About
A partnered research organisation
A partnered research, education and applied quantitative-finance organisation focused on Indian index options, stock options and equities, market microstructure and computational finance.
What this organisation is
Precision Capital Academy LLP is a partnered organisation combining market research, financial-market education and applied computational work across Indian index options, stock options and equities. It exists because a set of questions about how those markets actually behave turned out to be answerable — but only with infrastructure that did not exist yet.
The research arm asks empirical questions and answers them with data, code and a stated method. The education arm explains what those questions mean to people participating in the same markets. The applied work is the software that makes the first possible and the second concrete.
What it is not
It is not affiliated with any university, exchange, broker, asset manager or regulator, and makes no claim to be. It does not sell signals, manage money for anyone else or offer personalised investment advice. Published research is research: it describes what was found, including when what was found was nothing.
Where machine learning fits
Machine learning is one of the methods here, not a separate business. It is used where a question genuinely needs it — feature construction on event-aligned data, classification, forecasting — and held to the same standard as everything else: a model that cannot survive a period it has never seen is a description of the past, not a finding.
The organisation does not present itself as a general machine-learning consultancy. It researches markets, and uses whichever methods that requires.
Why the research comes first
Most market education is downstream of someone's convictions. The ordering here is deliberately the other way round: the research decides what is worth teaching, and a result that fails out of sample changes what gets said — publicly, in the paper and in what is taught.
Co-founder
Subhash Sadacharam
Co-founder & Market Researcher
Market researcher working on Indian index options, stock options and equities — derivatives, market microstructure, systematic strategies and applied machine learning.
The method is the one this site describes: propose a signal, specify the test before looking, control the obvious confound, and hold a period back for a single honest evaluation. Findings that do not replicate get published as findings that do not replicate.
Before Precision Capital
An engineer who learned to model the real world before he modelled markets.
Before co-founding Precision Capital, Subhash spent nearly five years leading operations and data analytics at Sri Sadacharam Farms — forecasting demand, planning procurement and pricing, and turning noisy, incomplete data from dairy, agriculture and B2B distribution into decisions that had to work on the ground. It taught him the discipline this work depends on: a model is only as good as the constraints it survives. Trained as an electrical engineer at PSG Institute of Technology and Applied Research, with early internships at Bharat Electronics and Real AI, he went on to earn QuantInsti's EPAT with a Certificate of Excellence and an M.Tech in AI & ML from BITS Pilani.
- M.Tech (WILP), Artificial Intelligence & Machine Learning — BITS Pilani
- EPAT, Certificate of Excellence — QuantInsti
- B.E., Electrical & Electronics Engineering — PSG Institute of Technology and Applied Research
- NISM Series VIII — Equity Derivatives
Research interests
- Index options
- Stock options
- Equities
- Options & volatility
- Market microstructure
- Options price discovery
- Quantitative modelling
- Applied machine learning
- Time-series modelling
- Systematic strategies
- Financial data engineering
- Real-time market analytics
Collaboration
Enquiries about research collaboration, data access, replication or academic correspondence are read personally.
Timeline
How the work developed
Only work that has actually been done. Each entry describes what was built, not what it proved.
2025
Market-structure engine
A mechanical break-of-structure and change-of-character detector built in two parallel implementations — a TradingView Pine Script v6 indicator for visual inspection and a Python implementation for research — so that structure could be defined precisely enough to test.
2025
Options-chain archive and research pipeline
Weekly NIFTY 50 options-chain data assembled into a research archive, and a deterministic Python pipeline built to reconstruct positioning around detected events.
2025
In-sample study
A full calendar year — all four quarters, 55 weekly expiries — used to specify and estimate the association between open-interest repositioning and structural break direction.
2026
Out-of-sample validation
A held-out later period evaluated once. The reactive finding replicated; the predictive specification decayed to approximately chance and was discarded rather than re-tuned.
2026
Working paper and reproduction notebook
The study written up as a working paper through successive review rounds, accompanied by a consolidated notebook that reproduces every headline figure end to end from the source data.