Collie Gold 1.12: 2 stops in 2026 backtests became 0. Here's the catch
Collie Gold 1.12 is out. It is a free update for everyone who bought or rented Collie, and it brings small adjustments to Collie's core, made for the way gold has been moving lately. Most of the time, it trades exactly as in 1.11.
Why we made this change
In our 2026 backtests, two baskets reached their stop. We studied both and adjusted the core around what we found. We then checked the change on every year since 2021, not only on the moments that inspired it.
What the backtests show
Strategy Tester backtests with every tick based on real ticks, Moderate profile, starting with 1,000 USD. The drawdown is the maximum equity drawdown, measured on every tick.
| 2026 through 14 October | Collie 1.11 | Collie 1.12 |
|---|---|---|
| Baskets that reached their stop | 2 | 0 |
| Maximum drawdown | 30% | 30% |
| Profit | base | about +37% |
- Over the whole run since 2021 (one continuous backtest, the lot growing with the account), profit is about 35% higher than with 1.11.
- The two baskets that reached their stop in 1.11 did not reach it in 1.12. The full results by year and profile are on the product page.
- From 2021 to 2025, almost every trade is the same as in 1.11. In the one year where a few trades differ, profit was about 14% higher (114 dollars for every 100), with the same drawdown.
Was this fitted to 2026?
Partly, yes: we built it on those two baskets, so 2026 is where it was always going to look best. That is why the years it was not built on matter more: from 2021 to 2025 it changed almost nothing, and where it did, the result was slightly better with the same drawdown. It adds no new setting and works the same way on every profile. The real test is the next time those conditions show up on a live account, and we will report it here, whether that basket ends at its target or at its stop.
The catch
The adjustments have a cost: at times Collie makes less profit than 1.11 would have. In the backtests, that cost was smaller than what it saved, but you should know it is there.
What we don't know yet
We built this change after studying those baskets, and are a small sample. They show how the change works, not what will happen next time. A basket can still reach its stop: stops are part of the design, and the risk profile you choose still sets what one costs.
What stays the same
- A defined stop on every basket, the same lot on every entry (never martingale), and the lot growing with your balance.
- The cost of a basket at its stop is unchanged: about 15% of the balance on Moderate, and the prop firm lot table stays the same.
- The five risk profiles, the news pause, Short Friday and every other setting. Your current settings carry over.
- The results on the product page are now from 1.12.
How to update
- If you can, update when Collie has no basket open (the panel shows open trades). If one is open, Collie picks it up again when you re-attach, but the stop is executed by the EA, so keep the gap short.
- In MetaTrader 5, open the Market tab, go to Purchased, and click Update next to Collie Gold.
- Remove Collie from the chart and attach it again, so the chart loads the new version. Check that the panel and the Experts tab show version 1.12.
- Your settings stay as they were. If you use a preset file, load it again after attaching.
If you trade Collie on a VPS, migrate again after updating, so the VPS gets the new version too.
Links
- Collie Gold on the MQL5 Market, with the full results table: https://www.mql5.com/en/market/product/200222
- The user manual: https://www.mql5.com/en/blogs/post/777124
- The Lucky Dogs channel, for updates and questions: https://www.mql5.com/en/channels/021c0d9e2157dd01
- The Lucky Dogs users' group on MQL5 (setups, results and questions about all our EAs): https://www.mql5.com/en/messages/02ac45d27c5ddd01
Trading gold with leverage carries a risk of significant losses. The results in this post come from backtests in the Strategy Tester; live results will differ. Past results, whether tested or live, do not guarantee future results.


