Global Open Competition · 2026

SoAI 2026
AI Algorithmic
Trading Competition

Build. Deploy. Trade. Compete.

Design and deploy AI-driven trading strategies in a fair, standardized global simulation. Compete worldwide. Present in Singapore.

Code Submission Deadline

--Days

--Hours

--Minutes

--Seconds

About the Competition

The Premier AI Trading Arena of 2026

The SoAI 2026 AI Algorithmic Trading Competition invites participants worldwide to design and deploy AI-driven trading strategies. Open to all backgrounds — AI, finance, computer science, mathematics, and beyond.

All strategies compete in a fair, standardized simulation environment: centralized execution, identical data feeds, zero hardware bias.

🏙️ Part of IntelligenceX 2026 – The Global Quantum × AI Frontier
24–26 September 2026 | Singapore

🏆

SGD 3,000

Total prize pool

📈

Aug 16 – Sep 15

Official trading window

🌏

Global Open

Worldwide eligibility

👥

1–5 Members

Individual or team

Key Dates

Competition Timeline

9 August 2026

Code Submission Deadline

23:59:59 SGT (UTC+8)

📅

🔍

10–12 August 2026

Verification & Test Run

Official validation by technical team

16 August 2026

Official Trading Begins

All strategies go live — 00:00:00 SGT

🚀

🏁

15 September 2026

Fund Closure & Final Results

Full portfolio liquidation at 23:59:59 SGT

18 September 2026

Winners Announcement

SGT — top teams invited to IntelligenceX 2026

🏆

Process

How It Works

STEP 01

🧠

Develop Your Strategy

Build a fully autonomous trading system using the Lumibot framework.

Signal generation & alpha research
Portfolio construction
Risk management
Inherit from lumibot.strategies.Strategy

STEP 02

📦

Submit to GitHub

Fork the template repo and push your strategy by the deadline.

Fully reproducible code required
All deps pinned in requirements.txt
Clear README documentation
No secrets or hardcoded paths

STEP 03

⚡

Compete Officially

IntelligenceX technical team executes all strategies identically.

Standardized, centralized environment
Identical data feeds for all participants
Zero latency or hardware advantages
Only official results count

Markets

Asset Universe

Trade across two major asset classes with unrestricted access.

💡 Trading Fee: A uniform fee of 2 basis points (bps) applies to all trades across both asset classes.

₿

Crypto Markets

Any CCXT-supported spot pair. Full access to the digital asset ecosystem — from blue-chip to long-tail. Orders subject to volume-aware slippage constraints.

BTC ETH Altcoins Meme Coins via CCXT

📊

US Equities

Full public market access via the Massive data provider. Large caps, small caps, ETFs, ADRs — the entire US equity universe available to your strategy.

Large-Caps ETFs ADRs Small/Micro-Caps via Massive

Developer Guide

Technical Requirements

🔧

Lumibot Framework

Your Strategy class must inherit from lumibot.strategies.Strategy. The template provides the base scaffold.

⚙️

Lifecycle Methods

Implement initialize() for one-time setup and on_trading_iteration() for each trading cycle — fetch data, compute signals, submit orders.

⏱️

Supported Frequencies

Minute, hourly, and daily cadences only. Sub-minute scheduling will be automatically rejected by the execution environment.

📡

Data Feed

Official feed provides OHLCV bars only. No order-book depth, news feeds, or alternative data sources are available during official execution.

strategies/strategy.py

from lumibot.strategies import Strategy

class Strategy(Strategy):

    def initialize(self):
        # One-time setup
        self.sleeptime = "1D"  # "1M" | "1H" | "1D"

    def on_trading_iteration(self):
        # Runs every sleeptime cycle
        bars = self.get_historical_prices(
            "BTC/USDT", 30, "day"
        )
        # Compute signal...
        order = self.create_order(
            "BTC/USDT",
            quantity=0.01,
            side="buy"
        )
        self.submit_order(order)

Judging Criteria

Evaluation & Fairness

Primary Metric

Terminal Return

Final portfolio value after full liquidation at market close on 15 September 2026 (23:59:59 SGT). The single metric that determines the leaderboard.

Only results generated by the official IntelligenceX execution system are valid. Self-reported backtest results do not count.

Centralized Execution

Fairness & Reproducibility

All strategies run in an identical, controlled environment. No participant gains an unfair edge.

Latency advantages eliminated
Hardware differences neutralized
Execution bias removed
Identical OHLCV data feeds
Pinned dependency environments

Prize Pool

Awards

SGD 3,000

Total Prize Pool

⚠️

Attendance Requirement: Cash prizes are awarded only to winners who present in person at IntelligenceX 2026 in Singapore.
No physical attendance → No cash payout. No money transfer or remote disbursement will be arranged.

Top finalists will present live and connect with

💼

Venture Capital Firms

📉

Quant Funds

🏦

Financial Institutions

Reasons to Compete

Why Join?

📈

Real Trading Scenarios

Apply AI in live market conditions — not just backtests. Your strategy trades real market data from 16 August to 15 September 2026.

🤖

Hands-On Experience

Gain deep expertise in algorithmic trading, signal engineering, portfolio optimization, and execution systems.

🌐

Global Showcase

Benchmark your performance against top AI and quant talent worldwide. Present your work at IntelligenceX 2026 in Singapore.

🤝

Investor Connections

Network with venture capital firms, quant funds, and financial institutions at the IntelligenceX 2026 conference.