Expert Advisor optional features

I designed the EA Studio backtester to generate and test strategies with a fixed-lot entry at bar-open. I spent months and years trying to make the world's fastest and most precise backtesting engine.

Now, I'm almost satisfied with that. However, it is obvious that the lack of money-management features and dynamic parameters is a drawback. For that reason, I started the work on a new product with a lot of new functionality. It sounds great, but the development is slow and will take another two years to finish.

Having a burst of inspiration, I decided that I could add additional features to the EA Studio's Expert Advisors, even if they can be tested in MetaTrader only.

I intend to implement an "Averaging down" functionality for the exported expert advisors.

This feature will be optional and will be included in the robots only if it is enabled. The Averaging options will be controlled from the Expert's Input. All options will be off by default, so they will not affect the original behavior of the strategy when disabled.

When enabled, the Averaging functionality will alternate the Expert Advisor's work in MetaTrader. It will be able to test and adjust the parameters in the MetaTrader's Strategy Tester.

Variants of Grid (Averaging Down)

- Trade levels:
   * add to a losing position at a predefined step - Grid
   * add to a losing position at a previous top/bottom (ZigZag, Fractal, Round price, ...)

- Trade volume:
  * add a fixed lot size
  * scale the entry volume up with a fixed or dynamic coefficient.

- Exit on profit
  * using a fixed Take Profit, which is dynamically calculated as per the current average position price

- Exit on loss
  * Exit on a predefined loss level as a percentage of the account at the first entry.

- Averaging conditions
  * always average a losing position
  * apply averaging only if particular indicators show a ranging market: (ADX less than 25, RSI between 30 and 70, narrow Bollinger Bands, ...)

Grid Input Options

Hello, Traders,
I started designing the options for the Grid system. Here is the proposal:

Enable Grid (Averaging Down) - true or false - it enables or disables the grid system. It is off by default.

Grid Distance (initial, pips) - 50 - the distance from the initial entry to the first grid entry in pips.

Grid distance sizing - this option determines how the spacing between grid entries changes:
- Equal distance (pips) - all grid entries are equally spaced from the initial entry
    Example: Initial at 1.2000, grid entries at 1.1950, 1.1900, 1.1850 (50 pips apart each).

- Increment distance (pips) - each grid entry adds a fixed increment to the previous spacing
    Example: With initial 50 pips and increment 20, entries are at: 50 pips from initial (1.1950), then 70 pips below (1.1880), then 90 pips below (1.1790).

- Scale distance (coefficient) - each spacing is multiplied by the coefficient
    Example: With initial 50 pips and scale 1.25, entries are at: 50 pips from initial (1.1950), then 62 pips below, then 78 pips below.

Next connected distance sizing options are:

Grid increment distance (pips) - 20  - the increment value added to the spacing for each subsequent grid entry. Only used when "Grid distance sizing" is "Increment distance (pips)".

Grid scale distance (coefficient) - 1.25 - the multiplier applied to the spacing for each subsequent grid entry. Only used when "Grid distance sizing" is "Scale distance (coefficient)".

Then we have very similar options for the Grid volume sizing:

Grid volume sizing - this option determines how the lot size changes for grid entries:
- Equal volume (equal to the initial entry) - all grid entries use the same lot size as the initial entry
    Example: If initial is 0.05 lots, all grid entries are 0.05 lots

- Increment volume (lots) - each grid entry adds a fixed increment to the lot size
    Example: If initial is 0.05 lots and increment is 0.01, entries are: 0.06, 0.07, 0.08 lots

- Scale volume (coefficient) - each lot size is multiplied by the coefficient (like Martingale)
    Example: If initial is 0.05 lots and scale is 1.25, entries are: 0.06, 0.08, 0.10 lots

These options work with the following two:

Grid increment volume (lots) - 0.01 - the increment value added to the lot size for each subsequent grid entry. Only used when "Grid volume sizing" is "Increment volume (lots)".

Grid scale volume (coefficient) - 1.25 - the multiplier applied to the lot size for each subsequent grid entry. Only used when "Grid volume sizing" is "Scale volume (coefficient)".

The Grid will exit at Take Profit

Grid Take Profit (pips) - 20 - The Take Profit of all open positions is calculated at every addition. It is measured from the aggregate positions price.

The following two options are the grid limitations:

Grid maximum total positions - 10 - the count of the initial entry and the subsequent entries

Maximum grid lots - 0 - the maximum number of open lots including all entries. Zero means off.

Then we have two options to limit the total loss:

Grid total Stop Loss - it is measured from the initial entry. Zero means off.

Grid total Loss Limit (%) - the maximum loss as a percentage of the account balance at the time the initial entry was opened. For example, if balance was $10,000 when the initial entry opened and this is set to 5%, all positions close when the total floating loss reaches $500. Zero means this limit is disabled.

> I started designing the options for the Grid system. Here is the proposal:

Awesome, I've been waiting for that for a long time!

I do not see any Take Profit strategy?
Optimization directly on MT5 will be possible?

First Test of the Grid functionality - Roller-Coaster style

With Grid

Who needs protections:

Positions with distance and volume scaling:

Nightmare Balance-equity curve:

The same strategy with Grid off.

We have a saying in Bulgarian (however it sounds more colorful)

Hi Popov,

Nice to hear that this option (grid) is now also on your desk.

I’ve been selecting and trading third-party grid systems for several years (over 30 different commercial
products), and in my experience only a few manage to deliver consistent results over time. We’ve also spent significant time refining the settings and working on cross-adaptation across different configurations.

