Investors and partners
A hedge fund in the making.
IndiQuant trades Indian equities on signals from independent researchers. A trust-weighted meta-model combines them, and a portfolio optimiser turns the result into one risk-managed book.
Rounds, scoring and the meta-model run every NSE session today. Paper trading of the book starts next, and live capital follows within the next few months.
Thesis
Crowdsourced signals, one disciplined book.
Good equity signals are scarce, and they fade. A single research team tends to find the same kind of signal over and over. We hand anonymised NSE data to a wide pool of independent researchers, score what they send back against realised returns, and keep what holds up out of sample.
Contributors never see tickers or company names, and they never see the book. What reaches the portfolio is the meta-model: their signals, weighted by each model's scored record.
India is the market we chose to do this in. We trade NSE equities on a point-in-time universe, with Indian trading costs built into the optimiser from the start.
Phases
Where we are.
01
Proving
Now
The whole pipeline runs end to end every session: daily rounds, scoring, the meta-model, the weekly portfolio and execution. The point is to prove each stage works, and keeps working, before real money is involved. Core rounds began on 21 September 2026.
02
Live capital
Next, within the next few months
Live capital, through an institutional execution partner, trading on IndiQuant's signals. The pipeline and the risk limits stay the same, and it starts once the record from the first phase and our own checks say it is ready.
03
Fund
After a live record
A fund for eligible investors, set up under the applicable regulation once live trading has a record to show.
Portfolio and risk
How the book is built.
Everything below runs every session today, ahead of live capital.
- Signal
- Contributors predict a 20-session, sector-neutral return for every name in the universe. Submissions close at 09:00 IST each NSE session.
- Meta-model
- After each session the meta-model is rebuilt from the latest scores. Each model is weighted by its trust, an estimate of skill from its scored record, and near-duplicate models are discounted so one idea sent twice does not count twice.
- Rebalance
- Weekly, on the first session of the week. With a 20-session target the book holds positions for weeks, not hours.
- Optimiser
- A constrained optimiser sets the weights. Gross and net exposure, sector, liquidity and turnover limits sit inside the solver, and one-way turnover is capped at 30% per rebalance.
- Costs
- Trading costs are priced into the objective rather than subtracted afterwards, so a trade has to cover its cost before the optimiser will make it.
- Execution
- Orders are built inside the Execution Trust Zone, a separate environment with its own credentials. It re-checks every limit before trading, and stock-level signals never leave it.
- Live guard
- The Execution Trust Zone routes no real orders yet. A runtime guard refuses live orders until live trading is deliberately switched on.
Entity structure
One company, two offices.
Talk to us
Follow the build with us.
If you are a prospective investor, allocator or partner, write to office@indiquantresearch.in. Tell us who you are and what you would like to see, and we will reply by email.