Diversified Trend Trading Approach

After downgrading possibility of rate hike in US, dollar is being sold. Volatility is coming back.

Why people need talking heads on television to tell explanations/stories why gold is up and dollar down, and oil down or up? CNBC or Bloomerg are just story telling nannies to children. I see this bias/craving in my own mind...always looking for explanation

http://lifehacker.com/5965703/the-sc...ate-our-brains

https://www.ethos3.com/2014/10/the-n...-storytelling/

Great article R....The Neuroscience of Storytelling;-)

EDTT - Preliminary Results of Medium Range Tests - Discretionary Systems

.......continued from post 2029

After fighting with my telecommunications provider (Telstra) from an outage of my cable service since 7 Sep....I am back and have now managed to backtest another 2 inclusions in the portfolio:

1. CAC40; and
2. AUDUSD.

All I can say is .....yuck.....and without my dose of Youtube music to keep my sanity. These problematic beasts seemed to read my mind and defeated me regularly in my attempts to apply momentum breakouts. No matter....that's just the way it goes. Let's have a quick squiz at their individual results and see what damage they then do to the portfolio.

CAC40

AUDUSD

Attached Image (click to enlarge)

Attached Image (click to enlarge)

Now to any normal trader, these results would appear to be abject failures. A trader using the EDTT on these instruments would conclude that the strategy has no edge and would likely throw it in the bin and strategy hop to another......however what simply needs to be remembered is that market conditions for the timeframe under back-test were simply unfavourable to the deployment of this strategy. This however is not always the case as discussed in many of my previous examples.

Before we conclude that the strategy is a dog, we must test it under portfolio conditions and discern the impact upon the portfolio of these poor performers. So let's have a look at the damage.

Prior to inclusion


After inclusion

Attached Image (click to enlarge)

Have a close look at the before and after performance metrics above. The inclusion of these horrid performers using a 7 x position sizing multiplier has reduced CAGR from 65.6% to 63.4% and the max drawdown from 11.3% to 12.0%. That's right....there has been hardly any material impact to the portfolio. This is one of the significant benefits of trading a diversified portfolio. The blend significantly reduces the impact of individual detrimental performers. This is one critical aspect of risk management that is absent when devoting your trading efforts to a single trading instrument.

Risk Management
As we are getting to the tail end of this portfolio blending exercise with 20 out of 24 liquid instruments concluded I am going to turn down the portfolio heat once again using a 6 x position sizing multiplier to keep the max drawdown of the back-test at a stable level of around 10%. This is going to have to be further tweaked at the end of the entire exercise as we need to embed assumptions regarding SWAP and slippage into our return distributions to approach a more realistic live performance level. The intended result will be a max drawdown tolerance of 10% including the impacts of SWAP and slippage.

So here are the revised results so far for the portfolio.

Annual Results


Summary


Portfolio Composition Return Profiles



Stay tuned...... :-)

20 instruments down now only 4 to go.

C

Food for Thought

Attached Image

 

1. When High Yield Becomes Low
2. Valuation and Volatility
3. The Power Of Compounding: You Can Achieve Anything, If You Stop Trying To Do Everything
4. Scientists discover there are 27 different emotions
5. Businesses are finding previously undetectable data trends using location intelligence
6. Why Do We Believe Things?
7. The Logic of Risk Taking
8. Harvey, Irma, and the Future of Weather Volatilty
9. Why You Diversify
10. ZeroMQ – How To Interface Python/R with MetaTrader 4
11. Hedge Fund Performance – Growth, Macro, and EM Best Strategies
12. How good are your entries and exits…really?
13. How to quickly identify when an edge has gone
14. Software millionaire fooled by ‘the dream’ and saved by global macro w/ Erik Townsend
15. "Always be paranoid about [your strategies], and always be paranoid about every part of it."
16. “Risk Cannot Be Destroyed, Only Transformed”
17. Confident Traders

Enjoy

C

That's a lovely and rich list and nr 12 got my attention
thank you for sharing

EDTT - Lining the Ducks Up - The Logic of the Application of Momentum Breakouts

Since it is a lazy Saturday and given my rising faith in the EDTT following my backtest efforts, I thought now may be a good time to explain some of the most recent amendments that are occurring as a result of my back-test findings.

