Diversified Trend Trading Approach

Example 1A - Gold (XAUUSD)

Update - The Momentum Ratchet and Volatility Extension

You may think that the SDC represents a linear representation of price action.....but think again. The SDC breathes in association with the data sample that comprises it and responds to non-linear mechanics.

In a previous post I showcased how with momentum increases in the direction of the underlying bias, that the SDC ratchets in the direction of the bias. What you will also observe in the chart below is how the SDC also extends or contracts with volatility. The blue rectangles represent the original representation of closing prices across the data range since trend commencement up until trade entry of 1.0 sigma. Now observe how the momentum push since entry has increased the variation of closing prices within 1.0 sigma according to the now extended data set. The orange rectangle represents the current volatility range of closing prices within 1.0 sigma.

What this actually achieves is that with momentum accelerations in the direction of price bias, you are afforded more breathing room for your stop. If momentum in the direction of historic bias declined, then the narrowing of the SDC naturally reduces the breathing room. This has important risk management benefits which is the name of the trading game.

What this tool explicitly ensures is that when your are right, you are given room to allow profits to run. When you are wrong (which is more often than right), the SDC ensures that you are out of the trade via SDC compression in quick time in a very objective (non-subjective) manner (cut losses short). Once again we are using statistics of price action to make the judgements....not our intuition.

Thank you for the link FXEZ, I'll look forward to reading the paper..

Very interesting paper.



Not surprising regarding the findings of back-test overfitting (which can be addressed by robust techniques that specifically avoid over-fitting), but what was surprising was the importance of the tail ratio. What this suggests to me is that only those metrics which specifically account for more extreme (non-linear) anomalies such as annual volatility or drawdown exhibited positive correlations between In-Sample (IS) and Out of Sample (OOS) return distributions. The remaining metrics such as Sharpe, Sortino, Calmar et al may have had short term predictive ability but appear to be massively diluted by 'noise' over the longer term. Is it the fat tails alone that have predictive substance eg. the anomalies given that their impact exists outside the day to day price churning? It makes sense if you consider that it is only during extreme events that market behaviour becomes unidirectional in nature.

I'd be interested to hear your views on this paper FXEZ. :-)

C

Words of Wisdom from Peter Brandt - 50 + years of experience actively trading the markets

Summarized transcript of major points:

