TradePatio uses cookies to provide necessary site functionality and improve your experience. By using our website, you agree to our privacy policy.

2019-11-29 07:07:12

What is asymmetric trailing?

First, let's look at the classic trailing stop strategy approach.

The word “trail” has a translation from English “follow”, in the case of exchange trading, the option “floating stop” is more suitable - the process of pulling stop loss to a growing trend (limited responding to course subsidence). This is a classic approach at this point in time.

Only 2 variables participate in it, the first is current price and the second is the step (step) in percent (or in units of price) which indicates when to lower the lower threshold ( stop loss ) is closer to the price. When crossing stop loss we place an order and fix the current price.

Let's look at an example:

Suppose we have coins that we bought at 98, and which we want to sell. We assume that the initial exchange rate of the coin is 100 - that is, if we sell right now, we will already make a profit, but we want to get more. At the same time, if the rate suddenly starts to fall and falls by 98 or lower, we should immediately sell what we bought in order to minimize losses.

Thus, we will set the lower limit (stop-loss) to 98, and then we will increase it every time the difference between the current exchange rate and the lower border is the specified value - in our case, for clarity of calculations let it be 5%. In practice, values of 1-2% are most often used, but this largely depends on the volatility of the pair and other factors.

Suppose our rate rises to 102 - the difference with 98 of the initial lower limit is less than 5%, so we do nothing. As soon as the price exceeds 98 + 5% = 102.9 - we tighten the stop-loss. On our test data set, the rate rose to 107 (see the figure below) and the lower border moves to the mark 107 - 5% = 101.65. Now the rate is 107, but if it drops to 101.65 - we will sell, and we will already be in the black.

The next price step (step 5) is the rate 106. The price has fallen, but we do not change the lower border on the fall, we change only on the increase, and stop loss remains 101.65. The price jumped to 110 - stop loss pulled up to 110 - 5% = 104.5   etc.

In the figure below you can see a graph where the price movement is shown in blue and the selected stop loss in red.

Initially, we expected to sell at 105, and sold at 108. This is profitable - of course, it would be more profitable to sell at 115, this is the maximum price on the chart until the time of sale, but ours the algorithm sells only at a rollback of the price, so when a stop-loss was reached (in our case, 109.25), the position was closed.

What is asymmetric trailing?

With this strategy, we also follow the price, but 4 variables are involved here:

    1. Current price : current price.
    2. Current rollback: this is a floating price upon reaching which we fix the position and place an order. It is configured by the parameter Price rollback% . We cannot pick up a price lower than the original (in this case 100) at Current rollback , there is a parameter Stop loss .
    3. Next price : this is the floating price at which we move Current rollback to Current price and push back Next price from Current price . It is configured by the parameter Price step% .
    4. Stop loss : this is the lower threshold to which the price can go down and we will place orders so as not to go at a greater loss. It is set at a specific price, never change and is optional.


Let's look at an example:

Let's say we bought some coins at the rate of 100, and we want to sell them at a profit. Now the course is 100, but we expect to get more. To minimize losses, set Stop loss 93 - if the price drops to this threshold, we will sell coins. Price step to 5%, that is, the first Next price will be at 105. Price rollback to 3% until the price reaches the first Next price (in this case 105) - Current rollback not active.

At a price of 103, and this is an increase of 3% from the original rate, we do not sell or increase anything (we are waiting for 5%, that is, the price is 105). As soon as the price reaches 105, we tighten the lower border. Because the course has now jumped to 105 (see the figure below), then the lower border will already be 101.85 (3% of Current price : 105). The current rate is 105, but if it drops to 101.85, then we will sell (and we will already be in the black). The difference between 105 and 101.85 is 3%. If the course grows higher, we will increase again, but we are already waiting for the mark in 110.25 (5% of 105 ) .

The next step of the chart is the rate 104. The price has fallen, but we do not move the lower border on the fall, we change only on growth, Current rollback 101.85 remains. Price jumped to 111 - Current rollback pulled up to 106.94 (3% of Current price ), expect Next price at 115.76 (5% of Current price ) etc.

Below is a graph of the movement along the steps, how the algorithm works.

Also, if you draw a graph, you get a corridor with a given offset.

Here we originally planned to sell at 105 and sold at 117. In this case, the upper threshold was at 122 and in the end we deviated from the maximum profit in 4.09%.

When Asymmetric Trailing if set Price step minimum and Price rollback more Price step then, we will be more susceptible to statistical noise and will not be able to reach the most favorable price.

To purchase, the algorithm becomes mirrored:

By stop loss :

Fine-tuning this tool for each pair individually will lead to the most profitable deals, since you can take into account the features for each of them and choose the best strategy.