tech/a
No Ordinary Duck
- Joined
- 14 October 2004
- Posts
- 20,400
- Reactions
- 6,317
Long-term trading
- Thread starter dougy
- Start date
Wysiwyg
Everyone wants money
- Joined
- 8 August 2006
- Posts
- 8,428
- Reactions
- 284
If you are interested ..... Here are two stock that have crossed up 20 on the Stochastic %K(55,40) , I simplified the 39 SMA to 40. These two could be referenced in the future for profit or loss on the 80 cross down. Trouble is ... we could be waiting years.
ASB :- AUSTAL LIMITED FPO
Date :- 22/1/2010
Last Price :- $2.37
ATR :- ASTRON LIMITED FPO
Date :- 21/1/2010
Last Price :- $1.805
Well Dougy if you are still around this is the latest on the longer term buy/sell on stochastic crosses. So far so good.
ASB :- AUSTAL LIMITED FPO
Date :- 17/12/2010
Last Price :- $3.30
Up 93 cents (39.24%). Stochastic(55,40) %K at 77.10 and walking the 80 line but not yet a sell.
ATR :- ASTRON LIMITED FPO
Date :- 17/12/2010
Last Price :- $2.44
Up 63.5 cents (35.18%). Stochastic(55,40) %K at 76.86 and also close to the 80 line but not yet a sell.
Wysiwyg
Everyone wants money
- Joined
- 8 August 2006
- Posts
- 8,428
- Reactions
- 284
I didn't quite understand the Stochastic values.
It should read ... 39 to 40 period to calculate MA of %K.
So Stochastic %K(55, 40)
55 = period to determine highest high/lowest low
40 = period to determine MA of %K value
Well Dougy if you are still around this is the latest on the longer term buy/sell on stochastic crosses. So far so good.
ASB :- AUSTAL LIMITED FPO
Date :- 17/12/2010
Last Price :- $3.30
Up 93 cents (39.24%). Stochastic(55,40) %K at 77.10 and walking the 80 line but not yet a sell.
ATR :- ASTRON LIMITED FPO
Date :- 17/12/2010
Last Price :- $2.44
Knew I should have bought those ones!
tech/a
No Ordinary Duck
- Joined
- 14 October 2004
- Posts
- 20,400
- Reactions
- 6,317
W
Firstly I have no problems with Longterm trading and a longer term Stochastic is very similar to a n M/A crossover.While in this case you have a profit the exit may never be reached! You would need an EXTREME upthrust over a short period to cross 80---very unlikely.
For balance I include your 2 charts and one of mine with a 67% increase in 14 days.
The Portfolio is shown and its 3rd down 20000 at .4706.
The main thing I want to point out is the time spent in drawdown and the time spent in the trade in total of 12 mths.
Trading is about refining and making the most of our capital.
Just presenting this as food for thought.
Firstly I have no problems with Longterm trading and a longer term Stochastic is very similar to a n M/A crossover.While in this case you have a profit the exit may never be reached! You would need an EXTREME upthrust over a short period to cross 80---very unlikely.
For balance I include your 2 charts and one of mine with a 67% increase in 14 days.
The Portfolio is shown and its 3rd down 20000 at .4706.
The main thing I want to point out is the time spent in drawdown and the time spent in the trade in total of 12 mths.
Trading is about refining and making the most of our capital.
Just presenting this as food for thought.
- Joined
- 13 June 2007
- Posts
- 838
- Reactions
- 136
Greetings --
I ran the test of the crossover alone against the 500 stocks that are currently components of the S&P 500, from 1/1/2000 through 12/14/2010. (Because of survivorship bias, the results will over-estimate performance.)
I consider the Monday part and the portfolio part to be secondary. If the base algorithm is reasonably profitable, then they can be added.
In AmiBroker:
// ASF_StochLongTerm.afl
SetTradeDelays (0,0,0,0);
BuyPrice = Open;
SellPrice = Open;
weekDays = DayOfWeek();
BuyR = 1; //weekDays == 1;
SellR = 1; //weekDays == 1;
Buy = Cross(StochD(55,39,3), 20) AND BuyR;
Sell = Cross(80, StochD(55,39,3)) AND SellR;
----------------------------------------
For the 500 stocks and 11 years:
Median exposure is about 30%
Median Compound Annual Rate of Return (CAR) is about 1%
Median Return While Invested (RAR) is about 3.5%
Median Maximum Drawdown is about 55%
Median CAR/MDD is about 0.02.
Median k-ratio is 0.004.
Median Number of Trades in 11 years is 5.
Median Bars held per trade is about 170.
Whether trading only on Monday adds value or not, my first assessment is:
I am risking drawdowns in the range of 55% to make 1% a year and able to compound my account about every 26 months.
Terminal Wealth Relative, a ratio of final account balance to initial balance is:
TWR = (1+GM)^N
where GM is the geometric mean of the trade, N is the number of trades, and ^ indicates raised to the power.
With holding periods so long and CAR so low, assume GM can be replaced by CAR.
So TWR = (1.01)^5 = 1.051
Assuming the future resembles the past, which is implicit in all our analysis -- fundamental or technical, long term or short.
