Research Results — Shaurya Singh

Core factor research

14 completed tests.

Hypothesis	What it tested	Outcome
Turnover-conditioned momentum	Momentum stronger in high-turnover names	Inconclusive
Volume shock	Price shocks + volume predict reversal	Statistically significant, failed robustness scrutiny — see case study
Lottery/MAX effect	Large recent gains predict underperformance	No signal
Cross-sectional momentum	Past-year winners keep winning	No signal
Volatility–return relationship	Realized volatility predicts returns	No signal
Overnight-return continuation	Overnight price moves continue intraday	No signal
Residual momentum	Stock-specific (not market-wide) momentum	No signal
Realized skewness	Return skewness predicts future returns	No signal
Moving-average distance	Distance from moving average predicts reversion	Wrong-signed
Volume-conditioned price-shock reversal	Price shocks reverse, conditional on volume	No signal (both directions tested)
Moving-average crossover (5,50)	Classic technical crossover signal	No signal
Short-term reversal	Recent losers/winners reverse	No signal
Regime-gated momentum	Momentum varies by market regime	Inconclusive
Low-beta / betting-against-beta	Low-volatility stocks outperform, risk-adjusted	Wrong-signed

Strategy transfer experiments

3 externally-sourced strategies. A separate track testing whether specific published strategies, not just individual factors, transfer to Indian markets at this universe size.

Strategy	Outcome	Why
Short-term reversal, adapted	No transfer	Result was wrong-signed (continuation, not reversal) and did not survive correction for testing multiple candidates at once
Turnover-conditioned double-sort	Implementation failure	The strategy's own sort mechanically collapsed to a single stock per rebalance at this universe size — not a valid test of the underlying idea as designed
Sector momentum rotation	No transfer	Wrong-signed result on a clean, broad-based test, but small sample (7 rebalances) limits confidence either way