Niti Pitakteeratham, Country Head – Japan at Banpu Public Company Limited (SET: BANPU), announced that the company has enhanced its energy trading capabilities in the Japanese electricity market under Banpu Power Trading G.K. by applying artificial intelligence to support power trading decisions.
BANPU developed AI models that integrate trader expertise with market data analysis and energy price trends to evaluate trading opportunities, set trade sizes, and manage risk, alongside elevating energy trading knowledge within its power generation and related business group (Power+). This marks a transition from traditional energy trading, which previously relied on traders’ personal experience, intuition, and situational decision criteria.
BANPU brought this expertise into a testable and auditable system prior to expanding and developing it into AI models. The system assists in analyzing procurement volumes, optimal timing for selling electricity into the market, and trading signal uncertainties, while human traders maintain oversight and retain final decision-making authority.
With the electricity market subject to volatility driven by demand, supply, weather conditions, and energy costs, timely and accurate data is critical for power trading decisions. As such, two AI models were developed to analyze market signals and manage risk.
The first model analyzes data to evaluate energy price trends and identify trading opportunities while assessing signal confidence; if a signal lacks clarity, the system can reduce trade sizes or refrain from trading to manage risk. The second model evaluates overall market conditions to select strategies and position sizes aligned with market dynamics.
Both AI models underwent simulated power trading tests between January and March 2026, utilizing historical data from the Japanese electricity market. The simulations achieved superior results compared to traditional trading methods by delivering a better balance between risk management and financial returns.
Following the testing, BANPU has deployed the AI models alongside traders to execute power trades in the Japanese energy market, accounting for more than 90% of the company’s total power trading volume across six regions.
The models operate in live environments as “intelligent assistants” supporting traders, with personnel reviewing AI recommendations and making final decisions before trade execution. This approach enhances decision-making efficiency, strengthens the company’s energy trading expertise, and fosters seamless collaboration between human capability and AI.
Additionally, Banpu is studying the feasibility of expanding these AI models to electricity trading markets in other countries as well as into other business segments to bolster organizational and workforce competitiveness in support of long-term energy business growth.
