and chipping away realized profits from me. There is no solution but to ride it out as I don't believe (or want to) trade based on market conditions or avoid news releases. Looks like you also need volatility. Me2 and I have a way of tackling it by dissecting the 24x5 clockwork function of the global currency market sessions....
Spot on mate. I just ride them out and don't predict. It is damn annoying though at times. :-)

Nice to catch up V.

C
  • #3,365
  • May 4, 2018 12:34am May 4, 2018 12:34am
Quoting Copernicus
{quote} Yep mate. That is what I am meaning with divergent versus convergent. Sounds great V.....and I am doing the same thing.....however I am always biasing my result with divergence as it tends to be more sustainable long term. As a result....my primary strategy is always divergent in nature.....but I do get a little bit pregnant with convergence when it assists. The result is that return distributions of my portfolios are necessarily more volatile that pure convergent plays....however that is the nature of the game in the divergent space. What...
Yep.. Looks like you seem to nicely understand what I am trying to convey. Here's a bit more. Convergent CAN be successfully incorporated into a trend following divergent strategy. It works beautifully.... and it works even more beautiful if you ignore the divergent signals and trade convergence in the direction of the trend (what I mean by this is playing the pullbacks in the direction of the trending signal) without any regard to levels or trend lines or even looking at charts or measuring anything on charts with fancy calculations or looking into the historical price movements or patterns. When I lack my knowledge is in portfolio comparison to pick which one works best in which market conditions. I struggled but at the end I settled on "it does not matter.... just focus on Return on Investment over MaxDD" and with this simplified mindset, a tremendously realization will come to us... trust me. All these trading stats measurements on TE and myFXBook are simply obstacles in my view. If you let go of crap and just focus on why we are participating in the markets, a light bulb moment will happen. We participate because

  1. We have control over risk that is within our control and
  2. We want to make a decent return compared to any other highly regulated centralized market via leverag

That's all there is to it for me. Add and remove exposure in a strategic, methodical, sensible and logical manner.

Please tell me what other measures are absolutely critical for your principles of portfolio diversification. I am a noob so please dumb it down for me. Market is just a giant seesaw and every statistical component could be easily manipulated to our own advantage. Want a 100% win rate? - doable via running the floating losses, want 300 consecutive wins and one consecutive loss - doable, want a average win/loss - doable, want a perfect profit factor - doable, want a perfect expectancy - doable but doing all that does not equate to real tangible returns at the end of accounting period unless you do a close-all and then do the accounting. I am sure you know all this as you have lived and breathe this shit far longer than I have :-)

Staying in my lane...
  • #3,366
  • May 4, 2018 1:00am May 4, 2018 1:00am
Quoting VEEFX
{quote} Yep.. Looks like you seem to nicely understand what I am trying to convey. Here's a bit more. Convergent CAN be successfully incorporated into a trend following divergent strategy. It works beautifully.... and it works even more beautiful if you ignore the divergent signals and trade convergence in the direction of the trend (what I mean by this is playing the pullbacks in the direction of the trending signal) without any regard to levels or trend lines or even looking at charts or measuring anything on charts with fancy calculations or...
Yep mate. ROI/MaxDD or the MAR ratio is my preferred choice of risk-weighted metric as well.

Totally agree with you on pull-backs in trend. Here is how I play my current portfolio.

Attached Image (click to enlarge)
Click to Enlarge Name: CaptureA.JPG Size: 101 KB


Attached Image (click to enlarge)
Click to Enlarge Name: CaptureB.JPG Size: 96 KB


Attached Image (click to enlarge)
Click to Enlarge Name: Capture.JPG Size: 100 KB


Attached Image (click to enlarge)
Click to Enlarge Name: Capture2.JPG Size: 104 KB


Attached Image (click to enlarge)
Click to Enlarge Name: Capture3.JPG Size: 92 KB


The Blend

Attached Image (click to enlarge)
Click to Enlarge Name: Capture4.JPG Size: 110 KB





Quoting VEEFX
{quote}All these trading stats measurements on TE and myFXBook are simply obstacles in my view. If you let go of crap and just focus on why we are participating in the markets, a light bulb moment will happen. We participate because We have control over risk that is within our control and We want to make a decent return compared to any other highly regulated centralized market via leverag That's all there is to it for me. Add and remove exposure in a strategic, methodical, sensible and logical manner.
Spot on V

Quoting VEEFX
{quote}Please tell me what other measures are absolutely critical for your principles of portfolio diversification. I am a noob so please dumb it down for me. Market is just a giant seesaw and every statistical component could be easily manipulated to our own advantage. Want a 100% win rate? - doable via running the floating losses, want 300 consecutive wins and one consecutive loss - doable, want a average win/loss - doable, want a perfect profit factor - doable, want a perfect expectancy - doable but doing all that does not equate to real tangible...
I follow a strategy factory approach which is a never ending exercise.

