Trading Automation

Every trade, the way it was planned.

We help professional traders and proprietary trading desks put their process into code, so it watches, decides and acts by the same rules every time, in any tradeable market.

The problem

Plans rarely fail on knowledge. They fail in execution.

Most traders know what they should do. The gap opens in the moment, when discipline is tested and no one is watching.

Stops get widened

A loss that should have stayed small is given room, and becomes the trade that sets back the month.

Entries get skipped

A valid setup is passed over after a run of losers, and the winner that would have paid for them is missed.

Setups get missed

Markets move while you sleep, travel or watch something else. A rule-based system doesn’t look away.

How it works

Watch. Decide. Act.

Three jobs a trader does by hand, done by a system that applies your rules exactly as written and keeps a record of why.

  1. 01

    Watch

    Every instrument in scope, monitored continuously on live market data, with scanners and automated alerts.

  2. 02

    Decide

    Precisely written setups, applied identically each time, so every decision can be reviewed afterwards.

  3. 03

    Act

    Entry, sizing and exit handled with a record of the reasoning, inside limits that don’t bend on request.

Opening-range breakout, five-minute bars

    Illustrative
    Stop and target are set at entry and never moved by hand. Illustrative data.

    Decision log

    1. WatchOpening range set at 5,812.25 to 5,826.50
    2. DecideBar closes above the range high. Setup confirmed
    3. ActLong 2 at 5,827.00. Stop 5,815.75, target 5,849.50, risk within limit
    4. ActTarget reached. Exit 2 at 5,849.50 for +2.0R
    5. RecordTrade written to the log with its reasoning

    Every step is recorded, so any trade can be reconstructed.

    Markets

    Any tradeable market.

    Equities, futures, options or crypto, on any venue with reliable market data and execution access. The discipline is the same wherever the instrument trades.

    How we build it

    From written rules
    to a live system.

    Nothing is automated until it has been traded by hand, written down precisely and tested on data it has never seen.

    01

    Specify

    Each setup, traded by hand first, is written precisely enough to code: entries, exits, sizing and limits.

    02

    Validate

    Backtested on the longest reliable history available and validated on out-of-sample periods before anything trades.

    03

    Build

    Position sizing, risk limits and execution enforced in code, never by hand.

    04

    Monitor & refine

    Scanners, alerts and a record of every decision. Each issue is traced to its cause and fixed, so the system gets stronger with every fix.

    Validation

    Tested on two decades of data, judged on data it never saw.

    Our own strategies are backtested across more than 20 years of market data. The most recent years are set aside as an out-of-sample period and used once, as the final test before anything goes live. Client strategies follow the same method, on the longest reliable history their market offers.

    • Backtests on the longest reliable history available
    • Out-of-sample periods kept apart from strategy design
    • Nothing goes live until it passes out-of-sample

    Strategy backtest, indexed to 100

    Illustrative
    Log scale, with drawdown from the previous high below. The shaded out-of-sample period is never used to design or tune the strategy. Illustrative data showing the method, not the performance of any strategy.

    In practice

    We build on what we run.

    Our own strategies went through this process before they were deployed on live capital: specified, backtested, validated out-of-sample, then automated.

    20+ years

    of market data behind their backtests, with the most recent years held out as the final test.

    In code

    Position sizing and risk limits are enforced by the system, never by hand.

    Real time

    Scanners and automated alerting running on live market data.

    12+ months

    Live in production. Issues are diagnosed and fixed at the root, and the systems are self-improving.

    We design, build and validate systems that clients run on their own accounts, for professional traders and proprietary trading desks outside India, and in India only where we hold the requisite registrations. We do not manage client money, operate client accounts or accept client capital.

    Contact

    Start a conversation.

    Tell us about your trading process and the markets you trade, and we’ll come back to you directly.