My final balance in 11 years will be about 5% higher than my initial balance today, but I will suffer a 55% drawdown somewhere along the way.
I want:
GM to be high.
N to be high.
CAR/MDD greater than 1.0.
k-ratio around 0.10.
The "sweet spot" for a trading system that will be traded to maximize terminal wealth while controlling risk is a holding period of a few days, 50 or more trades per year, and an accuracy rate of 60% or higher. When possible, add the ability and willingness to use leveraged trading instruments.
Longer holding periods, less frequent trading, and lower accuracy, all make creating a high final balance very difficult, even if the ratio of average amount won to amount lost is high.
I do not consider the system described to be a good system under any circumstances -- swing or long term -- and it certainly cannot be traded in any way that creates a high final balance.
Thanks for listening,
Howard
I ran the test of the crossover alone against the 500 stocks that are currently components of the S&P 500, from 1/1/2000 through 12/14/2010. (Because of survivorship bias, the results will over-estimate performance.)
I consider the Monday part and the portfolio part to be secondary. If the base algorithm is reasonably profitable, then they can be added.
In AmiBroker:
// ASF_StochLongTerm.afl
SetTradeDelays (0,0,0,0);
BuyPrice = Open;
SellPrice = Open;
weekDays = DayOfWeek();
BuyR = 1; //weekDays == 1;
SellR = 1; //weekDays == 1;
Buy = Cross(StochD(55,39,3), 20) AND BuyR;
Sell = Cross(80, StochD(55,39,3)) AND SellR;
----------------------------------------
For the 500 stocks and 11 years:
Median exposure is about 30%
Median Compound Annual Rate of Return (CAR) is about 1%
Median Return While Invested (RAR) is about 3.5%
Median Maximum Drawdown is about 55%
Median CAR/MDD is about 0.02.
Median k-ratio is 0.004.
Median Number of Trades in 11 years is 5.
Median Bars held per trade is about 170.
Whether trading only on Monday adds value or not, my first assessment is:
I am risking drawdowns in the range of 55% to make 1% a year and able to compound my account about every 26 months.
Terminal Wealth Relative, a ratio of final account balance to initial balance is:
TWR = (1+GM)^N
where GM is the geometric mean of the trade, N is the number of trades, and ^ indicates raised to the power.
With holding periods so long and CAR so low, assume GM can be replaced by CAR.
So TWR = (1.01)^5 = 1.051
Assuming the future resembles the past, which is implicit in all our analysis -- fundamental or technical, long term or short.
My final balance in 11 years will be about 5% higher than my initial balance today, but I will suffer a 55% drawdown somewhere along the way.
I want:
GM to be high.
N to be high.
CAR/MDD greater than 1.0.
k-ratio around 0.10.
The "sweet spot" for a trading system that will be traded to maximize terminal wealth while controlling risk is a holding period of a few days, 50 or more trades per year, and an accuracy rate of 60% or higher. When possible, add the ability and willingness to use leveraged trading instruments.
Longer holding periods, less frequent trading, and lower accuracy, all make creating a high final balance very difficult, even if the ratio of average amount won to amount lost is high.
I do not consider the system described to be a good system under any circumstances -- swing or long term -- and it certainly cannot be traded in any way that creates a high final balance.
Thanks for listening,
Howard
Wysiwyg
Everyone wants money
- Joined
- 8 August 2006
- Posts
- 8,428
- Reactions
- 284
Long term results so far.
One stock (ATR) crossed down over the stochk 80 line for a closure on the 17 March for a 59 cent gain or a tad under 25% profit.
ATR :- ASTRON LIMITED FPO
Date :- 17/3/2011
Sell Price :- $2.40
While ASB continues along for a gain of 35% so far
ASB :- AUSTAL LIMITED FPO
Date :- 11/4/2010
Last Price :- $3.21
One stock (ATR) crossed down over the stochk 80 line for a closure on the 17 March for a 59 cent gain or a tad under 25% profit.
ATR :- ASTRON LIMITED FPO
Date :- 17/3/2011
Sell Price :- $2.40
While ASB continues along for a gain of 35% so far
ASB :- AUSTAL LIMITED FPO
Date :- 11/4/2010
Last Price :- $3.21
tech/a
No Ordinary Duck
- Joined
- 14 October 2004
- Posts
- 20,400
- Reactions
- 6,317
Both still open on my charts.
But they weren't below the 20 line in January so I dont understand the trade.
This is ATR
I must be doing something wrong?
Wysiwyg
Everyone wants money
- Joined
- 8 August 2006
- Posts
- 8,428
- Reactions
- 284
I am using Yahoo OHLC.
Attachments
Wysiwyg
Everyone wants money
- Joined
- 8 August 2006
- Posts
- 8,428
- Reactions
- 284
As a comparison I have used a Commsec chart above to show the data is extremely similar. Maybe our stochastic indicators are different?
Attachments
tech/a
No Ordinary Duck
- Joined
- 14 October 2004
- Posts
- 20,400
- Reactions
- 6,317
Must be.