Step 1

Take any simple divergent or convergent strategy that over as great a data set as possible simply stacks up and offers an edge....no matter how slight it might be. Don't be put off by the MAR ratio of the individual strategy.

That will filter out a billion strategies leaving some very simple classic strategies that you may have heard of that according to retail traders don't work any more.

Step 2
Logically look at each strategy and classify it in terms of whether it is divergent or convergent in nature. Bias your selection to those with a demonstrated positive skew.

Step 3
Plot the return series of each isolated strategy to generate an equity curve and develop a daily....weekly....monthly return series......whatever floats your boat. Note the trade risk % of each variant and use this with multipliers in your blended solution.

Step 4
Examine the return series of each alternative and work out the reasons for the profile and in what market conditions it thrived in and in what conditions it struggled in.

Step 5
Start superimposing the return series on top of each other to deliberately address each systems weaknesses. This will ensure that your choice of blended components is not just a random selection....but a selection that addresses risk inherent in each return stream.

Step 6
Start using allocation weightings to each strategy to view the outcome of the blended return series. You should find far higher risk weighted returns in terms of the MAR ratio.

Step 7
Scale up or down using position sizing multipliers on each component to generate a MAR ratio and more importantly a drawdown tolerance you are comfortable with recognizing that your worst drawdown is ahead of you.

Step 8
Calculate your risk of ruin etc. and ensure you are happy with the outcome and that it address as many market conditions as you can throw at it.

Step 9
Walk forward test for say 6 months and ensure everything is in order and that you have defined your tolerances correctly.

Step 10
Dip your toes into live trading the portfolio and continue to monitor against your backtest results.

Step 11
Improve...improve .....improve.
  • #3,367
  • May 4, 2018 1:48am May 4, 2018 1:48am
Quoting Copernicus
{quote} Yep mate. ROI/MaxDD or the MAR ratio is my preferred choice of risk-weighted metric as well. Totally agree with you on pull-backs in trend. Here is how I play my current portfolio. {image} {image} {image} {image} {image} The Blend {image} {quote} Spot on V {quote} I follow a strategy factory approach which is a never ending exercise. Step 1 Take any simple divergent or convergent strategy that over as great a data set as possible simply stacks up and offers an edge....no matter how slight it might be. Don't be put off by the MAR...
Thanks for sharing. Seems sensible approach to filtering out the nonsensical strategies. Wondering what would you need to assess my strategies (and in what format) to see how they compare with the ones you have chosen and what's the minimum sampling size? Unfortunately, none of my strategies can be backtested or optimized. It can only be forward tested and I track/assess performance only on weekly scale to meet my short term income objectives and a cumulative of 4 weeks to meet my growth objectives. I hate losing sleep over weekend gaps.

Also / do you own the strategies you run in your portfolio or do you use their service independently and then conduct portfolio analysis offline in your excel framework? Just trying to get the sense of the underlying design schematics. And / how long the blended portfolio in a live environment? I am trying to see how I fair compared to a proven, qualified diversified portfolio as a model to achieve (in my journey)

Also / how would you rate a strategy that performs better ROMAD on a high leverage 200+:1 compared to lower leverage. Mine runs around the clock 24x5 and miserably fails on 50:1 with the same %risk position size. I know why and I can do nothing much about it other than going fixed lot sizing which is not ideal for compounding with fat tails.

Quoting VEEFX
{quote} Thanks for sharing. Seems sensible approach to filtering out the nonsensical strategies. Wondering what would you need to assess my strategies (and in what format) to see how they compare with the ones you have chosen and what's the minimum sampling size? Unfortunately, none of my strategies can be backtested or optimized. It can only be forward tested and I track/assess performance only on weekly scale to meet my short term income objectives and a cumulative of 4 weeks to meet my growth objectives. I hate losing sleep over weekend gaps....
+ I forgot to ask the most important question. How do you (or your professional chosen strategies) measure drawdown?

