We give you tools to generate Alpha, not signals.
Give it a thesis. It returns the highest-scoring stocks and ETFs, plus the option strategy with a price edge.
The best trading regime and strategy for any single stock or ETF.
Undervalued stocks set to rise toward theoretical value, and overvalued ones due for a correction.
Your analysts read all of it. Your CIO combines it into one ranked setup per name: the best fundamental, technical, and option structure together. Not a signal on a single indicator. The whole stack behind every trade.
Positions flow in from your desks or you add them manually. Every night it marks the book, checks exposure, Greeks, stops, and drawdown against your limits, and proposes a rebalance when one trips. You approve every change. It never trades for you. Shipping to beta soon.
since Jan 2 · EOD marks · sample portfolio, live in the Forward Test
Cash sits below the 10% floor. The plan trims the largest LEAPS gainer, closes PYPL before its stop, and sweeps the rest to money market.
A book built with edge, not the model portfolio every advisor sells.
Screenshot artists on X sell entries. Nobody prices them.
Your money dies after the fill. Thesis gone, no stop, still holding.
Influencer track records are screenshots. Wins posted, losses deleted.
Figuring out what works is a full-time job. Expensive & scattered sources, stale data, you do the syncing.
Every trade priced against theoretical value, scored on fundamentals, technicals, and macro. The gap is the edge.edge +9.8%
Every idea ships with its invalidation. The book is watched nightly. Thesis breaks, exit proposed.
We display a public, append-only log. Losses included.
One data spine. Vetted sources, updated every session, already synced. You just ask.
Unedited runs of the actual system. No mockups.
Your algos are built for specific purposes. Our desks scan the universe for the best names and option strategies, define the regime and strategy for any single ticker, and surface undervalued stocks from complex models, in 30 seconds.
We tailor our models and deploy them on your infrastructure, optimized to your investment needs.
Three tiers. Real prices. Cancel anytime. The forward test is the same on every tier.
Depending on data, it refreshes every 5 minutes, 1 hour or EOD. Universe-wide scans take substantial compute and time for an LLM-supported deterministic system, so running on live tick data is not feasible at this stage. StockAgents is not built for high-frequency trading. It is built for longer-horizon strategy generation: days, weeks, and months.
The gap between what an option actually trades for and what our model says it should trade for, given the underlying's volatility surface and term structure. We use stochastic-volatility-inspired (SVI) parameterization to fit the surface and price the option against it. Positive edge means the option is undervalued (long bias). Negative edge means it is overvalued (short bias). The number renders as a percentage on every trade idea on the platform.
A Blueprint is a saved find. Save any Scout scan, Deep Dive ticker setup, or Mispriced book filter. You define the types of alerts and run frequency, and it alerts you when something changes: a new name crosses the threshold, an existing match falls below it, or the top edge shifts. Alerts go to push, SMS, and email on Pro and above.
Positions flow in from Scout and Deep Dive, or you add them manually. Every night the manager marks the book and checks exposure, Greeks, stops, and drawdown against your limits. When a limit trips, it proposes a rebalance. You approve every change. It manages the book with you. It never trades for you.
Not yet. Broker connectivity is on the roadmap and will ship after the public beta release. For now, StockAgents is a research and signal layer. You take the trade in your broker of choice.
Honesty about cost. Frontier models cost real money per query, and a single deep analysis can cost many times a basic scan. Session limits let us keep real prices on the page instead of hiding the economics behind a contact-sales button.
For funds and quants, yes. The scoring engine, Mispriced, Deep Dive, and Catalyst are available over a private API. We also tailor our models and deploy them on your infrastructure, optimized to your investment needs. Talk to us.
No. StockAgents is a research tool. Trade ideas, scores, and Blueprint outputs reflect historical and forward-tracked performance only and do not guarantee future results. Options trading involves substantial risk of loss and is not suitable for all investors.