Diversified Trend Trading Approach

Why Diversification Matters


Given that this thread has explored a particular trading approach that capitalises on any trending instrument and is not wedded to any particular market or asset class, it is now probably a good time to look at the subject of diversification and how these principles can be employed to make the most efficient use of available capital.

Rather than being focussed on any particular style of trading strategy (eg. whether it employs trend, swing, momentum or mean reversion principles) this thread will now spend some time examining the theory of diversification (whether in terms of diversification in instrument selection, system selection or timeframe selection) and how the theory can be applied to manage the correlation of expected returns delivered by a diversified approach and manage the equity curve.

Diversification in a nutshell allows for the most efficient use of available capital to increase overall system performance. This is attributed to the fact that when systems, instruments or timeframes are combined, each of which possess their own distinctive equity curve profile, it is possible using correlation assessments to combine these elements to produce a cumulative profile where a trader can enjoy the positive contributory benefits of each element to collectively magnify performance and at the same time dilute the combined drawdown of a diversified approach. The combined result allows for lower capitalisation that what would be required by an individual component and thereby increasing the overall rate of return.

What must be noted however is that diversification requires selective application as it is essential that the best results of each individual component are not diluted via carte blanche diversification. Rather the focus of methods of diversification are used to apply measures to simply 'plug the equity drawdown' or 'fill the gaps' of an equity curve to reduce the volatility of returns and lift the curve than could be achieved by a single component from summing collective positive performance.

.....anyway, it is time for a change in tack to the direction of this thread and a more in depth look at diversification as a method to improve performance is now warranted.

Rather than creating a new thread where the subject could lose it's context to the trend trading technique discusseed in this thread, I would like to continue a discussion on diversification methods and varying approaches within this thread context.

Looking forward to some robust discussions on the subject of diversification. When things slow down a bit for me at my primary work, I will be able to get into this in more detail, but thought now might be a good time to change the thread direction.


Regards

C

I don't have much to say on diversification that isn't already known.

Basically just spreading your risk among as many loosely correlated markets as possible.

I have always tended to disregard portfolio diversification as applicable to my retail world, but have only more recently started to regret not getting deeper into it because the nuts and bolts of it, once digested, are fairly easy to apply but dramatic as a performance measure.

The more I get into portfolio diversification (asset allocation) the more I realise how too much of my trading attention has been focussed towards attempting to be a great trader or a rocket scientist using unnecessarily complex approaches that pay too great an emphasis towards the trade technique itself, when sitting right in front of me is a well known and understood approach that offers excellent long term performance results while minimizing unnecessary risks and costs.

The correct application of diversification principles (using equity profile contributions that don't correlate very well) reduces portfolio volatility which costs real dollars through a phenomenon referred to as 'risk drag'. Effectively risk drag reduces performance efficiency (by tying up available capital significant reducing earnings potential) and simply by addressing this factor and optimising performance efficiency, you can significantly accelerate performance above and beyond that which can be delivered by more accepted approaches in this retail world such as optimal position sizing or money management methods. Any asset with a volatility of returns of greater than 0 will see it's multiyear returns erode from risk drag.

We are all familiar with the notion that 'higher returns typically carry more risk' but many of us are less familiar with the ability to 'increase returns with no additional risk' by evaluating the portfolio components against a measure of performance efficiency referred to as 'the efficient frontier'. The reason for this lack of knowledge is attributed to an assumption buried in the former quoted statement relating to 'perfect performance efficiency'. Of course for a perfectly efficient system, 'higher returns are 'always' associated with more risk' but for less than optimal efficient systems portfolio diversification using 'Modern Portfolio Theory' allow you to achieve enhanced performance with no greater risk simply through optimising strategy efficiency.

Too few of us retail traders venture into this world, probably because it is not regarded as sexy, however when an emblazoned opportunity such as the terrific power of diversification shows up on the doorstep, for those traders who truly are concerned in obtaining an edge in this market, it really is a 'free lunch' opportunity. I would always prefer to play roulette with a biased wheel.

....anway, will start getting into it some more on the weekend and hope to get input from those on FF that are better versed than me at it.

I told u u r a nerd but u wont listen, "I realise how too much of my trading attention has been focussed towards attempting to be a great trader or a rocket scientist using unnecessarily complex approaches ".

I was trying to find when you mentioned this in my forum correspondence but just can't seem to be able to place my finger on it.

This is new territory for me. Can't wait to hear what you have.

Gone thru the thread. Robust trading approach!

For those of you that are interested in following the paths of others and seeking to not re-invent the wheel but obtain a sound fundamental knowledge base upon which to apply the science (with art) of trading, there is a freely available lecture series offerred by MIT which is worth the while getting the head into.

