Our own systematic trading research

What we research

Systematic, rules-based strategies, positional and intraday, across:

• US indices: the S&P 500 and the Dow Jones
• Commodities: gold, silver and other commodities
• Indian indices: the Nifty 50 and Bank Nifty
• Stocks in the US and India

The research draws on 12 to 25 years of price history, depending on the market.

How we test a strategy

A backtest is easy to flatter. These are the checks a strategy has to pass before we take it further:

• In-sample and out-of-sample. The rules are built on one part of the history and then tested on a part they have never seen.
• Walk-forward testing. The test is repeated as the window moves forward through time, so a strategy has to hold up period after period, not once.
• Costs and slippage. Every simulated trade pays commission and an allowance for slippage, so results are never worked out as if trading were free.
• Robustness checks. A result has to survive changes to its settings and to the period tested, rather than rest on one lucky choice.

Our published in-house studies show parts of this method in practice, for example look-ahead-free signal detection, deliberately pessimistic fills and walk-forward validation in our linear regression study.

Why there are no figures on this site

Past results, backtested or live, are easy to misread and easy to misuse. In India, SEBI has set up a framework (PaRRVA) for verifying past risk and return figures before they are published. Until our figures are verified that way, we publish none, here or anywhere else. We keep a live track record on a separate account so that results can be checked later. Qualified institutions can see the detail under NDA.

Who leads the research

Ankur Parikh, our co-founder, leads this research. He co-founded Trade Vectors in 2012, leads its engineering, and has trained 400+ professionals in algorithmic trading in Singapore and India. More about him on our founders page.

Kept apart from client work

Client systems are built to each client’s own rules. They never carry our research logic, and we never say a client uses our strategies. If you have a strategy of your own, our custom trading software service works as it always has: you bring the strategy, and we build the software for it.

Working with institutions

Brokers, portfolio and asset managers, family offices and prop desks can discuss licensing, white label arrangements for brokers and research partnerships with us, under NDA. We do not offer our strategies to individual traders or investors.

See how partnerships work

Our published in-house studies

• Automating a 0DTE index options strategy on Interactive Brokers: live data, strike selection and two leg execution, built for live trading.
• Backtesting a linear regression support strategy across markets: look-ahead-free detection, deliberately pessimistic fills and walk-forward validation.
• Turning chart patterns into tested rules: six classic chart patterns written as exact rules, with no look-ahead.
• Why we test trend strategies on the index first: gaps, outliers, sideways markets and the bar worth measuring against.
• A multi-market equity backtesting and research platform: Indian and US equities from one codebase, with per-sector risk caps.