1. It took me 4-5 years before I became accustomed to the market and identify how to trade and build sufficient capital that I needed to trade. I don't know how many small accounts I blew up before I finally caught the scent for the approach to markets that suited my personality and made sense to me. I had to experience the blow-ups and make enough mistakes for me to at least identify the themes upon which would contribute to me becoming a successful trader.
2. My dear friend, an excellent trader, as good as the best that I have known in my lifetime shared with me his philosophy on risk management such as keeping your capital together and taking losses quickly.... not his technique...... which is what so many novice traders today do not consider sufficiently. Those of us that started on exchanges had a big advantage in that we were part of a trading community. It is a very lonely game today. Fellowship, community and sharing of knowledge is an essential pre-requisite of trading.
3. You have to make mistakes, again and again....this is the only way you learn. Learning to trade is an exercise in problem solving more than anything else. Nearly all my early assumptions were just wrong. A big mistake I made was that I could hook my wagon onto somebody else's technique. It takes a few years to work out that if you are going to trade, you have to do it on your own and make your own mistakes, work out what makes sense to you. In the process you actually learn for yourself. bit by bit. Learning to trade is like going to university but it is the market that decides whether you are right or wrong. Learning to trade is a process that takes between 3-7 years or so before you get sufficient confidence before you can go all in. Novice traders who come to the conclusion too quickly that they know what they are doing will blow up and pay the ultimate price. ;
4. What kept me going during the blow-ups was that I had an alternate income stream so that I did not put my lifestyle at risk. It was a huge advantage that I did not have to become successful right away. I had the time to learn it properly and did not have to dip into my trading account to pay for living expenses. A very bad place to start off is in being immediately dependent on trading income. A blow up therefore was not the end of the game for me. I still faced some nervous times when I eventually cut out on my own and it is scary now to think about the risks I took in those days. It was lady luck that allows me to talk to you today as a trader.
5. To any new trader out there......go back and get a job. It scares me when I hear a new trader saying I am going to quit a job paying $X a year with $X saved away and begin full-time trading and be successful from the get go. It is a recipe for disaster. Trading can be a wonderful hobby. You don't have to trade full-time. It can be a rewarding venture on the side. You need at least 2 years of savings in the bank to pay for your living expenses. The capital you trade with should not be capital you inherited or have saved from your prior job. It should be capital you have generated from your trading. Until you generate sufficient capital from your trading to be the primary source of your future revenue ....do not trade full-time. Only trade full-time when you can demonstrate that you are successful.....then you can quit your job.
6. In the 70's I tried this and that..before I got interested in classical charting. It really was Richard Schabacker....(not Edwards and Magee) that in a pure sense turned me onto classical charting principles. It just made sense to me. It is a craft that starts from apprenticeship to journeyship. I continue to tweak things that suit my philosophy but despite the nuances, I have stayed true to the classical principles. I am however now more long term and prefer to trade the weekly timeframe using the daily charts for timing. I like horizontal boundaries like triangles, head and shoulders, rectangles .....but hate diagonal patterns and trend lines....they freak me out. I look for patterns that evolve over at least 8 weeks. I trade patterns that go up to several years of duration but the breakouts on these longer term patterns are harder. I specifically consider myself a breakout trader. I have never been really good at selling rallies or buying breaks. It is a skill set I wish I have....but I don't have it.
7. I trade futures and forex. Very few equities...but I do trade indices globally...but 90% forex and futures.
8. Breakouts over the years have become a lot less clean cut. Charts particularly daily charts were more reliable in the 80s and early 90s. I still find reliability in weekly and monthly charts. The main reason of course is HFT. In my opinion HFT is illegal, unethical and immoral given that they possess a split-second advantage from adjusting prices from my actual orders. They trade on advance knowledge of price behaviour.
9. I have also over the years placed more dependence on the closing price. What markets do during the day doesn't matter much. Volatility is created to suck people in. I don't want to be pulled into markets from intra-day volatility.
10. I stay away from diagonal boundaries due to intraday volatility. That's why I like horizontal boundaries.
11. Prices when plotted and viewed on a price chart can be seen to be either in a trend or in congestion. In the process of congesting, there is a tendency for prices to form geometric patterns. It is the composite of buying and selling that creates these areas of support and resistance......in the past it was collusion, but today there is coordinated behaviour 'like minded' and large moves require that the right people are prepared for that move....during congestion, big money comes in and creates tops/bottoms and consolidation areas and hence we get a geometry.
12. I don't believe that charts are predictive. I take exception to Elliot Wave in suggesting that you know where you are in terms of wave structure. I do think that 2-3 times a year, a chart pattern gives me a reason to trade. More often than not, there is not any chart setup. There is pattern or there is not a pattern. When there is a pattern there is a possible trade.
13. My win rate is 40% and the default position I take is that I will lose money.
14. When I enter a position, it get's complicated with many nuances. However if each trade over my history were to be printed out on a sheet of paper, then of that pile of papers, starting with the largest profits.......the top 10-15% keep me above breakeven. If I look at these 10-15% 'bottom liners', there is one thing common to almost all of them. They tend to work straight away and they never look back. They don't hesitate and they don't put me into trouble. If my timing is not right then my trade will not be right. You need to be right on direction AND timing. You must be right with both of them. I only risk <1% of each trade of my capital. I trade multiple positions and treat each position with different rules.......(Copernicus hint...diversification)
15. As long as the market is going in my direction and doesn't struggle, I will let it go. I do not use trailing stops.
16. You make decisions based on markets not based on equity. Trailing a stop to simply lock in profit is not the solution.
17. I don't know whether a trade will be profitable or not. I have no control. The only control I have is over whether I enter or not. Do the orders make sense. If I look at these trades 6 months from now will they still make sense. I still do stupid things and cannot purge myself of stupidity.
18. I just need to stay distant to my emotions and make sure my orders make sense to me....then the profits will just take care of themselves.
19. I am mid range between a swing trader and a position trader.
20. Whatever your approach is, it will be in synch with the markets at times and out of synch at other times. I don't want to change my approach just because the market is unforgiving at times. I only have a plan A. I don't pull Plan B out of the closet if the markets are unfriendly. I assume that that Plan A will work at times and not at others. It is what happens in those out of step phases that is important. The market is a marathon and not a sprint. It is a long term thing. I have traded in 5 different decades ad it is my goal to preserve my capital when things are tough. When I feel that the markets are messy I become more picky with my trade entries. You exercise diligence and the rest is in the details.
21. If you trade long enough, sooner or later you are going to make it. You can't help it. It is not the making of the money that is the challenge. It is the keeping of the money that you have made.
22. I have found that the traders I admire are the traders over the long haul (who are not the best)...but those traders than can keep their account together. A focus on trade identification and trade signalling is grossly overestimated in my opinion. Money management and keeping capital intact is where a trader needs to focus. Novice traders place way more emphasis on trade identification. The emphasis should be on money management, risk management and the process of trading being repetitive. Do the same things over and over again. Check your checklist and be very purposeful.
23. A common misconception is that you think you will become successful because you are just as smart as the other guys and if I worked hard and was studious it would bring me results.......they play a factor....but providence (luck) is a major factor....early on , a few trades enabled me to get to where I am. Those trades could have gone the other way. My outcome wasn't based on my intelligence or my work ethic.
24. The majority of traders never make it to the big time because they are afraid to lose. They are too concerned with being right all the time. You are going to be wrong a lot of times....don't be afraid. Assume it will happen. If you assume this, you will approach money management differently and will undersize your position. The other way ensures you oversize your position and that is where the problem lies. Other traders cannot believe the small size of what I trade. Most novices trade way too large a size.

