Developing a mechanical system from scratch

[captain black]

I read a paper the other day to do with volatility, basically for trend following you are better off with low volatility stocks compared to higher volatility. Can't remember where or who it was though.... not real handy lol

Is this the one from Connors research?

(pdf download link)

http://connorsresearch.com/wp-content/uploads/2015/01/CR_Historical_Volatility.pdf

 

[Roller_1]

Is this the one from Connors research?

(pdf download link)

http://connorsresearch.com/wp-content/uploads/2015/01/CR_Historical_Volatility.pdf

Yep, that's it

 

[Gringotts Bank]

Came across this in my travels.
Search "soy" in...

https://authoritynutrition.com/optimize-omega-6-omega-3-ratio/

 

[Wysiwyg]

From the Connors Research paper.

What is price? It’s essentially a culmination of all the known information of a company. Therefore, price is real as it represents what the majority of the market participants know at that given time.

Not wholly true! Price is also an anticipation of better (or worse) future company success, mediocrity or failure. In other words unknown information. Not the "majority of the market participants" either but the majority of securities held in one place.

 

[CanOz]

Came across this in my travels.
Search "soy" in...

https://authoritynutrition.com/optimize-omega-6-omega-3-ratio/

Dead link....

 

[Gringotts Bank]

Dead link....

https://authoritynutrition.com/optimize-omega-6-omega-3-ratio/

works for me.

 

[Newt]

I read a paper the other day to do with volatility, basically for trend following you are better off with low volatility stocks compared to higher volatility. Can't remember where or who it was though.... not real handy lol

Nick Radge made mention of this on his Facebook page recently if I recall correctly. Someone (Alverez?) had commented on combining Radge's Weekend trend trader with volatility screening/metrics. Its an interesting idea. For my longer term trend following there is a lot to support gut feel for less volatile trending stocks (per unit price).

Suspect ATR/price is just the tip of the iceberg though Canoz - probably need someway to incorporate linearity /correlation and rate of change in price. Arguably then too much at risk of lagging moving average sort of situation. Will have to play more in Amibroker when some quiet time comes along. Might add some value/edge to new high screens perhaps?


p.s. Seem to also recall Pixel commenting previously on using %ATR successfully in his trading over many years. Would have to go digging for the thread though. Not something I succeeded in distilling into a useful approach last time I played with it - which is of course usually the way until you invest a lot of effort, experience and thought into these things!

 

[CanOz]

Here's an interesting statistic, after 500 minutes of morning trend, there is a 62% chance the that the 6A will reverse the trend. No Stops, Average trade is $300. Exit is the reverse signal.

This system did not work before 2009.

 

[CanOz]

This Dax momentum system trades just after the open and holds through until the next day. The ATR% filter will shut this down soon. I'll have the filter for this tomorrow. It should take the chop out of the equity curve, as well as some of the best trades.... a smoother curve is what we want.

 

[Roller_1]

What sort of % returns and DD are you looking at with these system CanOz?

This one looks to hve some deep and sustained DD's. Could be tough to trade

 

[Lone Wolf]

Here's an interesting statistic, after 500 minutes of morning trend, there is a 62% chance the that the 6A will reverse the trend.

500 minutes (8.33 hours) of morning trend? Typo?

 

[Trendnomics]

Nick Radge made mention of this on his Facebook page recently if I recall correctly. Someone (Alverez?) had commented on combining Radge's Weekend trend trader with volatility screening/metrics. Its an interesting idea. For my longer term trend following there is a lot to support gut feel for less volatile trending stocks (per unit price).

Suspect ATR/price is just the tip of the iceberg though Canoz - probably need someway to incorporate linearity /correlation and rate of change in price. Arguably then too much at risk of lagging moving average sort of situation. Will have to play more in Amibroker when some quiet time comes along. Might add some value/edge to new high screens perhaps?


p.s. Seem to also recall Pixel commenting previously on using %ATR successfully in his trading over many years. Would have to go digging for the thread though. Not something I succeeded in distilling into a useful approach last time I played with it - which is of course usually the way until you invest a lot of effort, experience and thought into these things!

My trend-following system is constructed from two custom indicators (self designed/constructed - unaware of any other indicators that are similar). One of the indicators is a trend-strength + volatility indicator. Trading high volatility is perfectly acceptable, given the volatility has consequentially resulted in a comparatively strong trend.

 

[CanOz]

What sort of % returns and DD are you looking at with these system CanOz?

This one looks to hve some deep and sustained DD's. Could be tough to trade

Yeah this one had a doozy at the start, 30% plus. I don't plan on trading it live until i can smooth things out a bit. If that means giving up too much in returns then i'll bin it. I haven't done anything with annual returns of any sort yet.

 

[CanOz]

500 minutes (8.33 hours) of morning trend? Typo?

Nah its literally 500 minutes of trend, now that could be all kinds of different price patterns. This one is based on a certain number of higher closes.

