Trend UP (MIMIC model) 1.0

The model was trained to trade in the direction of the trend. It performs well in an uptrending market but struggles in flat or reversal conditions. In a falling market, the model hardly trades at all — the reason for this behavior during training is currently unclear.

It is recommended to use the model only during strong upward trends.

Please note: the default settings are not optimal.
Make sure to run tests, use optimization, and adjust parameters based on specific market conditions.

The model was trained using a feature based on price slope.
In the settings, you will find the slope periods and the entry threshold.

Moving On with MIMIC Models!​

The DeMark model (expert advisor) is pretty straightforward: it performs great in flat, non-trending markets. Works well — and sometimes even really well!
(No image attached — but you already know what I’m talking about.)


But what happens when the market starts trending upward?
Exactly — DeMark and similar models eagerly jump against the trend, as you can see in image #1.


So I asked myself: can we build a model that actually trades with the trend?
And here it is — meet Trend UP! (Image #2)

Important to know:​

For some reason, this model learned to buy extremely well.
But when it comes to selling — not so much.
When the market starts falling, the model simply stops buying and becomes very reluctant to take any sell entries. (See image #3)

P.S.​

And yes, don’t forget: this model was trained exclusively for trend-following.
In sideways or reversal conditions, it doesn’t quite know what to do — and its performance drops accordingly.
But I think you’ve already figured that out: every model is designed for a specific market type.


Also, building a model like this was quite a challenge — much harder than DeMark, for example.
It is what it is — just sharing the experience.

---

A few important notes:​

The default settings are not optimal.
I haven’t optimized them at all.
After training the model, I just upload it as-is — no backtests, no polishing. (Too many ideas, too little time.)


So let’s agree on this:
I build the models — you test and optimize.


Try different parameters, symbols, timeframes, tweak the entry threshold, adjust the feature periods — all of this is available in the model’s settings block.


If you find a way to improve it — leave a comment on the site or drop a message in Telegram.
Just keep in mind: I don’t store source code forever. After a while, I may not be able to return to a specific EA and make changes.


Good luck with your tests!