Theorize
AI reads global economic news, earnings reports, and market signals. It forms investment theories grounded in data.
Backtest
Run theories against historical data with institutional-grade backtesting. Walk-forward optimize to avoid overfitting.
Allocate
Deploy capital with dynamic exposure management. As conditions change, your portfolio adapts — with human approval or autonomously.
Test ideas before you risk capital
Every strategy can be validated against historical data before a dollar moves. Thesis provides the same caliber of backtesting tools that quant funds charge six figures for — rebuilt for speed, composability, and agent interaction via MCP.
- Multi-asset backtesting with realistic fills and slippage
- Walk-forward optimization to validate out-of-sample
- Monte Carlo simulation for stress testing
- Full MCP exposure so agents can run backtests autonomously
Let Claude be your research analyst
Thesis is MCP-first. Your AI agents can read market data, form hypotheses, construct strategies, run backtests, and propose allocations — all through tool calls.
- thesis.* MCP tools for every capability
- Human-in-the-loop or autonomous modes
- Composable with Context for full financial picture
Start simple. Go pro when ready.
Agency includes Era-managed model portfolios for hands-off investing. When you're ready to take control — to define your own strategies, run backtests, and manage exposure — upgrade your brokerage to Thesis. Same account. Same data. More power.
The tools quant funds use. For everyone.
Thesis is in private beta. Join the waitlist to get early access to backtesting, walk-forward optimization, and agent-assisted portfolio management.