 

[Newt]

My trend-following system is constructed from two custom indicators (self designed/constructed - unaware of any other indicators that are similar). One of the indicators is a trend-strength + volatility indicator. Trading high volatility is perfectly acceptable, given the volatility has consequentially resulted in a comparatively strong trend.

Agree there will usually be expanding volatility as the trend accelerates, but the gist of the Connors PDF referenced earlier was that (supposedly) over the longer term lower volatility stocks give more reliable gains that those with high volatility. I haven't been able to get my head around the volatility measures they talk about - may even include intra-day volatility if I've read it right, but not ATR or similar many of us grab when "thinking volatility".

Glad to hear your metrics are working for you trendnomics. I personally think honing a few well thought out indicators/calculations can really help you explore different dimensions of YOUR trading system as you move from modelling to developing confidence trading it. Certainly for the way my mind seems to tick.

 

[Wyatt]

Nick Radge made mention of this on his Facebook page recently if I recall correctly. Someone (Alverez?) had commented on combining Radge's Weekend trend trader with volatility screening/metrics. Its an interesting idea. For my longer term trend following there is a lot to support gut feel for less volatile trending stocks (per unit price).

Here is the link http://bettersystemtrader.com/037-cesar-alvarez-studies-stop-losses/

Suspect ATR/price is just the tip of the iceberg though Canoz - probably need someway to incorporate linearity /correlation and rate of change in price. Arguably then too much at risk of lagging moving average sort of situation. Will have to play more in Amibroker when some quiet time comes along. Might add some value/edge to new high screens perhaps?

Apparently, Volatility (ATR) based position sizing for trend following systems is a 2 edge sword, it will lower your drawdowns but also your returns at times.

Trendnomics system gives his stocks lots of time and room to move, so that when the inevitable dips occur, the loss or reduced profit is not taken at that time if the dip recovers within the shorter term. The attached chart shows some of the benefits of such concepts. Interestingly, on this model, the Feb '16 D/D was more acute than any other during the period.

View attachment 67763

 

[CanOz]

Suspect ATR/price is just the tip of the iceberg though Canoz - probably need someway to incorporate linearity /correlation and rate of change in price. Arguably then too much at risk of lagging moving average sort of situation. Will have to play more in Amibroker when some quiet time comes along. Might add some value/edge to new high screens perhaps?!

Sorry Newt, just saw this ...

I like ATR% because for most of my momentum systems, that are equity index, they rely of volatility to drive the big moves. The problem with indices like the Dax and the NK is that they get expensive as well as volatile. I just want volatile. With a filter then yeah of course its lagging and i'll miss out on some...only testing will tell if that still keeps things trade-able for me...

I did notice i had one system that actually benefits from lower volatility, i'll have to go through my notes again to see which was it was, but i'll be able to apply my filter to that as well.

 

[CanOz]

Here is the link http://bettersystemtrader.com/037-cesar-alvarez-studies-stop-losses/



Apparently, Volatility (ATR) based position sizing for trend following systems is a 2 edge sword, it will lower your drawdowns but also your returns at times.

Trendnomics system gives his stocks lots of time and room to move, so that when the inevitable dips occur, the loss or reduced profit is not taken at that time if the dip recovers within the shorter term. The attached chart shows some of the benefits of such concepts. Interestingly, on this model, the Feb '16 D/D was more acute than any other during the period.

View attachment 67763

I think it was Perry Kaufman on bettersystemtrader that was talking about using position sizing to improve performance of good performing system. So not to fix a poor system but to smooth returns. He was suggesting however, that a normal size should be used in volatile times and a larger size in less volatile periods. This has the effect of creating volatility in returns when it is needed, and less so when its not.

 

[Newt]

Here is the link http://bettersystemtrader.com/037-cesar-alvarez-studies-stop-losses/



Apparently, Volatility (ATR) based position sizing for trend following systems is a 2 edge sword, it will lower your drawdowns but also your returns at times.

Trendnomics system gives his stocks lots of time and room to move, so that when the inevitable dips occur, the loss or reduced profit is not taken at that time if the dip recovers within the shorter term. The attached chart shows some of the benefits of such concepts. Interestingly, on this model, the Feb '16 D/D was more acute than any other during the period.

View attachment 67763

Thanks Wyatt
And here's the link to Radge's page I was thinking of:
https://www.thechartist.com.au/Better-Trading-Series/make-it-your-own.html

(disclaimer - not currently a subscriber, but definitely an admirer of The Chartist )

Linda Bradford Raschke often talks about markets transitioning between low and high volatility. Makes sense shorter term trader will seek out Volatility, while long trend traders may prefer entering on low (but increasing or soon to be increasing) volatility.....

 

[Newt]

Sorry Newt, just saw this ...

I like ATR% because for most of my momentum systems, that are equity index, they rely of volatility to drive the big moves. The problem with indices like the Dax and the NK is that they get expensive as well as volatile. I just want volatile. With a filter then yeah of course its lagging and i'll miss out on some...only testing will tell if that still keeps things trade-able for me...

Interesting to hear your take on shorter term index trading utility too Canoz...