If you’re interested, we’d be happy to share our findings and setups.

Best regards,
Vincenzo

Fibonacci numbers are everywhere.

I found them in the Grid options also.

> Soon or later, you blow-off the account or, with the hard stops...
Yes, exactly, but it is controlled. This was my main motivation to start doing Grid expert.
What I noticed, Petko trades separate account for each Expert Advisor. Some of his Grid accounts blow, but the rest covers the loss.

> Is this seen as a limit? Why only small accounts?
It is because the risk is limited. Prop firms and the small accounts do not love risk.
It does not mean the EA Studio Expert Advisors cannot trade on larger accounts, of course.

...

I do not recommend Grid trading for everyone. I'm doing this Grid expert for research. I'm exploring the opportunity and provide tools for the user which want to try it.

Hi Popov,
Based on long experience with grid systems, the objective is to reduce exposure by systematically pairing the largest order with earlier orders, using full closures first, and partial closures only when necessary, while keeping the grid mathematically consistent.
Base setup:
•    Fixed grid step (e.g. 50 pips), of course it is better to have a variable space like the ones you included in the EA “fib ratios”.
•    Initial lot size: 0.01
•    Lot multiplier: 2
•    Orders are opened against price at each grid level
________________________________________
Example State
After a 150-pip adverse move, the EA has:
1.    Order #1 → 0.01 lot
2.    Order #2 → 0.02 lot
3.    Order #3 → 0.04 lot
All orders are in drawdown.
________________________________________
Exit & Reduction Logic (Sequential)
Step 1 — Full pair closure
When price retraces and Order #3 (0.04) and Order #1 (0.01) are both in profit, this can be fixed or variable dollar value:
•    Close Order #3 (0.04) and Order #1 (0.01) completely.
•    Remaining open order:
o    Order #2 → 0.02 lot
This step is always executed first whenever possible.
________________________________________
Step 2 — Partial pairing (forced continuation)
If price retraces again and no smaller full-lot order exists for pairing:
•    When a new order#3 again and Order #2 (0.02) are both in profit:
o    Close the largest new order (0.04) completely with
o    Partially close Order #2 by 0.01, reducing it from 0.02 → 0.01.
After this step:
•    Remaining exposure:
o    Order #2 → 0.01 lot
________________________________________
Re-entry Logic (Important)
After any partial close, the grid must restart from the remaining lot size, not from the historical maximum.
•    If Order #2 is reduced to 0.01:
o    This becomes the new base lot.
•    If price moves again against the position:
o    The next grid order size must be:
    0.01 × lot multiplier = 0.02
    NOT 0.04
From this point forward:
•    Lot progression continues normally:
o    0.01 → 0.02 → 0.04 → 0.08 …

I implemented the pair-closing strategy. It works well.

However, I reconsidered my approach to the Grid system expert.

My initial plan was to execute the original strategy closing logic (including SL) only if the positions is at a profit. However, it completely changes the strategy behavior.
Now, I'm rewriting the Grid logic to be blended in the original strategy. Think of it like "possibility" (Losers Average Losers) to make a recovery entry at a more favorable price.

The new options are:

- Grid pair position closing: true or false - When on, it closes the last open position with the full or partial first position as described by Quantbot (https://forexsb.com/forum/post/83088/#p83088)

- Grid Take Profit (currency) - Take Profit in dollars (or whatever is the account currency)

- Grid Stop Loss (currency) - Stop Loss in dollars (or whatever is the account currency). It is used by the pair closing described above.

- Major change - the Grid option acts together with the original strategy's exit rules.

...

ToDo:
- check if the Grid Expert works as in EA Studio when the Grid is off
- the same as above when the Grid is on, but the Grid max open positions = 1
- add some warnings if the options contradict (Pair closing on and Take Profit currency = 0)
- check if the position and account protections work with the Grid.

If everything is OK, I'll release the Grid functionality for public testing for all EA Studio users.

Hi Popov, this is what I would test:

    1.    The strategy combines a statistical edge with a structured recovery grid.
    2.    The EA trades normally and closes at TP during standard conditions.
    3.    If price moves against the trade by the original SL distance, recovery mode is activated instead of closing the trade.
    4.    The original TP is removed and the position is managed as a basket.
    5.    A grid sequence opens positions against price using a lot multiplier (e.g., 2).
    6.    The objective is not classic martingale recovery, but active exposure reduction.
    7.    The largest lot is always paired with the smallest lot and both are closed when in profit.
    8.    If needed, partial closes are used to reduce exposure back to the minimum base lot.
    9.    After any reduction, the lot progression resets from the new base size, avoiding historical escalation.
    10.    When exposure is minimized and the basket reaches breakeven or small profit, all trades close and the system returns to normal edge trading.

This approach aims to reduce the percentage of losing trades by actively managing and redistributing risk instead of closing positions at the original stop loss.

However, in a strong and persistent trend with little or no retracement, and without a hard stop mechanism, the risk of significant drawdown or even account blow-up is real.

On the other hand, it is important to remember that every large stop-out has a nonlinear recovery cost. For example, a 20% loss requires a 25% gain to return to breakeven (not 20%). The deeper the drawdown, the disproportionately larger the recovery required.

If you get 2 stop-outs in a row of 20% each, the account would decline by 36% overall, and you would then need approximately a 56.25% gain to return to breakeven. At 5% per months (without any other stop out) you would need approximately 9/10 months.

Just to keep all in mind!