I remember when I first became interested in science and as a naive newbie thought I could single handedly dance on a similar level with the most learned physicists of our time. That was my ego speaking. Fortunately, despite the shellacking I received from the scientific community in being so naive, a very patient and tolerant expert in his field (David Deutsch...I owe a lot to this great guy even though he probably doesn't realise it) suggested to me that it is perhaps wise to not jump to any conclusions before at least taking the same to read and at least acknowledge the principles of science and the path it has taken to get to this point. After sucking in my ego, I commenced my education and while I initially thought it would bias me, it actually empowered me and my confidence in navigating through the scientific world and has not only rigurously answered my initial naive questions but has armed me a knowledge base to peer further into the unknowns than I thought was hitherto impossible.

The rationale being that the many naive viewpoints that I previously had, have already been fully explored by others and already lie in that body of available information we call science. The greats of science were not those visionaries who took the subject on with a naive personal mindset but those like Einstein, Wheeler, Feynman et al who stood on the shoulders of others such as Newton and took on where they left off. To be able to to do this however entails an understanding that has gone on before but does require an open mind in identifying the assumptions that have been made before and being able to identify the weak spots that are worth investigating further.

In the late 19th century Lord Kelvin pronounced with authority that most of science was all known with no new discoveries to be found aside from more specific measurements.......but he did pay note to only 3 small clouds on the horizon that required further interrogation to have a Theory of Everything. Well as it turned out, those 3 small clouds led to major revolutions in scientific theory being the Special and General theories of relativity and the theory of Quantum Mechanics. Without an understanding of what has gone before, these theories would never have evolved in such a small spate of time. The development of these more accurate models that extended from less effective models was attributed to the original attempt to oversimplify or apply occams razor to previous assumptions to make the model more tractable. There lay the problem in applying the principles of parsimony naively. It was only through revisiting these assumptions that the theories could be extended to incorporate these 3 small clouds on the horizon and thereby develop more effective models that applied to a greater domain of enquiry. Failure to question the original underlying assumptions was the fundamental root cause of the dillema. But to be able to progress in science it is essential that you first understand the theory and then specifically look at the assumptions underpinning the theory. Only then can you hope to make better models.

....anyway, the moral of this story is as follows. If you want to be a retail trader who is truly successful, rather than taking your best bet using your own mind to take on this challenge, then a quicker way to become a more successful trader is to stand on the shoulders of others and use their insight and hard work to then identify where any unexploited opportunity may lie. In trading this market recognise what a small fish you actually are and at least learn what other participants may be applying so you can combat your enemy.

All this takes is a bit of time learning the theory of where we are today. While individual trading books and forums such and these are always recommended as a way to build the knowledge base, a recognised course is the best way to initially learn from the experts as you need to look at things holistically before understanding how to navigate the specific. It is such a complex world out there that you need a context before you can navigate it. How do you know that the content of that book you are reading is the answer to your trading woes when there may be another book that explains things from a different context and may be a more efficient approach to take. That is why context is critical in your journey to fast track in the most efficient way what potentially is a never ending quest for knowledge.

While this thread will over the course of time explore the principles of diversification, it does require a certain pre-requisite knowledge of the financial markets. I therefore strongly suggest that participants of this thread who are willing to learn spend some time listening to the following course content offerred by the Massachusetts Institute of Technology (MIT) and then delve specifically onto the sections on risk and return, portfolio theory and the Capital Asset Pricing Model (CAPM).

I have only just started getting into this so hopefully it is a journey that others will be taking with me but already I am identifying ways that my performance can be significantly improved simply using other's hard work in applying their principles to my technique.

The lectures are interesting but be prepared to dust off the cobwebs and get into a bit of maths. Mathematics has never been my strong suit but it is amazing that when applied to a subject of interest such as trading the markets, it actually is quite inspiring and far less dull that what was taught at school. While not an expert, they lead you through the maths in an understandable way to the extent that you actually learn something that can be usefully applied in your trading.

I hope you enjoy the journey as I am. :-)

C

The 4 essential components of successful diversification include:

1. Understanding the relationship between risk and return - A key pre-requisite to understanding why diversification and it's principles actually work.

