Home / Research / Research Behind Our Signals
Research Brief

The Research Behind
Our Signals

Most "signals" rest on folklore that breaks in real markets. TradePulse grounds each signal in peer-reviewed research, adapts it to the instrument, and validates it continuously. Here's how.

A condensed summary of TradePulse's internal signal-validation research.

Folklore breaks in real markets

"High PCR is bullish." "RSI below 30 means buy." "Negative dealer delta is bullish support." Each of these is a half-truth that inverts in a trending market. A signal engine that applies them naively will issue buy calls into a falling market — the exact failure mode this research was built to eliminate.

Example: PCR is conditional on IV

A high put-call ratio only means put writing (bullish) when put implied volatility is falling. When put IV is rising alongside PCR, the dominant flow is directional put buying — which is bearish. Reading PCR without the IV context is how naive systems get the day backwards.

Put writing → Bullish Put buying → Bearish Call writing → bearish Call buying → bullish High PCR Low PCR Put IV falling Put IV rising
The same PCR reading flips meaning with the IV trend — TradePulse scores PCR conditionally on implied volatility, as the microstructure literature prescribes.

Regime-aware, not one-size-fits-all

RSI below 30 confirms strength in a downtrend rather than signalling a bounce; max pain only exerts gravity in the final expiry window; dealer delta/gamma flips meaning outside the OI walls. TradePulse interprets each signal in the context of the prevailing regime instead of applying a fixed rule.

Adaptive thresholds, not hardcoded numbers

A "trend established" threshold that fits NIFTY under normal volatility is wrong for crude oil or for a quiet session. TradePulse normalises thresholds to each instrument's own historical volatility distribution, so the same logic behaves correctly across NIFTY, Bank Nifty and high-IV commodities.

Grounded in peer-reviewed research

The signal interpretations draw on established work in options microstructure and quantitative finance — including Pan & Poteshman (2006) and Ni, Pan & Poteshman (2008) on the information in option volume and IV, gamma-exposure research on hedging-driven momentum, and Hidden-Markov / Kalman methods for regime classification. Where the live data available on NSE differs from the academic setup (e.g. open interest vs buy-to-open volume), the limitation is documented and weighted accordingly.

Validated continuously

Signal formulas are reviewed against the literature and corrected when they underperform — this brief itself summarises one such review. Rigour isn't a one-time claim; it's an ongoing process.

See the signals in action

TradePulse's AI signals apply this research to live option-chain data — with the reasoning shown, not hidden.

Keep reading