The first goal was to find a fully automated or mostly automated way to trade 60 second options. In this case I haven't been able to fully automate it yet.
The second goal was to treat it similarly to high frequency trading. Harvesting micro amounts of money from the market frequently. Although each trade is not worth much collectively it adds up over time.
Assuming a few things
1. Broker has customizable trade size. (Some brokers only have preset bets which this wouldn't work for).
2. Account $4,000
3. Minimum Bet $1 - I know of only one broker offering $1 options most minimums are $25 although there are a few offering $5.
4. 70% Return -
Obviously these numbers could be scaled up for larger minimums.
Okay lets get to the statistics then I'll explain the system.
------------------------------------------------------------------------------------------
Since this was a manual backtest and not automated by mt4 the sample size is fairly low at this point.
Two sets of 500ish trades below is the worst of the two.
557 Total Trades
Winning Trades = 282
Losing Trades = 275
Win to Loss Ratio 1.02:1
========================================================================
Not so good right? All most a 50/50 ratio. To make up for the standard 70% return per trade we are going to multiply each losing bet by 2.5.
Here is the chart
http://i863.photobucket.com/albums/a...psf233e19b.png
Because 2.5x is more than is actually needed to stay profitable each multiple earns slightly more net profit than the previous. As shown below
http://i863.photobucket.com/albums/a...ps01d90f24.png
The return is quite small considering trade size. Which is the case with high frequency trading or even scalping. Professional scalpers open large positions worth 100's of thousand dollars to earn a fraction of a percent. This is the same except the risk is higher.
In this case if you desire more return per trade a larger multiplier is needed instead of 2.5x
==========================================================================================================
The system is quite simple. Open a 1M chart and add the indicator Pattern_Recognition_Master_V3.
PRMv3 alerts for several candle stick patterns
1. Doji
2. Morning/Evening Star
3. Engulfing Candlestick
4. Dark Cloud
5. Piercing Line
For this run only Dark Cloud, Engulfing Candlestick, and Piercing Line were enabled.
--------------------------------------------------------------------------------------------------------------------------------------------------------------------
"Total Trades To Win" or TW for short
If 5 trades lost and the 6th wins the total trades to win is 6.
Easy enough right?
Here's the break down of each 282 winning trades in terms of total trades to win.
TW-Worth-Num----Total
1= $0.70 x 141 = $98.7 -
2= $0.75 x 73 = $54.75
3= $0.875 x 35 = $30.625
4= $1.191 x 14 = $16.674
5= $1.962 x 10 = $19.62
6= $3.29 x 4 = $13.16
7= $8.79 x 5 = $43.95
8= $21 x 1 = $21
Total Earned = $298.479
Chances of 8 TW happening are 1/557 or 0.00179
After two runs of 500+ trades 1 had a 7 loss 8 TW the other did not. The worst of the two is shown above as a worst case scenario. No 9 or 10 TW's were found with these or any other test runs not yet documented. Although it probably could happen. With that being said I'm hoping to lower the number of trades and increase accuracy by implementing some sort of trend following factor.
The test above did not apply any strategy other than the candlestick patterns. Many up trades were taken in down trends some won some did not. The same is true for up trends and down trades. I guess you could call it "blind trading" every signal no matter what the trend, volume, range etc.
It probably wouldn't be to difficult to increase accuracy by adding some sort of trend following factor and trade filter.
Merry Christmas and Happy Holidays,
John
Edit: Completely forgot to mention the currency pair XAU/USD (Gold/USD)
The second goal was to treat it similarly to high frequency trading. Harvesting micro amounts of money from the market frequently. Although each trade is not worth much collectively it adds up over time.
Assuming a few things
1. Broker has customizable trade size. (Some brokers only have preset bets which this wouldn't work for).
2. Account $4,000
3. Minimum Bet $1 - I know of only one broker offering $1 options most minimums are $25 although there are a few offering $5.
4. 70% Return -
Obviously these numbers could be scaled up for larger minimums.
Okay lets get to the statistics then I'll explain the system.
------------------------------------------------------------------------------------------
Since this was a manual backtest and not automated by mt4 the sample size is fairly low at this point.
Two sets of 500ish trades below is the worst of the two.
557 Total Trades
Winning Trades = 282
Losing Trades = 275
Win to Loss Ratio 1.02:1
========================================================================
Not so good right? All most a 50/50 ratio. To make up for the standard 70% return per trade we are going to multiply each losing bet by 2.5.
Here is the chart
http://i863.photobucket.com/albums/a...psf233e19b.png
Because 2.5x is more than is actually needed to stay profitable each multiple earns slightly more net profit than the previous. As shown below
http://i863.photobucket.com/albums/a...ps01d90f24.png
The return is quite small considering trade size. Which is the case with high frequency trading or even scalping. Professional scalpers open large positions worth 100's of thousand dollars to earn a fraction of a percent. This is the same except the risk is higher.
In this case if you desire more return per trade a larger multiplier is needed instead of 2.5x
==========================================================================================================
The system is quite simple. Open a 1M chart and add the indicator Pattern_Recognition_Master_V3.
PRMv3 alerts for several candle stick patterns
1. Doji
2. Morning/Evening Star
3. Engulfing Candlestick
4. Dark Cloud
5. Piercing Line
For this run only Dark Cloud, Engulfing Candlestick, and Piercing Line were enabled.
--------------------------------------------------------------------------------------------------------------------------------------------------------------------
"Total Trades To Win" or TW for short
If 5 trades lost and the 6th wins the total trades to win is 6.
Easy enough right?
Here's the break down of each 282 winning trades in terms of total trades to win.
TW-Worth-Num----Total
1= $0.70 x 141 = $98.7 -
2= $0.75 x 73 = $54.75
3= $0.875 x 35 = $30.625
4= $1.191 x 14 = $16.674
5= $1.962 x 10 = $19.62
6= $3.29 x 4 = $13.16
7= $8.79 x 5 = $43.95
8= $21 x 1 = $21
Total Earned = $298.479
Chances of 8 TW happening are 1/557 or 0.00179
After two runs of 500+ trades 1 had a 7 loss 8 TW the other did not. The worst of the two is shown above as a worst case scenario. No 9 or 10 TW's were found with these or any other test runs not yet documented. Although it probably could happen. With that being said I'm hoping to lower the number of trades and increase accuracy by implementing some sort of trend following factor.
The test above did not apply any strategy other than the candlestick patterns. Many up trades were taken in down trends some won some did not. The same is true for up trends and down trades. I guess you could call it "blind trading" every signal no matter what the trend, volume, range etc.
It probably wouldn't be to difficult to increase accuracy by adding some sort of trend following factor and trade filter.
Merry Christmas and Happy Holidays,
John
Edit: Completely forgot to mention the currency pair XAU/USD (Gold/USD)
Attached File(s)
pattern_recognition_master_v3.mq4
27 KB
|
290 downloads