First validation of Gonzo Special on domestic broker FXTF's MT5. Low-spread gold trading, dedicated version to be released soon
GONZO SPECIAL / 2026.09.07
Gonzo Special FXTF Edition
Gold short-term test and forward records from today.
At FXTF, gold (XAUUSD) trading with MT5 has become available.
Being able to use MT5, trade gold, and have low spreads. It’s the environment I’ve wanted for a long time to run Gonzo Special. I’m glad this option is increasing domestically.
FXTF’s MT5 began offering product CFDs and other services on September 1, 2026. I’m quickly confirming trades with Gonzo Special and proceeding with development of a FXTF-exclusive version.
Personal account product CFDs are subject to separate filing taxation
Profit from product CFDs on FXTF’s personal accounts is treated as “miscellaneous income related to futures trading,” with a tax rate of 20.315% including reconstruction special income tax. I’d like to check the tax system alongside the trading environment.
First, I compared at the same quantity
In the development stage of Gonzo Special, I compared FXTF with a “low-spread broker.” Initial margin is 1,000,000 yen, and the trading quantity of gold is aligned at 1 oz for both sides.
The real tick data obtained this time for FXTF is limited to after August 18. Among them, I confirmed the consistency of historical dataAugust 24–September 6 (excluding end date)and published the report.
Comparison version of development/M1・real tick・5 ms latency・20x leverage. Fixed quantity is 1 oz for both. Pyramid and rollover control are OFF.
FXTF with 5 trades +2,954 yenThe comparison target with 3 trades +1,834 yenBoth ended in profit.
Profit was higher with FXTF. However, the maximum drawdown including unrealized losses was 2,721 yen for FXTF and 663 yen for the comparison target. There are differences not visible by profit alone.
Top: low-spread broker / bottom: FXTF. This is a balance graph from the original report. The unrealized P/L trend is not included, and the vertical axis scales differ between the top and bottom.
This is still a short test of 3–5 trades. I won’t decide superiority from this result alone, and I’ll continue to observe actual trading as well.
Today’s forward: both sides take profit
Starting today, September 7, I began forward trading with FXTF as well. In the initial comparison trades, both entered selling almost at the same time, and both locked in profits within a few minutes.
Top: low-spread broker / bottom: FXTF. Because the displayed lot and actual traded quantity differ, you cannot directly compare the P/L amounts in the image.
When converting trade quantity to 1 oz of gold and comparing profits after deducting fees, the comparison target was +278.49 yenand FXTF was +304.00 yen.
In this single trade, FXTF came out slightly ahead. It’s encouraging to observe Gonzo Special’s characteristic short trades. I will continue trading to see similar movements and where differences arise.
Developing the FXTF-exclusive version
Gonzo Special can operate with FXTF in its current form, but detailed verification is not yet complete. Currently,we are developing a FXTF-specific version tailored from lot calculation to order placement and settlement to align with FXTF trading specifications. We will continue forward measurement and plan a forthcoming release.
FXTF’s gold is 1 lot = 1 oz. Handling the number of lots with the same feel as before may lead to unintended trading quantities. The exclusive version will allow settings based on FXTF’s actual lot size, and we will continue to verify TP/SL, trailing, and rollover controls.
This is the domestic trading environment we’ve been waiting for, so I plan to nurture it carefully. I will announce again once the release preparations are complete.
Content published as of September 7, 2026. The backtest results are from a development-stage comparison and have different settings from forward testing. Published results are from a short-term verification and do not guarantee future profits. Please review trading costs, including not only spreads but also fees corresponding to position size.