  1. realized loss (via consecutive losses) or
  2. floating equity loss or
  3. peak balance to trough balance for the life of the account (MT4 account statement and i think myfxbook also)
  4. peak equity (high watermark) to trough equity for the life of the account (TE equity tab)
  5. any other way?


My unique way of measuring drawdown is not reflected accurately neither in TE or myFXBook because i don't care about account balance or equity. go figure :-)

Staying in my lane...
  • #3,368
  • May 4, 2018 4:01am May 4, 2018 4:01am
Quoting Copernicus
“The information we consume matters just as much as the food we put in our body. It affects our thinking, our behavior, how we understand our place in the world and how we understand others.” .....so it begs the question....what skills can be deployed to access the right information in a world suffering from data overload? Michael Simmons has a few ideas. {image}
Yep, I have same reaction to fluff/gossip/irrelevant information as to some allergen Cringe.

I became very selective what I read/watch. 99% of it out there is junk.
Cut short your losses. Let your profits run on. David Ricardo (1772-1823)
  • #3,369
  • May 4, 2018 4:06am May 4, 2018 4:06am
Nice to see such long list of open orders all profitable. For now

First broker:

Attached Image (click to enlarge)
Click to Enlarge Name: PTlr23Y.png Size: 62 KB


Second broker:

Attached Image (click to enlarge)
Click to Enlarge Name: OkXDMIj.png Size: 17 KB


I hold cocoa for 2 months already
Cut short your losses. Let your profits run on. David Ricardo (1772-1823)
  • #3,370
  • May 4, 2018 4:11am May 4, 2018 4:11am
Quoting Ramadas
Nice to see such long list of open orders all profitable. For now First broker: {image} Second broker: {image} I hold cocoa for 2 months already
Nice R. You are due a windfall :-)
  • #3,371
  • Edited 9:09am May 4, 2018 5:00am | Edited 9:09am
Quoting VEEFX
{quote} Wondering what would you need to assess my strategies (and in what format) to see how they compare with the ones you have chosen and what's the minimum sampling size? Unfortunately, none of my strategies can be backtested or optimized. It can only be forward tested and I track/assess performance only on weekly scale to meet my short term income objectives and a cumulative of 4 weeks to meet my growth objectives. I hate losing sleep over weekend gaps.
Hey V

The simplest way for me to compare and contrast different strategies is to do the following:
1. Determine the monthly return for the strategy. In my testing I always use the same capital deployed so calculating monthly return is a piece of cake based on net profit/opening capital for the month. You can scale things and compound the returns later but the initial key ingredient is the % monthly return from which you can calculate CAGR, Drawdown and sharpe etc. Ensure you use a capital level that reflects your ability/inability to trade minimum lot sizes. For example if you live trade with approximately $10K capital then use $10K fixed as your testing capital.
2. Adopt a standard trade risk percentage of capital or $risk for each trade under back-test conditions. You use this later for determining optimal position sizing that relates to your drawdown tolerance.
3. Go for as long a time series as you can muster. Time is not as relevant as variability in market conditions. The longer the test range....the better the chance that you have covered a greater range of different market conditions in determining if the strategy has an edge or not. You can never be 100% confident but you can estimate within confidence intervals.
4. Ensure that in your bactest you factor in real spread or a conservative estimate of it, a slippage for every trade that is conservative and you also adopt conservative holding cost estimates (SWAP). Best to be safe than sorry and also ensures you accommodate things such as occasional gaps etc. I use risk-weighted return metrics such as MAR as my basis for selection and simply forget about CAGR.
5. I parallel test my live trading in association with my backtest to ensure there is no style drift when under the pump of live conditions. If your backtest is more conservatively costed out than your live trading....then you are on the right track.


Quoting VEEFX
{quote} Also / do you own the strategies you run in your portfolio or do you use their service independently and then conduct portfolio analysis offline in your excel framework? Just trying to get the sense of the underlying design schematics. And / how long the blended portfolio in a live environment? I am trying to see how I fair compared to a proven, qualified diversified portfolio as a model to achieve (in my journey) Also / how would you rate a strategy that performs better ROMAD on a high leverage 200+:1 compared to lower leverage. Mine runs...
I maintain monthly returns of the professional funds and all my backtests. All my systems are home grown....but that is not essential as you have identified when you diversify. I maintain all return results in a database (which I horde like a bower bird) as you never know when a particular return series with a characteristic equity curve signature will be needed for blending purposes. I frequently resurrect old strategies that now have a use for the portfolio.

