Skip to content
All library documents

Modeling Retail FX Trading with Data and Behavioral Trade Repertoires

Article Quant Q&A · Author: xyzt
Read the original

Summary

The document asks how to make a trading bot imitate a typical nonprofessional foreign exchange trader, including whether such traders can be treated as acting randomly regardless of market conditions. The response rejects assuming random behavior as a sound starting point. To imitate actual trading, it recommends first collecting data and using it to identify patterns in how retail traders behave.

As an alternative when direct activity data is unavailable, the response suggests assembling a set of candidate trades from books or online forums, then using ideas from psychology to model how a trader might choose among them. These are high-level suggestions, not a specified behavioral model or a tested bot design. The document provides no dataset, measurable behavioral findings, implementation details, or performance evidence, so it establishes possible research directions rather than demonstrating that either approach can reliably reproduce retail FX trading.

Key ideas

  • Do not assume nonprofessional FX traders act randomly without supporting evidence.
  • Use observed trading data to identify behaviors a model should imitate.
  • A repertoire of trades from books or forums can provide candidate actions.
  • Psychological ideas may help model how a trader selects among candidate trades.
  • The suggestions are conceptual and are not validated as a predictive or profitable bot.

Tags

ForexMachine learningSentiment
Full text
# How can I model the way that average Joe trades in FX markets?


# How can I model the way that average Joe trades in FX markets?












I want to develop a trading bot that trades like a non-professional trader in FX markets. I wonder if it is possible to think that an average trader trades randomly regardless of the market conditions? Or is there a behavioral model to model non-professional traders' trading activity?

## Answer by Stéphane (score 1)

https://quant.stackexchange.com/a/51421

I don't think that they are trading randomly. You'd need some data to figure out what they tend to do to imitate it.

Or, you can build a repertoire of trades from books or forums and exploit knowledge from psychology to determine how the bot will choose between them.

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.