Committed to sharing quantitative strategies, training videos, Python, algorithm research and other related content.
Hello everyone, I am Ukrainian Juggernaut.
In this issue, we are the second model model of the day trading model development series, and to be precise, it is an extension of the strategy of the first phase. SF27 has deletions and additions on the basis of 25. The removed module is ATR filtering, which has low code universality and increases the number of parameter optimizations. The added filtering module is a penalty module, and we will see how to change it.
Introduction:
Applicable period: 1 minute
Applicable varieties: Shanghai nickel, international copper, apple
Fee slip: Default firm fee, 1 slip in each direction
Note: The native variety of the strategy is Shanghai Nickel, and other varieties are recommended to readjust the parameters by themselves
principle
1. Intraday Weighted Average
Intraday trading often encounters reversal after opening, especially for non-ferrous precious metals that are greatly affected by the outer market. The place marked by the circle in the figure is to recalculate the intraday weighted average according to the night and day markets respectively. In this way, we judge the direction separately on the same trading day, so as not to adjust the direction in time when the market reverses.
2. Shining trading range (adaptive range)
Anyone who is familiar with the Shining trading system knows that it is a simulated trading loss before opening a position. This method is good and bad. For intraday, the number of transactions is more critical. You don't know whether you make money or lose money after opening the trade. Filtering too much will cause a suboptimal state and sacrifice trading opportunities.
We focus on the profit-to-loss ratio to reduce the magnitude of a single loss and increase trading opportunities. At the same time, a penalty module after a loss is added.
Here comes the point, the punishment module.
principle:
If your last trade lost, it may not be suitable for trading in this range, so we add or subtract a certain fluctuation value to the opening price of this trade. That is, the loss caused the range to enlarge.
Let's take a look at the following figure to analyze its advantages and disadvantages:
Pay attention to the closing signal in the blue box, it is a long and short loss exit. Then green is open long on rail, red is open short down rail. After a loss is incurred, the opening range will be added or subtracted by a certain range according to the opening price of the losing transaction. If it does not break through, no more positions will be opened. loss. On the contrary, there is a kind of market called piercing the head and breaking the foot, this kind of market will keep pulling back and forth, as shown below:
Like this kind of market, it breaks through a new high and then goes back, and then breaks again and then goes back. Then this kind of market is not profitable, and it will also cause a retracement, which is the disadvantage of this trading strategy.
A profitable market requires the opening and closing of high-volatility daily trends:
This is the intraday trend of the two trading days, and the shock of opening and closing. Falling today, rising tomorrow, then falling again, this kind of market will be very uncomfortable if it is an overnight strategy. But for intraday strategies, a big pull is very suitable for making money.
It is this kind of wide opening and closing, which is more comfortable.
There is also a unilateral trend at the daily level:
In this kind of market, there will be a small trial and error cost, but if it is a daily-level trend, this small loss is a drizzle.
think:
Q: If we add cross-cycle filtering, can it be better, such as the daily-level trend mentioned above, can we filter in this way?
A: Good idea, it is a good idea to combine intraday strategies with trend filtering of large cycles. Because in the small cycle, it does not know which stage of the historical market, whether it is the end of the trend or the middle of the trend, it does not know, so you can try this. In the next issue, we can add cross-cycle modules to try filtering.
Performance report (TB, Shanghai Nickel, 1 minute, firm cost, starting in 2015):
Apple:
International Copper:
MC (starting in 2020):
WH8：
PS: MC and Wenhua 8 trade after walking through the bar. This is not good for timely appearance within the day, so it causes performance differences. The native version is TB.
This strategy is only used for learning and communication, and investors are personally responsible for the profit and loss of real trading.