Raif Mondal
Founder & CEO
LinkedInAbout IndiQuant
A hedge fund built on measured research.
IndiQuant is building a quantitative hedge fund for Indian equities. Its signals come from independent researchers working on anonymised NSE data, and every one of them is scored the same way, on what the market did next.
Why we started
Most quant research happens behind closed doors.
At most funds a small internal team produces every signal the book trades. That limits how many ideas get tested, and the ideas tend to look alike.
We think a large, open pool of researchers, scored on identical rules, will find signals one team would miss. IndiQuant is set up to test that with real measurement now and real capital from the live phase.
What we believe
Four views behind how the fund is built.
Independent models
People who work apart are less likely to share the same blind spots. We combine their signals and weight them by evidence.
Out-of-sample or nothing
A model counts for what it scores on data it has never seen. Backtests and credentials don't enter into it.
A long record
One good week is noise. A model earns weight over many rounds, and it needs 20 resolved rounds before it appears on the leaderboard.
Costs are real
An edge that disappears after Indian trading costs was never an edge. Costs sit inside the portfolio optimiser.
The pipeline
From data to book in five stages.
The loop runs every NSE session, from the round's dataset to the meta-model. The weekly paper book starts trading next.
- 01
Data
Anonymised NSE features, published by 19:00 IST.
- 02
Models
One prediction per name, submitted before 09:00 IST.
- 03
Scoring
Each submission is scored once its 20-session outcome is known.
- 04
Meta-model
Signals combined, weighted by each model's record.
- 05
Book
A weekly, cost-aware portfolio.
Vision
An Indian equities fund built from the best models we can find, wherever they come from.
How we work
Rules we hold ourselves to.
They apply to the research platform today and will apply to the fund when it trades real capital.
Published rules
Contributors know how they are scored before they submit.
Evidence
Out-of-sample results settle arguments. Opinions don't.
Risk first
Limits live inside the optimiser, and execution checks them again before any trade.
Separation
Contributors see anonymised data. Stock-level signals stay inside the execution environment.
Reproducibility
Every round's dataset is fingerprinted, so any score can be traced to the data behind it.
Patience
We move to live capital when the record supports it, and not before.
Team
Who is building it.
The two people responsible for IndiQuant. Reach either of us on LinkedIn, or write to the office address on the contact page.
Diganta Sarkar
Co-Founder & CFO
LinkedIn
Bring a model. See how it scores.
If you build models, there's room for yours. Contributing is free, and your record grows with every round you submit to.