KLab Begins Development of an AI-Powered Automated Trading System for Financial Products

~ Simulations show more than double the benchmark performance ~

KLab Inc. (Head Office: Minato‑ku, Tokyo; President & CEO: Tetsuya Sanada; hereinafter "KLab") announces that since last year it has been developing an AI‑based automated trading system for financial products (hereinafter the "AI"). Backtesting is currently underway, and the test results have been favorable.

What Is the AI-Powered Automated Trading System?

By learning the relationships between various types of information and market price movements, the AI:

1. Collects information autonomously in real time
2. Predicts market price movements
3. Generates signals such as "buy" or "sell"
4. Enables users to execute automated trades based on those signals

When markets move significantly, human judgment can become unstable due to anxiety or fear. In contrast, AI can make mechanical decisions based on learned data.

Foreign exchange and Bitcoin prices fluctuate 24 hours a day, 365 days a year. AI continues collecting information even while people sleep and executes trades at optimal timing.

AI That Watches Not Only Charts but Also News and Social Media

Many conventional automated trading systems have focused primarily on charts such as prices and trading volume. However, in real markets, sudden movements can be triggered by policy announcements, remarks by key figures, or reactions on social media--factors that cannot be captured by market data alone.
The AI aims to collect and evaluate the following information in real time:

- Market Data

　Charts (price, volume, order book information, etc.)

- News

　Economic indicators and statements by key figures (media reports, official announcements, etc.)

- Expert Analysis and Interpretation

　Market Analysis and Insights by Financial Analysts

- Market Sentiment

　Reactions on social media and other platforms

By using diverse information sources as decision factors, the AI supports trading from short‑term to medium‑ and long‑term horizons.

Example 1: When Major Announcements Cause Sudden Market Changes

The AI evaluates not only abrupt chart movements but also the content of news and the spread of reactions on social media to determine actions such as:

• Reducing trading volume
• Switching to a wait‑and‑see stance
• Changing to risk‑controlled orders

and adjusts automated trading behavior accordingly.

Example 2: When Signs of Overheating (Overbought/Oversold) Appear

The AI shifts decisions toward:

• Strengthening profit‑taking
• Suppressing new entries
• Preparing for opposite‑direction risk

and adjusts automated trading behavior accordingly.

*Specifications are under development and subject to change without notice.*

Roadmap

1. Development of basic and applied prediction models

Development of an AI system that predicts market price movements based on training data

2. Learning and accuracy improvement through backtesting 　　　

Historical data is applied to the developed logic so the AI can learn through repeated trial and error to improve precision

3. Learning and accuracy improvement through demo trading

Using real‑time information with virtual funds to further train the AI and enhance precision

4. Live operation using proprietary funds

Actual automated trading conducted with company funds

5. Service provision for general users

Commercialization after completion of required legal procedures

Expansion to Bitcoin, Foreign Exchange, Gold, and Stock Indices

KLab is promoting the "Dual Gold Treasury Strategy," a financial strategy combining Bitcoin (digital gold) and gold (real gold), and will continue holding both assets.

The AI consists of a basic prediction model applicable to all financial products and applied prediction models that consider the specific characteristics of each instrument.
Considering synergy with the Dual Gold Treasury Strategy, development has first begun with applications for Bitcoin.

Going forward, the system will be expanded sequentially to foreign exchange, gold, stock indices, and other financial products.

*The "KLab with AI BTC Report," which collects, reviews, and analyzes analyst reports on Bitcoin and gold from around the world using AI, is a completely separate system from the AI.*

Backtesting Shows Performance More Than Double the Benchmark

Backtesting is progressing smoothly. In testing, the AI has produced performance exceeding more than double the benchmark of Bitcoin buy‑and‑hold. It also recorded positive performance during Bitcoin downturns by taking short positions.

However, these are simulation results based on historical data and do not guarantee future performance.

The project will soon transition to stage 3 (demo trading), followed by security measures and then stage 4 (live operation using proprietary funds).

Aiming for Global Commercialization

KLab was founded as a mobile software R&D company and has grown primarily through its mobile game business. Since fiscal year 2025, AI has been positioned as a key business area, and the company has been conducting R&D and business planning for numerous AI systems, including unannounced projects. The AI is one of them, with proprietary-fund operation planned for 2026 and commercialization targeted for 2027.

Because the AI targets borderless financial products such as Bitcoin and foreign exchange, service deployment is expected worldwide. Service countries will be determined based on regulatory frameworks and marketing considerations in each jurisdiction.

Disclaimer

- This document is not intended to solicit the acquisition of any specific financial products or investment activities.

- The plans and forecasts contained in this document are based on currently available information and are subject to change due to future changes in market conditions, laws, regulations, etc.

* It is assumed that businesses requiring licenses and permits will be implemented after obtaining the appropriate permits.

Return to List Page