[Episode 9] Shōkinryū Why not move with anything other than gold—The choice of "one currency pair specialization"
“Can this EA be run on a currency pair other than USD/JPY?”
“Is it compatible with EUR/USD as well?”
This is one of the most common questions about Shōkinryū. You can understand how you feel. If it can operate on multiple currency pairs, there are more profit opportunities and risk diversification. Most of all, it is convenient due to reusability. Intuitively, a general-purpose EA seems more convenient.
However, Shōkinryū’s answer is clear.“Do not run it on anything other than gold.”This is not because we want to limit performance or be conservative. There is a strength that can be realized only by specializing in one currency pair. This is my conclusion.
This time, I will explain why the intuition that “multi-currency support equals diversification” is wrong, and why I chose single-currency focus, with its rationality.
First, as a premise, please note how the “characteristics” of price movements differ between currency pairs.
Gold’s daily movement range is $50–$100. By contrast, USD/JPY is 50–100 pips, EUR/USD about 50–80 pips. The absolute movement ranges differ, and what drives them—geopolitical risk, interest rate differentials, central bank policy differences—are entirely different. The character of movement is also varied: “sharp rises and falls,” “trending,” “range-bound,” and “erratic.”
Consider this:Do you think an EA with the same parameters would optimally perform across all these completely different currency pairs?The answer is clearly no. A logic that profits in gold does not necessarily profit in USD/JPY. The reverse is also true.
Nevertheless, some EA claim “multi-currency support,” using a coarse, common parameter set for each currency pair. It may be acceptable for any one currency, but never the best for all. This is the reality of generic EAs.
The strengths Shōkinryū achieves by concentrating only on gold are threefold.
All parameters—RSI2 threshold, ATR dynamic grid spacing, take-profit, stop-loss, entry pause time—are optimized for gold’s quirks. For example, gold tends to move more in the New York session, so grid spacing is designed with the price range in that period in mind. This level of tuning is impossible with a generic EA.
All 11 years and 22,189 trades are gold data. If this were a 5-currency EA, each currency pair would have only about 4,000–5,000 trades in the same test period. Focusing on a single currency pair dramatically increases the density of validation, yielding statistical reliability.
As a developer, I face gold charts every day. I hardly look at other currencies because deep understanding of one pair is far more valuable than shallow understanding of many. This depth is reflected in EA updates.
The claim that “multi-currency EAs reduce risk via diversification” sounds appealing, but contains a major misconception.
The real purpose of diversified investing is“to reduce overall risk by combining assets with different price movements”The key is different price movements, not simply different currency pairs.
Even applying the same logic to multiple currency pairs may still cause all pairs to suffer if the overall market is risk-off. Conversely, combining different logics (for example, a trend-following with a momentum and a mean-reversion scalper) allows one to profit when the other struggles, achieving true diversification.
In other words“Running multiple single-currency-focused EAs in parallel is the strongest combination.”I currently run gold-focused Shōkinryū (buying/scalping) alongside CHAOS_GOLD (hedging/day trading), both gold-focused but with completely different logics, so when one is in drawdown the other can realize profits, providing true complementarity.
If single-currency focus is advantageous, why gold? There are three reasons.
Reason ①: Structural upward trend
Gold rose from the $1,000s in 2015 to the $4,800s in 2026, about fourfold over 11 years. Although there are short-term drops, the long-term trend is upward. For a buy-focused EA, this structural tailwind alone is a significant advantage.
Reason ②: Abundant profit opportunities
Gold’s daily range is $50–$100, vastly larger than major currency pairs. For a scalping EA, “range” translates directly to profit opportunities. Scalp trading a currency pair with little movement yields little profit because you cannot reach take-profit.
Reason ③: Persistent demand
Gold is supported by central bank reserve assets, investment demand, and industrial demand. Since 2022, central banks have been buying over 1,000 tons annually, a trend likely to continue. When considering whether demand for this currency pair will disappear in the future, gold is among the safest bets.
Suppose Shōkinryū is run on USD/JPY; what problems would arise concretely?
The take-profit lines assume gold’s range (tens of dollars). If applied to USD/JPY (tens of pips), the distance to reach profit is too long, so you enter but never close, creating numerous dredging positions.
ATR-based dynamic grid spacing bases the spacing on the currency pair’s volatility. Gold’s ATR is $10–$30. For USD/JPY, it’s 20–50 pips. The same logic yields inappropriate unit sizes, causing over- or under-nanping and deviating from intended behavior.
Gold spreads vary widely by broker and time (especially at major news). Shōkinryū’s spread filter is designed around gold’s normal spread. Applying it to USD/JPY could either skip entries due to wide spreads or enter under unfavorable spreads if gold-based criteria deem the spread too narrow.
The most critical issue. All 11 years of testing are gold data. Running on USD/JPY provides no guarantee from that backtest; you would be testing unverified logic.
“Multi-currency support” may sound convenient, but it is essentially a design that compromises across all currencies. Specializing in one currency pair yields far superior performance for that pair. This is the truth of EA design.
If you want diversification, don’t run the same EA on multiple currencies, but“run multiple EAs with different logics in parallel”. I personally run gold-focused Shōkinryū and another gold-focused but different logic, CHAOS_GOLD, to create complementarity.
Concentrating on a single currency pair, gold, is not limiting performance but a strategic choice to maximize it. The figures 11 years, 22,189 trades, and 84.28% win rate came thanks to this specialization strategy.
If you want to run EAs on other currency pairs, use specialized EAs for those pairs. That is my honest advice. Next time we will cover a topic many are curious about: starting with 1,000,000 yen and 0.01 lot, and how far compounding can grow it.
※This article provides information only and is not investment solicitation. The results shown are past performance and do not guarantee future profits. FX/CFD trading involves risk. Please make investment decisions at your own responsibility.