Regards

C

Hi Copernicus,

With regards to the paper, I had similar thoughts to yours. On the subject of overfitting, I too thought that it would be prevalent among the users, and it isn't surprising that there is a correlation between the amount of backtests run and the amount of overfitting. This has been written about in depth by Marcos Lopez de Prado and others (citations at the end of the paper).

The conclusion by others that system performance degradation must be the result of markets becoming more efficient rather than from overfitting led me to a year of fixing the wrong thing. I also concur with the paper's authors that when a system falls apart it generally has much more to do with the way the system was created and the amount of overfitting than to markets becoming more efficient or anomalies getting arbitraged away. An exception would be a strategy that was presented to thousands at a conference (who all start trading it as-is) that then fails as too many trade it. Or the infamous Fap Turbo that had great performance but was a victim of its own success. I heard 2nd hand that Fap Turbo imploded because too many retail traders were competing for fills at the same time on a scalping EA, leading to its demise / unprofitability.

When I got to the conclusion of the paper, I had to go back and find the definition of tail ratio because I wasn't familiar with that ratio. It makes sense - compare the most extreme 5% of profits to the most extreme 5% of losses. A system that has a high tail ratio seems more likely to have positive upside surprises while a system with a low tail ratio might have some nasty negative loss series that could materially affect the P&L of a system going forward.

Maybe it is as you imply that when you look at the most extreme returns, the true nature of a system comes out. There may also be some relation between tail ratio and true Reward:risk ratio for a system. Perhaps tail ratio could encapsulate whether a system is likely to have good or bad luck once the system goes live. I just read an unrelated blog that basically said that in the short term (< 3 years), and even in the long term for low Sharpe systems, luck can dominate skill. So tail ratio may be a way of identifying good/bad luck that hasn't happened yet.

After I reread the tail ratio definition, I wondered if the rest of the distribution was compared (perhaps in 5% or decile slices), whether the other slices would also have predictive power. I guess that's a follow-up item.

It is also worth remembering that the paper mentioned that as the Sharpe ratio measurement period was shortened, that the predictive power increased. It's my thought from some testing evidence and some inference that this may hold true of other ratios and even shorter measurement periods than one year.

Im confident that any directional strategy worth running (read: robust) could be presented to the world and still succeed. It is the nature of markets to make sure of this. Unprofitable phases and drawdowns are the inbuilt mechanisms that shake out the weaker hands ensuring that a particular class of strategy never gets crowded for too long.
Good lord.

Hi FXEZ

Thanks for the response


Crumbs....I feel your pain.


It was only the other day on this forum that someone was wanting to know about the demise of Fap Turbo and why. Interesting.


It had me thinking about Kathryn Kaminski's and Alex Greyserman's work again "In Search of Crisis Alpha" in explaining where the dominant alpha arises from medium to long term diversified trend following strategies, namely from market crisis events and usually led by equity crisis events. The inter-correlation between asset classes during this time is quite pronounced leading to mass trend events across asset classes as traders/investors rush for liquidity. If you break down the returns of the long term CTA's that have demonstrated consistent performance with only minor tweaking of their composite strategies, you quickly realise that the majority of the time they are simply treading water. It is predominantly these fat tailed events that feed the traditional CTA segment and allow them to recover from persistent drawdowns.

Given that I am not familiar with other hedge fund segments that strive for alpha, their tactics, or their long term performance record, it is easy to assume that if you are striving for more than beta, your best option is to join these vanilla CTA lads in harvesting opportunities during these stampedes. Your article is certainly piquing my interest and making me wonder if simply staying alive in this market to be able to participate in these opportunities is really all that's needed for a robust edge. Obviously to achieve this your risk management during 'normal market conditions' must be extreme to avoid you becoming victim to effectively random 'noise'.