2. Selecting the component groupings (eg classes of groupings) that you wish to use for diversification processes (eg. instruments, timeframes and systems and combinations) and collecting the historical data of a sufficient sample size to plot their relationship on a risk/return graph. This is where you need to assess your component choices as a grouping in terms of risk and return to identify where they plot on the risk/return graph and isolate those optimal most efficient grouped choices against inferior choices using 'the efficient frontier' as your guiding principle. This is an essential step in analysis as it is where you examine the efficacy of alternative groupings before you move onto step 3;

3. Determining the right mix of those groupings to achieve your objectives that are based on objective historical statistics - This is where you apply principles of correlation to reduce 'risk drag' to compile your portfolio that is configured to your objectives; and

4. Rebalancing your portfolio periodically to maintain your desired mix and turning a whimsical objective into an objective target that will be achieved. This is where the magic of diversification seems to be able to generate extra returns out of thin air using a principle referred to as 'volatility pumping'.


.....to be continued. A busy weekend ahead and a nagging wife means I need to return to this at a later time. Have a great weekend all :-)

C

maybe your wife has a point???

Thats a great post C....

In any work-life event if you want to fast track to being in a position where you can truly understand a successful route over the landscape you have to be guided by someone or a group who has experienced the correct route, otherwise the initial experiences are just going to be a 'learn the the hard way' lesson, once these experiences have been absorbed we can then build our own route that will sit with our rationale while adhering to the ' successful methodology'.. sure we can end up in the same place with self education but the learning curve time is completely different.. Finding successful traders who are willing to share thoughts and idea's isn't easy, though they are around..




You know my thoughts on this C.. If you want to go fast, go alone. If you want to go far,go together. Why make it difficult for ourselves when we all have different unique strengths that we can lend to others to fill the spaces..


Two great posts... I know I don't post to much in here mate but I very much enjoy the thread... Speak soon

hello C,

am all for diversification, but not really to smooth the equity curve.
to me it doesn`t matter that some months are far better than others: at the end of the year the curve might be the same for most instruments.

however am more concerned about sudden flashes like the China scare in August.
look at all the forex pairs: they all reacted violently with the exception of USD/CAD.
the indices were down a lot.
am trading 10 different instruments for protection against the big storms.
but if you`re unlucky, you might be looking at a huge drawdown.

so what I`m saying is that with all the different instruments there are out there, I doubt there were a lot of them which didn`t react much.
am happy to be proven wrong.........

Cheers Okehiedon. Thanks for the comments

Totally agree with you Ja. In this lonely world of retail trading it just makes so much sense to collaborate and share the immense skill sets that many of us have.

BTW.....Absolutely loving Dance Spirit mate. You know my music tastes and you have nailed it with this. Speak very soon :-)

Hi yonnie. Thanks for dropping in.

The wife has released me temporarily from the weekend chores....*tongue in cheek*, so can spend a bit of time discusing your comments.

I understand where you are coming from relating to flash crashes. They are to be feared by all traders but with those specific one off events such as the 'CHF depegging' ,at least a diversified approach with a very low trade risk assigned to each instrument seems to me to be the safest way to ride them. At those times when volatility is extreme and liquidity is severely compromised, stops might not even save you. Perhaps the safest way to navigate these times (without having to pay an arm and leg using say guaranteed stops offerred by some brokers) is to only allocate a very small fraction of your capital to a specific instrument. I know in hindsight that the best way to avoid say the 'Swissy' event is to not trade pegged currencies, but I certainly was caught on the wrong side of this event but from diversification only paid a 2% price for the joy of this poor trading decision which included approximately 100% slippage.

With diversification (and I mean significant diversification with as little correlation as possible across asset classes) you simply cannot allocate a significant % to any particular instrument as your trade capital is finite. As a result, if a flash crash occurs, which seems to be fairly well restricted to a specific set of highly correlated instruments (eg. all Swissy pairs) the impact is never significant to the portfolio and you can ride them without having to have sleepless nights worrying about them. For example the correlated grouping mix may only account in total to a portfolio exposure of say 1%. Assuming the stops don't work and you get a total 2% hit from your exposure plus slippage, it allows 100% slippage but allows you to survive the holocaust in pretty good shape. You simply swear a bit and don't tell the wife *sniggers*....then carry on as if nothing never happened.

For those more enduring crisis events when all instruments seem to adopt almost perfect correlation (eg. the 2008 GFC), the nature of a diversified trend trading strategy actually capitalises on these events by jumping on board and that's where most of the alpha is generated from. Market failure from these type of crisis events is not instantaneous and rolls through asset classes like falling dominos. High correlation between instruments and asset classes takes time to unfold during these events. Riding these events with trailing stop losses ensures that in a brief spate of time your positions become profitable allowing you take new trades in more correlated positions. It is important to recognise that in diversification you need to closely look at the nature of each discreet asset class as correlation between asset classes is what saves you in these periods. For example when the equities market get's hit, almost all instruments become perfectly correlated almost instantaneously. For the futures market however and it's derivatives, this feature is far less pronounced. While in this thread we talk mostly about forex or indices that are available with most brokerages, it really pays to add to this mix by using brokers that offer exposure to a broad range of asset classes.