Mate with regards to leverage, provided you have an edge and you are trading the lower timeframes it is your biggest friend in regards to diversification. Lower leverage starts eating into your capital base and restricts your ability to widely diversify. As a result, you need to step towards rotational strategies under full capital commitment like Nick Radge does with his stock trading strategies. For example he deploys the same momentum strategies but is restricted to say 20 instruments on a $100K account with $5K allocated towards each instrument.

Of course as with everything there is a but......and with increased leverage, this does restrict your ability to step out to longer term solutions due to holding costs that ramp up as your gearing increases.....so as a general rule when trading diversified short term divergent systems....leverage is good......but as you go out to the medium to longer term you pay a price in trade efficiency due to holding costs. Also it goes without saying that with the two edged sword of leverage...if you don't have an edge then you will be out of the game quicker.


Quoting VEEFX
{quote}+ I forgot to ask the most important question. How do you (or your professional chosen strategies) measure drawdown? realized loss (via consecutive losses) or floating equity loss or peak balance to trough balance for the life of the account (MT4 account statement and i think myfxbook also) peak equity (high watermark) to trough equity for the life of the account (TE equity tab) any other way? My unique way of measuring drawdown is not reflected accurately neither in TE or myFXBook because i don't care about account balance or equity. go...
My preferred drawdown metric is simply realised losses from my equity curve. It tends to be an industry standard in the funds that I follow so to ensure apples for apples this is what I use. It allows me to therefore compare and contrast my MAR ratio with industry standards.

Also because I do not predict outcomes and let my profits run....I am really only concerned with realised losses. I measure portfolio heat which is my method of controlling risk on live trades. It simply assesses in a worst case scenario the result if all my trades instantaneously went pear shaped. If my actual current result is above my heat tolerance then cool. If it gets close to my max risk tolerance defined by heat then I start risk mitigating in the portfolio.


I hope this helps mate

C
  • #3,372
  • May 4, 2018 5:30am May 4, 2018 5:30am
You can't teach an old dog new tricks :-)

hehehehehehehe

https://dms.licdn.com/playback/C4D05...k6gkDouePVdW2M
  • #3,373
  • May 4, 2018 8:26am May 4, 2018 8:26am
Long term SDC so steep, short term SDC cannot be found. USD HUF
Attached Image (click to enlarge)
Click to Enlarge Name: 1ObyzBr.png Size: 50 KB
Cut short your losses. Let your profits run on. David Ricardo (1772-1823)
  • #3,374
  • May 4, 2018 8:58am May 4, 2018 8:58am
Inserted Video
Cut short your losses. Let your profits run on. David Ricardo (1772-1823)
  • #3,375
  • May 4, 2018 9:04am May 4, 2018 9:04am
Quoting Ramadas
Long term SDC so steep, short term SDC cannot be found. USD HUF {image}
I would give it a swerve till one can be found :-)
  • #3,376
  • May 4, 2018 10:58am May 4, 2018 10:58am
Quoting Copernicus
{quote} Yep mate. I hear you. The benefits of a dynamic portfolio in terms of attacking market inefficiencies is the way to go.....provided that we place our attention and effort in managing portfolio heat.....which can also mean that we may occasionally miss some great periods of momentum on those instruments we avoid due to heat. There are so many swings and roundabouts in this game. I would prefer to bundle up our instruments into asset classes and make some rules for dynamic selection within each asset class. That way...it may help in preventing...
Well, system diversification is good, but the systems must make money individually.

April was not good for simple divergent systems:

Attached Image (click to enlarge)
Click to Enlarge Name: jDt77J6.png Size: 179 KB


And in fact whole 2018....

Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot1.png Size: 232 KB


it doesn't even compare with EDTT + Reverter

Inserted Video
Cut short your losses. Let your profits run on. David Ricardo (1772-1823)
  • #3,377
  • May 4, 2018 3:53pm May 4, 2018 3:53pm
Quoting Copernicus
{quote} Hey V The simplest way for me to compare and contrast different strategies is to do the following: 1. Determine the monthly return for the strategy. In my testing I always use the same capital deployed so calculating monthly return is a piece of cake based on net profit/opening capital for the month. You can scale things and compound the returns later but the initial key ingredient is the % monthly return from which you can calculate CAGR, Drawdown and sharpe etc. Ensure you use a capital level that reflects your ability/inability to trade...
Thanks man. You and Mr. Vega are the only two dudes who make me read posts several times just to get through my pea sized brain.

