- Data for testing
- Point A
- Failed Point A
- Point C
- Pivot range
- W-pivot range -new
- Points A and C through the pivot
- Closing method -new
- Dynamics of Points A and C
- Significant time frames
- Entry and exit

Appendix

Testing, research and upgrade of Mark Fisher's ACD system, trading programs

- Data for testing
- Point A
- Failed Point A
- Point C
- Pivot range
- W-pivot range -new
- Points A and C through the pivot
- Closing method -new
- Dynamics of Points A and C
- Significant time frames
- Entry and exit

Appendix

In chapter 5 some of the drawbacks of Mark Fisher’s pivot were revealed, which prompted me to create my own pivot.

In order to somehow distinguish it from Mark Fisher’s pivot, I called it the w-pivot, because it is based on the weight of the candle, as described below. The w-pivot range is calculated in a more complex way. Before presenting the formulas, I will describe its calculation in words. Let's take the daily w-pivot range as a basis. Each candle on a 1-minute price chart can be represented as a set of prices from low to high in increments of 1 tick. We assign weight to each candle and the price inside it. Weight is equal to the trading volume for a candle divided by the number of ticks in it and multiplied by the serial number of the candle within the day. The trading day for stocks on the Moscow Exchange starts at 10 a.m., and ends at 18-45 p.m. and lasts 525 minutes. So for stocks, the serial number of candles during the trading day varies from 1 to 525. The greater the trading volume of the candle and the closer the candle is to the end of the trading day, the greater the weight of the candle and the price levels inside it.

Weighted prices in increments of 1 tick of all candles within the trading day are summed up and divided by the sum of the weights. Thus, the average weighted price *Ā* for the previous trading day is obtained. After that, the standard deviation *S* of the weighted prices is calculated. Now we can define the w-pivot range as *Ā±S/2*. Candles with a large trading volume and towards the end of the trading day make the largest contribution to the w-pivot due to weight.

The formulas for calculating *Ā, S* and the w-pivot range are as follows:

w-pivot range =*Ā± S/2*

where

*Ā* – average weighted price

– average weighted price squared

*S *– standard deviation of average weighted price

*P _{ij}* – price at j ticks above the low of candle number i (for the low j=1)

If the w-pivot is based on more than one day, then formulas are the same, only *N* is the number of candles for the previous several trading days.

The w-pivot is not the same for patterns (a) and (b) in Fig 5.2-5.3 from chapter 5 as Mark Fisher’s pivot. For pattern (a), the w-pivot is narrower and closer to the closing price of the trading day.

For a visual comparison of Mark Fisher’s pivot and the w-pivot, you can look at Fig 6.1.

Let's apply the mirror pivot approach described in chapter 5 for the w-pivot and compare the results with similar results for Mark Fisher's pivot. From a comparison of both pivots presented in Table A7, the following conclusions can be drawn:

- The w-pivot range is on average slightly larger.
- Opening outside the w-pivot range is more rare.
- The number of plus and minus days for the w-pivot is less.
- The probability of the w-pivot breakthrough is less, especially in the case of higher values of the number of days n in w-pivot: more green numbers in the last column of the table confirm this.

The latter conclusion seems to favor w-pivot. On the other hand, the advantage of w-pivot is not absolute, because w-pivot does not always show better results. Point A and C through w-pivot approach does not always bring more profit than Point A and C through the pivot approach described in Chapter 7.