QUANT / LAB
Choose a market to explore prediction ideas and trading rules. Each study has its own dates, sample and limitations.
Factor research asks whether a signal helps predict returns. Strategy research also considers positions, turnover and costs. Historical simulations do not guarantee future returns.
Test predictions and portfolio rules on historical stock data.
Factors and prediction · Evidence status：Exploratory
Prediction comparisons across two windows have not produced stable rankings.
Historical prediction research with limited samples and data-quality evidence.
Explore research →Strategies and backtests · Evidence status：Exploratory
Keeping stocks whose rankings slipped usually reduced trading; returns improved less often.
Historical account simulations; real execution remains unverified.
Explore research →Strategies and backtests · Evidence status：Exploratory
Compare CNN, momentum and equal-weight holding across costs and turnover.
Exploratory results; fixed membership, dividends and risk matching remain limitations.
Explore research →Explore fixed-universe strategy history and earlier CNN chart experiments.
Factors and prediction · Evidence status：Further tests needed
Momentum and moving-average strategies were compared, but independent signal-ranking and grouped-return tests are still needed.
A complete independent factor test is not yet available.
Explore research →Strategies and backtests · Evidence status：Exploratory
Historical comparisons cover momentum and rebalancing frequency; historical membership and point-in-time data have gaps.
The limited sample does not represent the full Hong Kong market.
Explore research →Strategies and backtests · Evidence status：Historical record, verification pending
See original images, six related entries and historical simulated returns. Training boundaries remain unverified.
Explore research →Study price signals, portfolio returns and trading costs.
Factors and prediction · Evidence status：Further tests needed
Price-signal research is available; decentralized-exchange factors are still being prepared.
Data and independent testing remain incomplete.
Explore research →Strategies and backtests · Evidence status：Exploratory
Historical simulations separate market, exposure, selection and cost contributions.
Results depend on historical windows and modeled costs.
Explore research →Strategies and backtests · Evidence status：Exploratory
See how turnover, costs and volatility controls affected eight historical windows.
Exploratory results; sentiment, asset events and actual fills remain unverified.
Explore research →