Contextual AI: The Next Step Beyond Raw Market Data in Trading

When two traders receive the same market signal, they can interpret it differently. One might be assessing a fresh trade, while the other could already hold a large position in that same asset.

At iFX EXPO Asia 2026 in Hong Kong, that was the key idea I addressed. Artificial intelligence must take a trader’s personal situation into account, not just market data.

From Trading Signals to Contextual Intelligence

Machine-learning models can sift through historical market data to uncover patterns that may signal trading possibilities.

A signal based solely on past prices cannot, on its own, factor in a trader’s existing positions. If it recommends buying EUR/USD, someone already heavily long would need to verify that increasing their exposure stays within acceptable risk parameters.

Contextual intelligence examines how three categories of information interact when trading decisions are made.

• Personal context: the trader’s approach, available capital, trading patterns, and risk appetite.
• Market context: economic events, volatility, and liquidity, which affects how easily an instrument can be traded.
• Trade context: details about the position, such as its size, planned holding duration, and stop-loss or take-profit levels.

When market volatility climbs while a trader has an open position, the range of possible gains and losses can expand. Whether an adjustment is needed depends on the size and direction of the trade, the trader’s strategy, and their risk limits.

If the trader adds to that position, their exposure grows. A contextual AI system would need to re-evaluate the trade against the trader’s strategy and risk boundaries, factoring in the increased volatility.

Scenario Analysis Before a Market-Moving Event

Take a trader who holds a forex position half an hour before a major economic release. An AI workflow could model how the trade might respond to various market outcomes, drawing on current market data and details of the trader’s exposure.

Rather than predicting what the market will do, the analysis would lay out potential scenarios.

The trader could then decide if the position still fits within their risk limits. Based on their plan, they might keep it open, cut exposure, or exit before the announcement.

Different Markets Can Carry the Same Risk

A trader may hold long positions in gold, EUR/USD, and the Nasdaq-100, thinking the risk is spread across separate markets. Yet all three can move in response to shifts in US interest-rate expectations.

On 23 September 2022, gold, EUR/USD, and US equity indexes all fell as markets anticipated further rate hikes and rising recession fears. A trader with buy positions in all those markets could have seen losses across multiple trades at once.

A contextual AI system could look at those overlapping exposures and estimate how simultaneous market moves might impact the account. The trader could then evaluate total exposure before deciding whether adjustments are warranted.

AI Can Generate a Strategy, That Does Not Validate It

Near the close of my iFX EXPO discussion, I flagged a concern about how some traders use generative AI.

A trader might ask a chatbot for a profitable trading strategy and receive detailed rules, an explanation of the approach, or even code.

A thorough answer can easily be confused with a validated one.

Large language models can produce false information and rely on unsupported assumptions. Without reliable data, a clear methodology, and proper testing, traders have little basis to assess whether the suggested strategy is robust.

I would urge traders to begin with a specific issue in their trading process and use AI to explore it using trustworthy data and a well-defined approach. The results can then be compared against actual trading records and live market conditions.

In my view, the next phase of trading AI involves moving toward contextual intelligence systems that extend existing trading tools. Such systems could help traders assess opportunities, manage positions, and review performance—with the analysis adapting as markets and individual circumstances shift.

Ultimately, trading AI should be judged not by how convincingly it produces a signal, but by how well it helps traders grasp the risks, assumptions, and consequences behind a potential trade.

AI can expand the analysis. Context can sharpen it. Human judgment remains indispensable.

Dr Parker Tian is Head of the EBC AI Lab at EBC Financial Group. At iFX EXPO Asia 2026 in Hong Kong, he presented The Data Isn't Enough: Contextual Intelligence for Trading Decisions, examining how AI-driven analysis can aid trading decisions.

EBC Financial Group was named Most Trusted FX Broker and Best CFD Broker at the World Finance Forex Awards 2026. At the UF Awards APAC 2026, held alongside iFX EXPO Asia in Hong Kong, EBC Financial Group received the Best Broker – APAC award.