Long week so I will read again over the weekend but want to quickly clarify the "realized drawdown from equity" comment. Would you consider a 0% drawdown if 100% of my closed trades are in profit and all my floating trades are in loss by the end of the week? You said you don't hold trades so it might work well for you to keep it simple. If I understand you, "floating equity curve analysis is replaced by Portfolio heat map which is fine also depending on the strategy. I am just not sure if heat map can be universally adopted to all strategies. For example: 1.5 %Risk position on a 65 open position would result in 100% red on your heat map. right? To expand further, the way I "over-leverage" to a good exposure spread via diversification of price, time and geo-economies is to keep adding positions up until my margin level drops to 150% (I could go below 100% on some brokers) and then I stop after some 60+ positions (on 400:1 leverage) are built up by organically optimizing my available margin. %risk stays the same but $ equivalent risk actually goes down with every new position added - ask me how :-) I don't think many understand this fabulous way of using leverage to introduce diversification like you do.

Now, take it a step further and ask ourselves how do we manage a account with 100+% risk exposure that will result in 100% profitable closed trades and floating equity capped at 10% in loss or max Equity Drawdown? This is where the gravy is :-)
The answer lies in a eye-opening post made by stt on October 23th 2017 to one of my inquiries and I have literally never slept since then. Lot of tears and frustration and mind-explosive revelation!

Quoting stt
{quote} Only way to really control the drawdown is to keep cutting your bet size as your equity goes down. This means that at your max drawdown limit, your trade size will become 0 and no further drawdown will happen (and also you will never come out of drawdown). There is no other way - I can even prove it by Proof_by_contradiction method. Assuming there exists a strategy that guarantees the max drawdown to be X%. Now suppose we are at X% drawdown currently. That means, the strategy will take next trade which will have 0 drawdown. i.e. strategy...
Thank you Mr. STT.
Staying in my lane...
  • #3,378
  • May 4, 2018 4:16pm May 4, 2018 4:16pm
Quoting VEEFX
{quote} Thanks man. You and Mr. Vega are the only two dudes who make me read posts several times just to get through my pea sized brain. Long week so I will read again over the weekend but want to quickly clarify the "realized drawdown from equity" comment. Would you consider a 0% drawdown if 100% of my closed trades are in profit and all my floating trades are in loss by the end of the week? You said you don't hold trades so it might work well for you to keep it simple. If I understand you, "floating equity curve analysis is replaced by Portfolio...
It seems appropriate to share my diversified portfolio since others are also doing the same.

Portfolio Start date: May1st
Type: 100% automated during week days... requires manual tweaks during weekends.
Time Horizon - Weekly
Strategies: 6 (The seventh one with "ALL" is a new addition in May so unproven yet)
Position Risk: 1.5%
Exits: Hybrid/Manual/Automated
Including screenshots of the best performing strategy in the portfolio (I have got to feel good also right :-)

DFM = Divergent or Diversion from Mean
RTM = Reversion to Mean or Convergent (in your terminology)

Please guide me if this "Risk of Ruin" is acceptable. I just have not spent enough time to study all these fancy stats so I have no idea if I am on the right track or not. If I am causing distraction to you or your readers, let me know and I will delete this post

Disclaimer: This is demo on myfxbook. I don't hook up live accounts on public forums

Quoting Ramadas
{quote} Well, system diversification is good, but the systems must make money individually. April was not good for simple divergent systems: {image} And in fact whole 2018.... {image} it doesn't even compare with EDTT + Reverter https://www.youtube.com/watch?v=loG3Ezz6dj8
I don't have the entire month of April to share as I create new demo accounts every weekend to link up but recall every week in April looks very similar to this one in last week. Mon/Tue/Wed nice runs and if I don't close the shop, nasty drawdown kicks in on thurs/Fri.
Attached Image(s) (click to enlarge)
Click to Enlarge Name: Image1.png Size: 20 KB
Click to Enlarge Name: Image2.png Size: 30 KB
Click to Enlarge Name: Image3.png Size: 12 KB
Click to Enlarge Name: Image1.png Size: 28 KB
Staying in my lane...
  • #3,379
  • Edited 10:26pm May 4, 2018 8:51pm | Edited 10:26pm
Quoting VEEFX
{quote} Thanks man. You and Mr. Vega are the only two dudes who make me read posts several times just to get through my pea sized brain. Long week so I will read again over the weekend but want to quickly clarify the "realized drawdown from equity" comment. Would you consider a 0% drawdown if 100% of my closed trades are in profit and all my floating trades are in loss by the end of the week? You said you don't hold trades so it might work well for you to keep it simple. If I understand you, "floating equity curve analysis is replaced by Portfolio...
Kaboom. Thanks mate. I am going to need some time to digest this and STT's post.....very helpful. It gives me fresh ideas to mull over .....

