I made a real-world AI hedge fund team. It has 4 agents: 1 • market data agent 2 • quant agent 3 • risk manager agent 4 • portfolio manager agent You can run all of my code below. No prior coding experience is required. The agents work sequentially,... passing messages to the next agent. 1 • Market Data Agent gathers market data and calculates technical signals. 2 • Quant Agent takes market data and generates a buy, sell, or hold signal. 3 • Risk Management Agent takes trading signal and generates a risk signal. 4 • Portfolio Management Agent considers all signals and makes a final trading decision.show more

Code:

A glimpse into the manager economy where we are just managing agents. Curious if you've dabbled into evals?

The included code runs backtests (evals).

Backtests finally... Good job Virat.

Thank you - evals are everything.

When I worked at a quantitative trading fund, the process involved submitting models to a portfolio manager, who would then evaluate and select models from a pool created by multiple quants. How does this approach compare to what you're building? A significant aspect of quantitative finance isn't just identifying technical signals but also testing and developing new, innovative ideas - often based on economic theories, market structure, personal experience, or other underexplored dynamics. If your AI agent is focused solely on creating models from existing signals and historical data, it seems to be more of a model optimizer than a true "quant."

I swear the best use of one’s time is learning how to “hire” - in other words or code up - a company of agents to do the work a company of humans does today.

How is this an agent and not a SaaS with 4 modules?

ChatGPT's knowledge cutoff should be before the backtest to avoid forward-looking issues.

Nice work. Code base looks clean.