You should all be aware from the performance metrics of the back-test that this strategy is now being applied at the short term horizon with an average holding period of about 12 days. The rationale being applied with this technique is that we are searching from definitive momentum breaks with low holding costs. What I mean by this is that momentum is either in your favour or it is not. As soon as it is not...defined by our medium to short term SDC's, then we are immediately out of the trade with a <1xR loss or better. Our short holding period ensures we also keep SWAP to a bare minimum.

Unlike most trend following strategies that give room to breathe, this particular technique is extremely selective about when to enter and how long to hold. We always keep an open-ended profit target however as you never know how long a momentum burst will last for.

I have recently enjoyed an excellent momentum burst short in USDSGD in my live walk forward testing and thought I would explain the rationale for this trade and how I did it.

Here is my timeframe setup for EDTT for a single instrument. In this case USDSGD. While I restrict my trades to M30, the multi-timeframe approach helps me see the wood for the trees in terms of momentum.

Let's start with interpreting price action in terms of momentum on this multi-timeframe perspective. This get's me warmed up and ready to strike on the short term for a momentum move with reduced possibility for a 'false breakout'.



On the monthly chart MN and the weekly chart W1 we see that price is testing it's longer term bullish momentum. The longer term investment community would be seeing this as a possible trend breakdown and would be avoiding long entries at this stage of the game waiting for either a retracement back into the prior bullish phase or preparing to go short once the trend change has been confirmed. The shorter term traders would be looking for a nice early entry into a possible short trend at this important inflection point. Selling momentum on these timeframes would be progressively building at this point with attempts by the bulls to hold back the impending pressure.

On the D1 timeframe we see excellent consistent momentum that sits within the SDC with no single bar or period overly biasing the general momentum surge. This is a good sign that we are not seeing shenanigans from the institutions holding back the surging tide of short selling momentum. It augers well for a smooth ride ahead with further selling momentum.

On the H4 and H1 timeframes we see that in the context of current momentum we are not in an oversold position. Price currently sits smack bang in the mid range of the SDC. This is a further line-up of a necessary duck that offers potential for more significant downside momentum......in other words there is the opportunity for a future short surge in momentum before we see a desperate struggle by the bulls to hold back the tide.

Ok....so now I am very interested in looking for a short term breakdown on M30 where I can capitalise on a short term momentum burst.

Let's go to M30.



We are at 6/9/2017 and our pending order short is about to be triggered. We have our multi-timeframe perspective in hand and we are looking for a nice short momentum burst.

The first thing we do is define our longer term momentum swing. In this case Options A and B in the chart above offer this opportunity. With less conviction, I would have preferred A....but with the multi-timeframe conviction I was pitching for a nice sharp breakdown. I therefore chose B as my swing high point to commence my SDC channel from. In retrospect...that was probably unwise and I should have opted for pt A as there were more bars of data confirming the overall bias in momentum......anyway.....neither is wrong....but A would be preferable.

Now as discussed previously in my posts, I always look to a shorter term momentum push to carry the longer term momentum drive in the correct direction. If there is no short term momentum push then I do not take either trade and simply extend my pending order further south until a shorter term momentum push is detected. This significantly helps in filtering out 'fakeout's'.

So in looking for my short term momentum push to give conviction to my longer term position, I look for recent swing highs to define my short term SDC. If the slope of the short term SDC is significantly greater than the slope of the longer term SDC.....then that is sufficient confirmation for entry on the pending order.

Now that I am locked and loaded with two trades on entry at 6/9/2017 at 16:00 (a short term SDC and a long term SDC) I initially manage the trade on the shorter term SDC until momentum is exhausted and I am stopped out with my trailing stop.....and then focus on managing the longer term SDC trade.

