{image}
Jan 1996 to Jan 2010
Source:
As a value investor Buffet's strategy suits convergence where diversification doesn't work. With divergence, diversification is essential.
Diversified Trend Trading Approach
- #4,241
- Edited 10:35pm Jul 4, 2019 10:09pm | Edited 10:35pm
- | Commercial User | Joined Apr 2013 | 4,366 Posts
Jan 1996 to Jan 2010
Source:
As a value investor Buffet's strategy suits convergence where diversification doesn't work. With divergence, diversification is essential.
- #4,242
- Jul 4, 2019 10:48pm Jul 4, 2019 10:48pm
- | Commercial User | Joined Apr 2013 | 4,366 Posts
How Diversification of Divergent Systems Smooths the Volatility of the Equity Curve.
- #4,247
- Edited 3:17am Jul 8, 2019 3:07am | Edited 3:17am
- | Commercial User | Joined Apr 2013 | 4,366 Posts
For Price Following Systems that operate off the philosophy of market divergence, your success is attributed to a handful of trades. The edge lies in the 'Positive Skew'. Keep to the right hand side of the market distribution by letting profits run with a trailing stop and no profit target and make sure the left hand side is managed by cutting losses short.
You cannot predict the 'when' of a large outlier.....so don't bother. You must catch all trends to participate in the occasional unpredictable large outlier. Type 2 errors (missing the big one) is the greatest sin....so avoid retracement entries where you can miss a few and stick to breakouts, where you can catch them all. False breakouts (the whipsaw) is something you just gotta live with.
To achieve this:
You need this:
In other words....."your success is defined by a handful of large positive outliers".
You cannot predict the 'when' of a large outlier.....so don't bother. You must catch all trends to participate in the occasional unpredictable large outlier. Type 2 errors (missing the big one) is the greatest sin....so avoid retracement entries where you can miss a few and stick to breakouts, where you can catch them all. False breakouts (the whipsaw) is something you just gotta live with.
To achieve this:
You need this:
In other words....."your success is defined by a handful of large positive outliers".
- #4,251
- Edited 7:16am Jul 8, 2019 7:06am | Edited 7:16am
- | Commercial User | Joined Apr 2013 | 4,366 Posts
ohh i just noticed since 2016 you DD is also far less, that could be the reason of low gains since 2017..... were you doubling up on your looser since 2000 till 2015 as trade management way ?
The strategy (displayed above) is a composite of 150 trend following systems across FX (Majors and crosses plus XAUUSD and XAGUSD). Single trades, no pyramiding or averaging. Just simple strategies that cut losses short and let profits run....but lot's of them.
It has been a better story for FX since 2014 however than for the CTA's as a whole. We picked up some good trends in 2015 and 2016 while the CTA's stagnated as a class.
- #4,254
- Edited 8:09pm Jul 8, 2019 6:06pm | Edited 8:09pm
- | Commercial User | Joined Apr 2013 | 4,366 Posts
Morning N :-)
For that example, I was using fixed lots across the time series according to the following sizing (in yellow below). This is all traded on the Daily timeframe with an average hold per trade of 30.2 days. No pips are targeted. Every strategy is unbounded with trailing stops....no profit targets are used. I am hunting anomalies with these techniques.
No martingales used. Each strategy cuts losses short and lets profits run with full risk release each trade. The portfolio comprises 21 separate sub portfolios (markets). Each sub portfolio has up to 7 separate strategies.
That's about it in a nutshell mate. A small edge applied across Large Numbers. 
{quote} what position sizing you are using?? are those size constant % of your account per trade ?? around what TF you are trading and approx what pips are targeted to achieve keeping in mind ??
No martingales used. Each strategy cuts losses short and lets profits run with full risk release each trade. The portfolio comprises 21 separate sub portfolios (markets). Each sub portfolio has up to 7 separate strategies.
{quote}last question how do you manage your trades ??? is little edge with n numbers of trades over time (which is getting averaged) is only reason in growth of your capital ?
- #4,255
- Edited 6:30pm Jul 8, 2019 6:18pm | Edited 6:30pm
- | Commercial User | Joined Apr 2013 | 4,366 Posts
Morning N :-) {quote} For that example, I was using fixed lots across the time series according to the following sizing (in yellow below). This is all traded on the Daily timeframe with an average hold per trade of 30.2 days. No pips are targeted. Every strategy is unbounded with trailing stops....no profit targets are used. I am hunting anomalies with these techniques. {image} No martingales used. Each strategy cuts losses short and lets profits run with full risk release each trade. The portfolio comprises 21 separate sub portfolios (markets)....
might possible you have a team (or a EA) to trade the way you are doing.
In the court of lord, only love and devotion is counted...
- #4,260
- Edited Jul 10, 2019 1:07am Jul 9, 2019 6:50pm | Edited Jul 10, 2019 1:07am
- | Commercial User | Joined Apr 2013 | 4,366 Posts
ATS 1000 Program
Guys....If you are road testing the ATS 1000 Program then to speed things up with backtesting use Control Points as opposed to Tick by Tick. With these particular D1 strategies that operate off the open on entries and close by a trailing stop, then there is little difference between the results as seen below
To illustrate the non-material difference here are the backtest results from Dukascopy for EURUSD over the test period 1/1/2000 to 21/6/2019. Note that I actually am testing these strategies on Pepperstone for live trading as the rebates with Pepperstone are nice added extras.
Dukascopy Tick by Tick
Dukascopy Control Points
Testing on tick by tick is very slow when using the Portfolio 1000 EA as it needs to sweep through 21 markets and 140 strategies at each iteration. There really is no need given the lack of material variation to Control Points with this particular type of strategy.
Cheers
C
Guys....If you are road testing the ATS 1000 Program then to speed things up with backtesting use Control Points as opposed to Tick by Tick. With these particular D1 strategies that operate off the open on entries and close by a trailing stop, then there is little difference between the results as seen below
To illustrate the non-material difference here are the backtest results from Dukascopy for EURUSD over the test period 1/1/2000 to 21/6/2019. Note that I actually am testing these strategies on Pepperstone for live trading as the rebates with Pepperstone are nice added extras.
Dukascopy Tick by Tick
Dukascopy Control Points
Testing on tick by tick is very slow when using the Portfolio 1000 EA as it needs to sweep through 21 markets and 140 strategies at each iteration. There really is no need given the lack of material variation to Control Points with this particular type of strategy.
Cheers
C