Verified record
—

Net return since inception, after fees.

Every trade settled on Kalshi and logged.

§ 01 · Membership

Learn the method

Every 15 minutes, our model's fair price for each market you subscribe to.

Fair value is what a contract should cost if the odds are right. Members see it for their markets, every window, along with the settled trades and the lessons behind them.

Members only
What it runs on
  • Stochastic calculus and probability modeling
  • Realized-volatility models on live market data
  • Quantitative, fully automated execution
  • Low-latency infrastructure
  • Systematic risk management
  • Every rule validated on data before it trades

Prices are placeholders until launch.

§ 02

Who we are

A builder and a psychologist. The system, and the discipline to follow it.

Co-founder · Systems

Harvard. Background in private equity and quantitative trading. Designed and built the system: the execution engine, the risk rules, and the data testing behind every rule.

Co-founder · Behavior

UCLA, Psychology. Training in statistics and research methods. Leads the behavioral side: why traders lose to overconfidence, chasing losses and revenge trading, and how to measure it in the data.

Not affiliated with or endorsed by Harvard or UCLA.

§ 03

The record

Every settled trade, both accounts, after fees. Delayed, never live.

More numbers
Each line is one trading day, every trade in orderDrawn from the record
§ 04

The right way to trade

Eight rules. None are secret. All are harder to follow than they look.

01

Only trade when the price is wrong

If the edge doesn't beat the fees, we pass.

02

Size small

A bad day should sting, not hurt.

03

Cap the worst case

Hard loss limits, set before the market opens.

04

Automatic circuit breakers

A loss stops everything. No exceptions, no overrides.

05

Never chase

No doubling up to win it back.

06

Discipline over prediction

We don't try to out-guess the market. We follow the rules every time.

07

Read the whole record

Wins are small, losses are bigger. Only the total counts.

08

Test before you trust

Nothing trades until the data says it works.

Ideas we tested and dropped (3)
  • Machine-learning price model Dropped
  • Chasing sharp price moves Dropped
  • Trading for rebates Dropped
§ 05

Every trade

One square per settled trade, oldest first. Green won, ink lost.