I am eager to find out more about this. Hopefully some more papers come to light shortly. My gut tells me the direction they are taking is a valid one.

I noticed this as well. I might have to dust off some incorrect assumptions about robustness of these ratios over the long term based on these findings.

Good find FXEZ. Thanks for sharing.

C

Simply Great.

ROFL....I agree that you can virtually guarantee that the trend following space will never be overcrowded. If flies in the face of human nature....thank god for trend following I say......... *sings* Give me land, lot's of land and the starry skies above........Don't fence me in :-)

It was on my third listen that I thought....Peter's simple words would remove the need for 3/4 of the noise of this forum. It is so easy to simply pass these things by. I almost did....but if you are lucky and take the time to digest it....you realise the gift he was giving......(thanks Bass for bringing him to my attention).

Cheers :-)

Are 3 year track records meaningful?

This is the blog post I referenced in the quote above. This post doesn't give a warm and fuzzy feeling about diversified programs that have Sharpe ratios < 0.5.


After watching the video recommended by bassramy regarding historical drawdowns and their occurrence in a long-only buy and hold the market "strategy" with drawdowns existing about 75% of the time, while only 25% of the time new equity highs were made, your conclusion may be on the mark, Copernicus. As a follow up to that study, and possibly of more practical interest to investors is estimating the probability of being in a drawdown for holding periods of say 1 year, 2 years, 3 years, etc., from any random entry point walking forward, answering the question, "If I buy here, what is the probability my position will be under water in X years' time?"

Thank you for sharing the Peter Brandt autobiography snippets. Very interesting reading. Thank goodness we don't have to do it the hard way anymore

Thanks for this :-)

Gulp....many of my automated strategies fall into this category....more work obviously needs to be done. I must admit that I am finding it very hard to outperform the discretionary strategies long term with full automation. It's a tough game. Good summary at the end of the article assessing asset management performance that sums it up.





Agreed. Just looking at the drawdown profile of my automated strategies, it is a rare occurrence to find that you are not in drawdown. No matter how you cut it, learning to live with and manage risk seems to be the name of the game. Those long term survivors (such as Brandt) might put their success to luck...... but when you dig a bit deeper, it was that they were simply around to participate when the opportunity struck. To achieve this I think you always need to be in the game at all times....but never exposed. That therefore necessitates a priority towards risk management at the expense of all else. The performance results as a consequence will just flow from this IMO.

No probs dkrock. :-)

Example 1A - Gold (XAUUSD)

Update - Minimising Losses

.....and it looks like this trade is not going to get anywhere.....but the fat lady is yet to sing.

However....look what has been achieved using the SDC trailing stop. The SD trailing stop has ensured within statistical limitations that my trade management must take place within the trend. It has also imposed a time limitation for breakeven and beyond to be achieved in accordance with historic momentum in the trend. The worst case scenario now is that I lose 0.1% of my trading capital as opposed to the 0.25% applied to the initial stop and associated position sizing.



If price touches the SDC it tells me two things:
1. exit immediately (cut the losses as short as you can)....but more importantly......
2. statistically the definition of my trend is over so my stop is optimised according to the available statistics. No intuition, no gut judgement...just process.

Do I worry about it....of course not!!!! It's a pin prick I tells ya. Where are you rover....come here fella.

its a start..

good stuff!

Woohooo. MS Access here we come. Great stuff 21 (or should I say J). Access is a great data warehouse solution and a repository for trade histories. You are going to need it if you are using M1 long range data.

PS Have been speaking with an old mate (ex collaborator) of yours who is practically my next door neighbour (in global terms).

PPS Diggin' the vibe from the Boiler Room. That goes straight into my You-tube converter........ if I can stop dancing on my chair.

I am exploring atm. Ill see what connections I will find. Definately writing the logic will be very challenging. I set up the starting point to see if I can manage to do something, otherwizze, ill try another work-out of mt4 and into some other multicurrency multiTF environment.

Yes, Freds a legend. We started with version one and I have to say he is class one programer. Im in v3 atm, but v1 still working on the comp 2,5y ++ on forward testing. I saw you also worked together if im correct. Nevertheless, say hello if you meet eachother downunder

hahaha lol, yeah my collegues also watch me move on the chair Few times on Carl Cox already.. Best time was on Ibiza Space! Crazy movement..the guy was awesome also. Happy faces
back to work!