When using a diversified portfolio I manage risk by looking at total portfolio exposure (the summated risk exposure of all positions still in loss to their respective stops) but disregard the exposure relating to all profitable positions in the portfolio. The profits are locked in by trailing stops. This way, I can afford to get aggressive in the longer term crisis events without having to unduly worry about positive correlation during these events.

With your specific reference to how all instruments appear to become correlated during crisis events, you are quite right but in terms of what can be achieved in trading forex and indices but these two particular asset classes are actually quite highly correlated when the proverbial hits the fan. I think we need to stretch further in terms of diversification to get a better result. A portfolio diversified across forex, indices, hard commodities, soft commodities, treasuries, a mix of discreet equities by their summated exposure provides far better protection of capital than that which could be achieved by a single instrument. So no matter the degree of correlation (provided it is not perfect correlation), diversification is always a safer bet than simply disregarding the role it plays in risk management.

We have not even yet begun to look at system diversification which IMO is an even better method of optimising portfolio risk as a diversified system approach allows you to specifically integrate systems into the portfolio that can never be perfectly correlated. For example you can design your system to incorporate two strategies (each with a slight positive edge) where when one strategy is 'turned off' the other strategy is 'turned on'. This ensures that you never run into correlation issues. For example if your fundamental bread winning strategy is a medium to long term diversified trend trading approach, you can supplement this with a diversified reversionary approach that is only active on a particular instrument when the fundamental strategy is not trading.

We often think that the most efficient method to obtain a positive result is by employing the most powerful trading approach but in the land of diversification this is not the case. For example by employing very simple but robust strategy components that only have a very slight edge (say 1-3% per annum), the total porfolio performance result can exceed that which can be delivered my a most efficient individual instrument trade strategy. The reason for this is that in a diversified portfolio your worst drawdown is lower than that which can otherwise be achieved with a single instrument and you can therefore commit more of your trade capital to work to achieve overall positive return.

Through diversification you can set more achievable long term performance targets that allow you to weather the storm of what the market throws at you without having to over-optimise or squeeze every last performance drop out of a specific singular non-diversified approach.

Interesting latest podcast from the perspective of a specialist short trader, Laurent Bernut.

It opens the mind to how to apply short selling strategies to any market (eg. equities and commodities markets that may traditionally be viewed as only being applicable to long only strategies) to achieve symmetrical long/short strategies by simply relative strength trading amongst pairs. I used to spread trade equities to remove market risk with a degree of success so I have an open mind in relation to his thoughts about short trading and in particular his methods of managing risk through 3 forms of stop loss technique. Interesting stuff.

Enjoy

C

hello C,

delved a bit deeper and looked at metals, softs, grains etc (not bonds) and all of them reacted more or less to the China scare.
I have come to the conclusion that in times of stress in the market place, diversification as a tool is of little use.
unless ofcourse you are lucky enough to be on the right side with some futures when it strikes.

have decided to ditch 9 other instruments I`m trading and trade the Nasdaq100 only.
that can only be good to have only 1 instrument to concentrate on.

good that you started the subject about diversification though: it made me have a look at the stress test

No probs yonnie. Keep me posted how it goes with the Nasdaq. It's nice to hear from traders dealing outside of the forex world. :-)

Here is a bit of an overview first of how to optimise risk and return.

Once you have an understanding of these principles and are comfortable about moving on to the detail then move onto the following excellent video introduction using excel solver to make the process easier on yourself. There is also a detailed review of the mathematics surrounding optimising portfolio weights (if you are interested).

Watch the excel solver tutorials in the following order:
1. Optimum Portfolio Weights for Maximum Sharpe Ratio;
2. Sharpe Ratio Maximisation with Excel Solver; and
3. Markowitz's Portfolio Risk Minimisation with Excel Solver.

Enjoy

C

PS Remember this process can be applied to any equity curve contribution to a total portfolio, and rather than merely focussing on individual instruments for diversification purposes, also consider individual equity contributions from different strategies and also strategies applied to different timeframes. If you are concerned about how correlations never stay still and that during particular market conditions, instruments can become remarkably correlated through market globalisation then specifically concentrate on system diversification as a measure to avoid this issue.

PPS An even better demonstration of how to use excel solver to optimise portfolio weightings w.r.t. sharpe ratio and a further good excel demonstration with a freely available template to assist that is used for Efficient Frontier determination....and for those seeking to optimise for a large diversified portfolio, here is an example on how to make a powerful excel model.

Very interesting C, thanks for the links and info.