Skip to content

Trading Agents

Paper trading with a path to live trading, gated by constitutional vote.

Paper Trading

All trading begins in paper mode using simulated portfolios:

Model Purpose
PaperPortfolio Tracks simulated holdings, cash balance, and total portfolio value.
PaperTrade Records individual buy/sell transactions with timestamps, prices, and quantities.

Paper trading runs continuously, building a track record that can be evaluated before any real money is at risk.

Strategy

The default strategy is long_term diversified:

  • Allocates across multiple asset classes.
  • Rebalances periodically based on drift thresholds.
  • Optimizes for risk-adjusted returns over months, not minutes.

Auto-Funding

When platform revenue exceeds a configurable threshold, a portion is automatically allocated to the paper trading portfolio. This grows the simulated portfolio in proportion to platform success, providing a realistic test of what live trading would look like.

Live Trading Activation

Transitioning from paper to live trading requires a constitutional vote via the thought experiment system (see thought-experiments.md):

  1. A thought experiment is created proposing live trading activation.
  2. Network peers review the paper trading track record.
  3. A majority vote is required to approve.
  4. Only after approval can real funds be deployed.

This ensures no single operator can unilaterally risk platform funds.

Source Files

  • integrations/social/models.py (PaperPortfolio, PaperTrade)
  • integrations/agent_engine/ (trading goal execution)