Break Hunter KINAKO – conservative breakout grid EA for GOLD (forward test log, Week 1)

1. What is Break Hunter KINAKO?

Break Hunter KINAKO is my own experimental EA for XAUUSD (GOLD) on MT4.

It is a breakout + limited grid system designed to:

• Jump on strong breakouts in the direction of the move

• Use small, controlled martingale only when needed

• Keep drawdown as low as possible while still trading actively

This post is a simple forward-test diary, not a sales page or investment advice.
I’m sharing the numbers and behaviour so far in case it helps other EA traders who like GOLD and breakout logic.

2. Forward test environment

For this first week I used:

• Account type: Demo, cent account (HFM)

• Balance: JPY 50,000 (demo funds)

• Symbol: XAUUSD

• Platform: MT4

• Timeframe: M15

• Mode: 24h full auto

• Initial lot: 0.10

The EA has both “aggressive” and “defensive” parameter profiles, but for this test I used a setting that trades about 20 times per day on average.

There is martingale (averaging) logic inside, but it is very limited and mostly stays at 1–4 steps.
Most of the profit actually comes from the first breakout entry.

3. Week 1 forward results (28 Nov – 6 Dec)

Forward-test period: 2025-11-28 to 2025-12-06

Summary:

• Total profit: +1,540 JPY

• Max drawdown: -1.35%

• Number of trades: 140

• Max grid / martingale steps: 4 levels

So far, this is exactly what I wanted to see for a “conservative breakout grid” idea:

• Active trading with many small wins

• Very controlled floating drawdown

• No extreme lot explosions or scary equity swings (yet!)

4. What I like about the behaviour

A few things that stood out to me in the first week:

• Sharp entries on breakouts
  The EA often catches clean breaks of recent highs/lows.
  When the breakout is strong, the first order alone can ride a decent move without needing any grid.

• Calm handling of retracements
  When price moves against the first position,
  the EA adds a few extra entries (max 4 steps in this test) and waits for a reasonable pullback.
  So far, drawdown has stayed relatively small compared to the lot size and frequency.

• Matches my original design idea
  I wanted something that can run, for example,
  0.10 lot on around 50,000 JPY of margin without blowing up the account in a few days.
  As long as the floating loss can stay within roughly 10,000 JPY in the worst case,
  I’m happy with this concept.

Of course, one week of data is not enough to draw big conclusions,
but the behaviour pattern is exactly what I was hoping for.

5. What I’m still watching / what could be improved

Even though the first week looks good, there are still many things to monitor:

• How the EA behaves during major news
  GOLD can move very violently on FOMC, NFP, CPI, etc.
  I need more data in those environments and may add stricter news filters if needed.

• Worst-case grid scenario
  In this test the maximum grid depth was 4 steps.
  I want to see how the EA behaves when it actually reaches that limit more often,
  and whether the risk is still acceptable.

• Optimal lot size vs. balance
  For 50,000 JPY balance, 0.10 lot might be slightly aggressive.
  I will also test more conservative profiles (e.g. 0.05) to see how much the stress on the account changes.

• Broker differences
  This first run is on one broker and one account type.
  Spreads, swaps, execution speed, and tick quality can change the behaviour,
  so I will need to compare a few environments.

6. Plan going forward

My plan for Break Hunter KINAKO is:

1. Continue the forward test under the same conditions for several more weeks

   • Same broker, same account type, same parameters

   • Weekly reviews of profit, drawdown, and trade quality

2. Then test slightly different risk profiles

   • Lower lot / higher balance

   • Maybe time filters for specific sessions (e.g. avoid rollover)

3. After I’m satisfied with the stability,
   I’ll decide whether to:

   • Release it publicly,

   • Keep it as a private tool,

   • Or build a separate “World Edition” like I did with some of my other EAs.

For now, this is still a research project and a forward-test log.

7. Disclaimer

This post is only a personal record of my EA experiments.

• The test was done on a demo cent account, not on live funds.

• Past performance (even forward tests) does not guarantee future results.

• GOLD is a highly volatile instrument and any grid / averaging logic can suffer large drawdowns or even total loss of capital.

Please do not treat this as investment advice.
If you experiment with similar ideas, always use demo accounts first and only risk money you can afford to lose.

It is Demo Forward now.