Group Collaboration System Ideas

[JulianCohen]

I have a monthly momentum system which gives about 20% CAGR with 30% MDD. From my investigations that is about par for the course with a longer term system and bearable for me. If anyone can guide me to one with a better MDD I’m all ears.

[Anonymous]

i have been laid out with the “worst flu ever” over the past week so have done nothing in that time.
other things ive been working on are:
– MR system over the asx
– position sizing. i keep coming back to this to see if i can make it work… it seems hard to find consistent results

[JulianCohen]

What’s your problem with position sizing Darryl?

[ScottMcNab]

Julian Cohen wrote:

I have a monthly momentum system which gives about 20% CAGR with 30% MDD. From my investigations that is about par for the course with a longer term system and bearable for me. If anyone can guide me to one with a better MDD I’m all ears.

Have you tried the MOC systems on weekly or monthly time frames Julian ?…I have found some of them to give CAR/MDD of between 1 and 2

[Anonymous]

Julian Cohen wrote:

What’s your problem with position sizing Darryl?

just experimenting with something like this:

Code:

formula = BuyPrice/ATR(10);
SetPositionSize(formula,spsPercentOfEquity);

[SaidBitar]

Darryl Vink wrote:

Julian Cohen wrote:

What’s your problem with position sizing Darryl?

just experimenting with something like this:

Code:

formula = BuyPrice/ATR(10);
SetPositionSize(formula,spsPercentOfEquity);

But in this case all the time you will have only one position (if you enabled allow position shrinking) or maybe zero (if you did not enable allow position shrinking)
because the BuyPrice normally is bigger than the ATR(10)

[Stephen James]

Looks like you may be position sizing after the fact as well – i.e. at EOD on the day you have already bought.

[Anonymous]

i blame my stupidity on the flu :sick:

[JulianCohen]

Scott McNab wrote:

Julian Cohen wrote:

I have a monthly momentum system which gives about 20% CAGR with 30% MDD. From my investigations that is about par for the course with a longer term system and bearable for me. If anyone can guide me to one with a better MDD I’m all ears.

Have you tried the MOC systems on weekly or monthly time frames Julian ?…I have found some of them to give CAR/MDD of between 1 and 2

I’ll have a look. Thanks Scott

[Anonymous]

Thanks Scott. I tried the Friday close on a weekly mean reversion type strategy. It works and has decent results, not as good as the daily results though. It will probably be too closely correlated to my daily mean reversion strategies so I keep on trying to come up with a trend following / momentum strategy that I am happy with. More days of “Let me try this idea, it sounds like it will work” and walking away at the end of the day with another great idea that didn’t live up to expectations.. Then trying again the next day, an accumulation of the ideas will surface at some stage.

Next 1000 ideas… :ohmy:

[ScottMcNab]

Very similar process I have been going though…weekly MOC systems ok but not as good as daily MOC so only live trading daily… have been trying to come up with other weekly systems but as of yet not been able to find one as good as the weekly MOC

[SaidBitar]

Regarding trend following/Momentum you can look at WTT / BBO they are really nice systems. But if you are expecting the returns to be comparable with the mean reversion strategies then i don’t know if there is any system that will give such high CAR and low DD.

the expected returns of TF / Momentum as per the books of Nick and Andreas Clenow you will notice there is not huge difference in the expected returns regardless if it is stock strategy of even futures.

So from my point of view if the returns are CAR between 20-25% and DD up to 30% winning percentage >= 40% then it is good strategy.

I trade the WTT on Russell 3000 i started trading it around April 2015, untill March/April 2016 the strategy was in DD but when the trends started on some stocks it started to catch up and now (OPEN profits) are enough to cover all the realised losses from last year and this year and to have some profits on top.

the intriguing question is why to trade TF when MR returns are much better, i do not know the answer and honestly i have no idea of any correct answer but i did the following i ran backtest on MRV strategy and TF strategy over the same period of time and i assumed that i allocated 50:50 between them and i combined the results and made graph for each year return (TF returns as % , MRV returns as %, and combined returns as %) and the results is somehow convincing of how the two styles compliment each other.

so for ideas, any simple idea of breakout to a new high or channel BO and follow it by trailing stop with Nick’s magical pull up during the periods when the index filter is down you will have a good TF strategy.

if you don’t want this long term strategies and prefer shorter term ones (swing trading style) [“trade like a hedge fund”] has some nice ideas. I am not sure they will work anymore as far as i remember i tested some they were not profitable but at least will give a starting point.

[Nick Radge]

Quote:

the intriguing question is why to trade TF when MR returns are much better, i do not know the answer and honestly i have no idea of any correct answer

The answer is DIVERSIFICATION.

Real world example; my main US mean reversion system is about +1% for the year yet my ASX trend system is +24% for the year.

Professional fund managers trade poor systems or poor markets because they still provide diversification over time.

Crude Oil was a very good example in the last few years. Many CTA managers have decided that trading commodity trends on the short was a losing proposition so they only trade the long side. Along comes Crude oil which trended smoothly from $110 down to $25. CTA’s that missed that trend lost money for the year and CTA’s that did catch it had a bumper year.

Also remember what your first reaction is – because usually your first reaction (or the reaction of most amateurs) is always wrong.

The first reaction is ‘I won’t trade a trend system when a MR is so much better’

[ScottMcNab]

Miguel Teixeira wrote:

Thanks Scott. I tried the Friday close on a weekly mean reversion type strategy. It works and has decent results, not as good as the daily results though. It will probably be too closely correlated to my daily mean reversion strategies

Next 1000 ideas… :ohmy:

Just looked at correlation of monthly returns between daily MOC and weekly MOC system between 1/1/96 and present and correlation was 0.3. Worth checking as my excel skills are as “good” as my amibroker but the result surprised me

[Anonymous]

Hi said I was wondering if you could do me a favour and send me the link you had set up for the excel sheet for optimisation I cant seem to find it on the forum it would be greatly appreciated.
Thanking you Craig

[Author]

Posts