In order to identify long-term and non-random components, it is necessary to know not only how much the price changed, but also how
these changes were happening. In other words – we are interested not only in the values of the price levels, but also in the order in which the price levels are calculated.
in which these levels succeeded each other. Thanks to this approach, it is possible to find long-term and stable factors that affect
(or may affect) the price change at a given time. And knowing these factors allows you to make a more or less accurate forecast.
these changes were happening. In other words – we are interested not only in the values of the price levels, but also in the order in which the price levels are calculated.
in which these levels succeeded each other. Thanks to this approach, it is possible to find long-term and stable factors that affect
(or may affect) the price change at a given time. And knowing these factors allows you to make a more or less accurate forecast.
Let’s define what specific results we want to get from the analysis of financial series over time:
In order to meet these conditions, we will use a regression analysis of relative price changes, and make based on
this model is an indicator. The algorithm of this indicator is based on training on historical data, and its operation is under full control.
under the control of the trader.
this model is an indicator. The algorithm of this indicator is based on training on historical data, and its operation is under full control.
under the control of the trader.
- LH is a parameter that determines the number of bars used to smooth the financial series. His
the allowed value is between 0-255. - UTS-a parameter that affects the speed and depth of training. Its value also lies in the range of 0-255. If
If the UTS value is zero, then training occurs throughout the entire history. In all other cases, a training sample from time to time
updated. This update occurs the more frequently the lower the UTS value.
[spoiler title=”Read More…”]
- LH is a parameter that determines the number of bars used to smooth the financial series. His
the allowed value is between 0-255. - UTS-a parameter that affects the speed and depth of training. Its value also lies in the range of 0-255. If
If the UTS value is zero, then training occurs throughout the entire history. In all other cases, a training sample from time to time
updated. This update occurs the more frequently the lower the UTS value.
The main drawback of this indicator is that when smoothing financial series are considered as stationary, which is why
there may be some delay.
there may be some delay.
[/spoiler]