There are many valid ways to measure drawdown but the most important feature of any definition is in how you use it and this will be system dependent. On a very simple level....using realised equity as my basis for Drawdown is simply a convenience for me in having a permanent record of the drawdown signature of a strategy which I can compare to other funds etc. The most stats you get from the fund managers relate to monthly returns hence there is no way I can gather info on floating equity from these guys to cross compare with.

Thanks for your comments on heat. They are very valid. I trail my stops which therefore allows me to add to my live trades when the cumulative sum of heat gets below my max tolerance as trades get to break even and beyond. This way when markets are correlated and heading in an adverse direction I don't tend to get overexposed to them and vice versa when correlation assists my portfolio. While a useful risk exposure tool in itself...it is my main weapon to manage correlations in the portfolio. Correlation can be good provided it is in your direction and bad when it is vice versa. Having a dynamic heat level that keeps track of the worst case scenario allows me to take advantage of good correlations when they matter and avoid bad correlations when things get ugly. :-)

Your idea of maximising your diversification using leverage is a good one and I like the objective way you are dealing with it. This game is about optimal use of finite capital...which many don't seem to understand. It is not about system performance which most tend to think relates to system profitability which cannot be controlled....but about methods to maximize your capital management efficiency by managing risk which can be controlled. I then leave things to the non gaussian nature of the market to build my wealth. Once this clicks then you are on the road to long term wealth.

Cheers

C
  • #3,380
  • Edited 11:07pm May 4, 2018 9:00pm | Edited 11:07pm
Quoting VEEFX
{quote} It seems appropriate to share my diversified portfolio since others are also doing the same. Portfolio Start date: May1st Type: 100% automated during week days... requires manual tweaks during weekends. Time Horizon - Weekly Strategies: 6 (The seventh one with "ALL" is a new addition in May so unproven yet) Position Risk: 1.5% Exits: Hybrid/Manual/Automated Including screenshots of the best performing strategy in the portfolio (I have got to feel good also right :-) DFM = Divergent or Diversion from Mean RTM = Reversion to Mean or Convergent...
Cheers V. This is great stuff. I am getting excited now. For some reason folks on FF don't spend much time assessing their performance metrics and are more interested in the system itself....but these results are like a fossil record where you can dig down into objective data and precisely determine the strengths and weaknesses in your system in response to market conditions at the time.

It is great to see someone who is stepping into the world of system diversification. Your chart below says it all for me. You can tell when your systems are correlated and when they are uncorrelated or negatively correlated....though we need a bigger sample size to be more confident in our assumptions. When you investigate market conditions during these times....that's what you need to know as it gives clues in how to construct systems that address the weaknesses and compound the strengths to get maximum bang for buck through risk management as opposed to risk guessing. For example the April 25-April 26 result tells me your systems were correlated and yet between 26 Apr and 27 April they were a mix of non-correlated and negatively correlated results. Now you need to ask yourself why? Look at the market conditions and look at your system design to find that out. So how then would you improve the result so that the composite result of the entire portfolio was more a linear line in the upwards direction?.....in other words a better risk-weighted result that you could then magnify through position sizing and dynamic capital allocation to get most bang for your finite total buck.

Attached Image (click to enlarge)
Click to Enlarge Name: Capture.JPG Size: 35 KB


There is so much useful info that can be gained in portfolio performance when you cross compare those equity curve signatures. When you get a few years of stats under the belt, those signatures give you the secret sauce in how to deliberately configure your portfolio to maximise your leverage potential using position sizing and in how to manage your drawdowns.

This game is not about letting the market decide your fate.....but rather in controlling those aspects of risk which can be controlled to push the odds in your favour through deliberate action.

C
8 traders viewing now