Our performance · 04 / 04
From ideas to impact.
Every strategy begins as a mathematical question rather than a market opinion, and reaches investors only after it has been coded, tested and run in paper mode on the same platform.
Figures as at 21 September 2026
How the platform works
- 01
Signal generation
Live market data in, a directional signal out — computed in code, on a fixed schedule.
- 02
Macro gates
A signal trades only when it agrees with the prevailing market regime.
- 03
Execution and exit
Entries, targets, stops and square-off are placed programmatically on pre-set rules.
- 04
Logging and roll-out
Every signal is logged before an order is sent; new strategies run in paper mode first.
Research foundations
Statistical learning
Regularised regression and sequence models map engineered features onto the distribution of next-period returns. Regularisation penalises complexity, so a fit survives outside the window it was trained on.
Mean reversion and relative value
Long-run statistical relationships between instruments, where deviations are measurable in standard deviations rather than opinions and carry little directional exposure.
Derivatives and volatility
Option strategies built on premium decay and the shape of the volatility surface, sized so the payoff profile is the source of return.
Market microstructure
Order-book depth and open-interest movement, treated as probabilistic signals weighted against the prevailing regime.
Continuous monitoring
- Live strategies
- 4 of 8
- Capital
- ₹11.38 Cr
- Investors
- 120
- Accounts closed
- 0
- CanopyIndex options · fully automated
- Backtest
- Paper
- Live
- HeartwoodShort index options · NIFTY / SENSEX / MIDCPNIFTY
- Backtest
- Paper
- Live
- AmberBidirectional LSTM · GOLDBEES
- Backtest
- Paper
- Live
- GroveMarket-neutral statistical arbitrage
- Backtest
- Paper
- Live
- Twin OakShort strangle · NIFTY / SENSEX options
- Backtest
- Paper
- Live
- WillowRidge regression · multi-pair FX
- Backtest
- Paper
- Live
- CedarBid-ask flow · OI momentum · Midcap 150 BTST
- Backtest
- Paper
- Live
- BirchOHLC candle patterns · OI & PCR regime
- Backtest
- Paper
- Live
What we do not publish
Model parameters, entry thresholds, instrument universes and execution timings stay inside the firm; they are the research itself. What we publish is the outcome: every strategy's backtest and live record on the same axes, including the months that went against us.
Way forward
Assets under management (₹ Cr)
View data table
| Assets under management | |
|---|---|
| Today | ₹11.38 Cr |
| Target, September 2027 | ₹40–60 Cr |
Investors
View data table
| Investors | |
|---|---|
| Today | 120 |
| Target, September 2027 | 300+ |
How we intend to get there
Broaden the base. Around 180 further investors over the year, at ticket sizes consistent with the existing book rather than through a single large mandate.
Move the remaining strategies live. Twin Oak, Willow, Cedar and Birch complete paper trading and join the live record.
Hold the discipline. Live results continue to track research, and every month, good or bad, continues to be published.
Targets set by the partners for the twelve months to September 2027. Targets are objectives, not forecasts or guarantees.
Ideas in the open
Notes from the field.
The value of a better question.
Why the most useful research often begins by looking at a familiar problem differently.
Illustrative articleGood research should be repeatable.
A practical look at the systems that make experiments easier to trust, revisit, and improve.
Illustrative articleLearning from uncertainty.
What changes when uncertainty becomes part of the question, rather than an afterthought.
Illustrative articleRooted in curiosity.
Driven by evidence.
The partners will walk any prospective investor through the strategy records line by line, including the periods that did not work.