As you can see in this example the short term breakout had significant momentum behind it and I was able to achieve 2.5xR whereby I reduced my short term SDC to 0.5 and capitalised on a good quick profit. I have now exited the short term trade and can relax and enjoy the fruits of the longer term SDC.....wherever it heads.

In summary I am using a combination of multi-timeframe logic and the power of statistics through the SDC to filter the noise and hone in on what appears to be genuine price momentum in the short term.

Cheers

C

Cheers J :-)

I've got that Saturday club itch :-)














Time to crank it up :-)

EDTT - Preliminary Results of Medium Range Tests - Discretionary Systems

.......continued from post 2044



Yahoooo!!!!!......The backtesting is drawing to a close with 24 liquid instruments now complete. The entire process took close to a month commencing on post 1932 on 19 August 2017....but would have taken eons if I hadn't been introduced to a great piece of backtesting software that is applied to discretionary systems.

Below are the raw results but over the next week or so we will be delving into the analytics in a bit more detail to showcase some important key metrics and challenge some of the assumptions and mantra's you may have heard on this forum. There is nothing like a bit of solid data to sort out the wheat from the chaff.

Also we need to convert performance metrics to R values so we can scale our portfolio to our liking.

With that said and done....we have a very powerful beast in the EDTT at least over the last 7 years or so with what appears to be a very solid edge.

We now have to tone things down and add a bit of healthy reality into the performance statistics to approach more realistic levels. Our key focus will be on conservatively incorporating SWAP and slippage costs into the equation. The SWAP fortunately won't be a significant burden given the relatively short holding period....but the slippage will. Breakout strategies in particular have slippage to deal with on an ongoing basis given the relatively high momentum in action during the point of entry. We will be factoring in a slippage on entry and a lower degree of slippage onto the exit to simulate realistic conditions. Once that is all concluded then we will be focusing on appropriately risk weighting the portfolio to keep drawdowns within acceptable levels of approximately 10%. .....anyway this will be the subject of the following weeks posts.

For now enjoy the raw results that you are now familiar with.

Annual Results


Summary


Portfolio Composition Return Profiles


Sector Performance


Stay tuned...... :-)

More crunching to follow over the next week or so

C

Dear C

Great work as always!

Of zero importance, but in order to be as perfect as possible, please see below

T

Cheers T. Good eyes. I will amend

Number of days means trading days or calendar days? 2782/1176=2.36 trades a day. Still pretty active portfolio.

All the people saying trend following is dead, probably did not research it.
From 1970 divergence in markets is still the same:



And trend leakage is still a thing:

2782 calendar days...which is about 1935 trading days. With 1176 trades that's about 0.6 trades a day. Enough time to sip your martinis. :-)

This week I will have a look at max trades in a given day just to ensure we aren't over-doing it....but it sure seems pretty manageable with 24 instruments.

Spot on R. That's a nice extract.

I was picking this up in the 54 trend following funds that I monitor (below). The regression line that represents CAGR is remarkably consistent. What is noticeable however is the widening spacing between divergent market periods inter-dispersed with protracted periods of mean reversion. In my mind this supports the notion that what is dampening volatility is central bank intervention represented by lengthy mean reversion periods until the market periodically snaps back to equilibrium. The next snap coming up looks like it is going to be particularly aggressive.......well here's hoping :-)




PS. Have at look at this R. I have lined up the EDTT equity curve with the TF 54 curve to see where the weak spots are. This gives me confidence that we are onto a cracker. Look what happens when TF takes off into the stratosphere. So does EDTT. Look when the TF community are in drawdown.....we don't suffer drawdowns to the same extent, probably attributed to our very selective trade entries. Anyway.....food for thought. :-)

Anynone wants a cracker?



Summary table will be wallpaper on my computer, because this is such a beautiful view!

and something to cheer you up!

Just a reminder, that markets can do anything....

Gold was looking good on Friday

Owwwww. Don't you just hate those gaps :-( It's like being poked in the eye with a blunt stick.