[Basis] Thorough verification of why and how accurately ABUADX (TF2) and ABUDMI (TF1) filters can predict trend endings

Good evening!

I’m Shirou.

The other day, Abu-san’s post

ABUADX (TF2) and ABUDMI (TF1) show the end of the trend, which can be useful in line trading

and shared this valuable information with us.

Is it true that you can tell when a trend ends?

How can you tell the end of a trend with ADX and DMI?

For those who are curious, in this article we will

• Explain why the trend end can be identified
• Explain the accuracy of ABUADX and ABUDMI filters

without any bias, thoroughly tested, so please read on!

These are my personal views; please understand in advance.

Reasons why ABUADX (TF2) and ABUDMI (TF1) filters can predict trend end

Here is a quick recap of the two indicators related to why the trend end can be predicted with ABUADX and ABUDMI.

The two are

• ABUADX = trend strength
• ABUDMI = trend direction

indicators focused on these aspects.

As a filter, in language, it is when

the higher timeframe shows trend strength and the execution timeframe shows the trend direction alignment

that this applies.

To make it easier to understand, let’s break it down one by one.

First, ABUDMI.

Let’s look at the color change in the image.

AUDCHF 1-hour chart

As you can see,ABUDMI changes color right after the peak ends.

In other words,

the current ABUDMI color change itself indicates the end of the trend

.

Next, ABUADX (TF2); this is the next higher timeframe, isn’t it?

The basics of FX aregoing with the trend on the higher timeframe.

If the trend strength on the higher timeframe is opposite, there is a risk of retracement on the execution timeframe.

This isa filter for aligning with the higher-timeframe trend.

Even if the execution timeframe has a direction aligned, if the higher timeframe differs, there’s a possibility of moving in the opposite direction.

Summary

Predict the trend end with ABUDMI and use ABUADX (TF2) to sense whether the trend strength aligns

as a filter.

Just these two together already seem to offer high edge

Next, the questions:

How accurate are ADX (TF2) and DMI (TF1) as filters?

We tested this.

Thorough verification of ABUADX (TF2) and ABUDMI (TF1) filter-based trend end predictions

A red line is drawn at where the end of the trend occurs (roughly gauged).

All on 1-hour charts

The chart is slightly condensed to fit one frame

Only six currencies

Results

As the filter accuracy,

• 10 times valid
• 4 times invalid

This was the result.

However,including the policy rate events and the first color change of DMI, if you properly do fundamentals and observe the first color changeinvalids become only once.

Furthermore, if you also include a DMI color-change stop-out for the invalid one, you only incur a small loss.

AUDJPY

All three indicators prevented losses in both indicators, right?

CADCHF

Predictable, isn’t it.

EURAUD

Only the middle red line is invalid.

With a DMI color change, it seems you could stop out safely

EURGBP

Both are effective as filters.

If fully automated, I think the short entry would hit the stop on the left side chart.

GBPNZD

The right side is invalid, but this is the time when NZ policy rate was announced.

USDJPY

The middle is valid, but the sides barely fail as filters.

In conclusion

From looking at the charts, it’s clear that ADX TF2 and DMITF1 can predict trend end.

I really think Abu-san’s perspective is amazing.

This may deviate slightly, but since trend end can be sensed to this extent,

it might be good to consider a strategy like a take-profit on a point basis with oscillator TF2 color changes for stop-out

as in the traditional heaven-and-earth EA approach.

During adjustments, there are times when it crosses the line and returns to Band 2.0,