Considering a normal "buy/sell" situation, an asset would be bought in average at the median price following a Gaussian like concept. A higher or lower average trend would significate that the current perceived value is respectively higher or lower than the current median price, which mean that the buyers are evaluating the price underpriced or overpriced.
This behaviour would be even more relevent depending on its derivative evolution.
Therefore, this Strategy setup is based on this Gaussian like concept anomaly of average close positionning compare to high-low average derivative, such as the derivative of the following ploted basic signal : 1-(high+low)/(2*close).
This Strategy can actually be used like a trend change and continuation strength indicator aswell.
In the Setup Signal part :
You can define the filtering of the basis signal "1-(high+low)/(2*close)" on EMA or SMA as you wish.
You can define the corresponding period and the threathold as a mutiply of the average 1/3 of all time value of the basis signal.
You can define the SMA filtering period of the Derivative signal and the corresponding threathold on the same mutiply of the average 1/3 of all time value of the derivative.
In the Setup Strategy part :
You can set up your strategy assesment based on Long and/or Short. You can also define the considered period.
The most successful tuned strategies I did were based on the derivative indicator with periods on the basis signal and the derivative under 30, can be 1 to 3 of te derivative and 7 to 21 for the basis signal. The threathold depends on the asset volatility aswell, 1 is usually the most efficient but 0 to 10 can be relevent depending on the situation I met. You can find an example of tuning for this strategy based on Kering's case hereafter.
I hoping that you will enjoy using this Strategy, don't hesitate to comment, to question, to correct or complete it ! I would be very curious about similar famous approaches that would have already been made.
For those who like cryptos, you can try the following parameters on BTC to experiment and backtest this strategy: INDICATOR SIGNAL SETUP MA Period = 18 Threath.. Ratio = 2 Derivative MA Period = 3 Threath.. Derivative = 1
TRADING STRATEGY SETUP Depending on your risk/performance risk choose Long and/or Short The Bar Index Start is the number of bars you want to consider in the past from today… if you want to test how effectively this strategy could perform the bear market, you can try 365*2=730 for instance
STRATEGY PROPERTIES Pyramidage = 2 Then set up your assets, order setup and fees… For example, 10000$ Asset, 10% equity, 1$ per order
How does it look ? How would you impove it ?
arunagirish03
⋅
looks attractive. seems excellent work and analysis. . does it work for crude and gold and equities in low time frames. however its one of the masterpiece
Koamind
⋅
@arunagirish03, Thanks for your feedback ! The best is to use it on a 1D basis mainly as it usually gives the best ratio betneew transaction frequency and benefits per trade allowing the transaction fees. The challenge is to find the most appropriate SMA periods (1. Basis signal, 2. Derivative) and then the most effective threathold on the derivative (0.5, 1, 1.5, 2, 2.5, ...) to amplify the gaussian anomaly phenomenon and limit the fake signals. I found this strategy very effective on ranging assets aswell. Using piramid order can also amplify the performance. So far, I mainly tried it on stocks and cryptos but it should work on Gold aswell. I think this strategy would not fit for gold physical buy, as the fees are very high and this strategy try to optimize the performance on a swing trading pace, but on CFDs it should give something, especially with leverage, otherwise I would be possible to find something but with a higer risk with higher equity ratio. Do you have an idea on the contract, fees, asset amount and average trade I could use for the strategy tuning on a Gold case ?
Koamind
⋅
@arunagirish03, will see with your inputs but for GOLD 1D with lever 1 (upper would multiply the performance), you can try the following setup : INDICATOR SIGNAL SETUP MA Period = 14 Threath.. Ratio = 2 Derivative MA Period = 8 Threath.. Derivative = 0
TRADING STRATEGY SETUP All ticked And Bar Index Start at 365 for one year backtrack or 730 for 2 years
STRATEGY PROPERTIES Asset = 10k$ Order = 10% equity Pyramid= 10 Fees = 1$ (or even better with french SRD like leverage contract fees @0.025%)
arunagirish03
⋅
@Koamind, . sure will try the same . amazing work. ..
INDICATOR SIGNAL SETUP
MA Period = 18
Threath.. Ratio = 2
Derivative MA Period = 3
Threath.. Derivative = 1
TRADING STRATEGY SETUP
Depending on your risk/performance risk choose Long and/or Short
The Bar Index Start is the number of bars you want to consider in the past from today… if you want to test how effectively this strategy could perform the bear market, you can try 365*2=730 for instance
STRATEGY PROPERTIES
Pyramidage = 2
Then set up your assets, order setup and fees… For example, 10000$ Asset, 10% equity, 1$ per order
How does it look ? How would you impove it ?