Page 1 of 1

Important - Know when to break the rules

Posted: Wed Sep 04, 2013 7:37 am
by Chris67
I think one of the problems I suffered with over the years [ apart from all the rest yes I know] is I have been relatively stubborn in my marriage to some of "those rules" you just dont break as a trend followr / systematic trader.

One of those rules is sticking to your system when times get tough - I thought I 'd share my experience - it maybe a valuable exercise if I get some feedback here of what others have done and found ?

I had 2 systems - both robust. I say robust because one of them was a 1 [effective] paramter system [ maybe 2] - trade a large bask of futures on the Bollinger band break-out system - use something like a 200 day MA and a 1 standard deviation break-out
Not only was this a system that by mid 2010 had performed very well - but it also seemed to have a steepening equity curve unlike so many basic systems that had flattened off a tad !

Anyway the worst drawdown it had suffered was about 25% and at 35% I bailed out sensing something just want at all right

Every "guru" , every TF and people like Mr Covell Im sure would argue how not sticking to your system and bailing is the wrong thing to do !
I just checked teh performance of this system - its now in a 92% draw down - all criterai totall blown away from anything that had happned in teh previous 30 years

Secondly the TMA which I have already written about on here - i bailed on that two having traded it for many years and made good money on it
a medium term system traded on a large basket of futures with good risk control in place and heat limiter saw this system totally disintergrate into oblivion.

When all is said and done it is actually amazing that if you look at some turtlish type systems traded over a much larger portfolio than the turtles used and without the aggressive "adding" they used - that in fact these have held up far better and are showing no real signs of degradation ?

its all a learning curve and all food for thought - can we get a discussion going here beacuse maybe I'll learn what I did wrong [ if anything?] maybe you'll learn what you did wrong / right and maybe we can share some ideas !!!!

Posted: Wed Sep 04, 2013 8:50 am
by sluggo
Be sure to include a few paragraphs in your Disclosure Document which alert potential customers / investors in your Fund. You can be (will be!) sued if you deviate from the Disclosure Document and investors suffer losses. Perhaps something like,

  • The Portfolio Manager invests the Fund's assets using several systematic trading strategies, which the Portfolio Manager firmly believes are robust and have stood the test of time. However, unusual market conditions may arise which cast doubt upon this belief, and consequently the Portfolio Manager may abandon the systematic strategies at any time. Investors are warned that if the trading strategies change, the Portfolio Manager's previous track record becomes irrelevant and moot. The old strategies he traded and then abandoned, do not predict the performance of the new strategies he trades afterwards.

Posted: Wed Sep 04, 2013 10:20 am
That Sluggo needs a new hobby. It is perfectly acceptable to adjust, modify, or introduce new strategies within a systematic framework. Now if your disclosure documents say diversified systematic and then you decide to trade nothing but corn based on weather patterns then........

Posted: Wed Sep 04, 2013 10:42 am
I don't think you did anything wrong at all. I also "turned off " my intermediate trend system in 2012 because the data indicated clearly it was out of its reasonably expected risk/return profile.

Sure saved me a lot of money and in the end that is the most important measure.

Long-term trend system while below trend for past two years is still within expectations and mean reversal systems are above expectations.

One needs to be very adaptable to survive in trading and have multiple systems to handle various markets. Option strategies, for example, have been outstanding for past two years for the same reason that intermediate trend systems are horrid.

Posted: Wed Sep 04, 2013 1:02 pm
by Chris67
We got that covered dont worry !!!

I think many systems is the answer and once again a pure understanding that 30-50 years of data is simply not enough !
Having said that 30-50 years puts us ahead of most for sure

Posted: Wed Sep 04, 2013 9:31 pm
by trending3029
Thanks for bringing this up. I've had similar instances where I've had to turn off systems. What I found interesting Tom Basso's comment in a podcast
" I would shut myself in a room and ask my secretary not to bother me unless there was something urgent and I would take a look at the market and the system and check with the given market condition, how did the system perform." Also as per reading trading systems methods book, what Perry Kaufman demonstrates is that the market is constantly changing, he measures noise levels in the markets. Therefore exactly understanding the market, system and thinking what could possibly happen if market changes (preempting changes such as increase in volatility) and how the system would react and testing the possible change may be way to go. I guess this seems like fluff and subjective and yes it indeed is, cause I am in a similar situation as yourself and in the process of articulating my thoughts and going back to drawing board to formulate strategies.