What is Quantitative Investing (量化交易)?
Quant investing uses data-driven methods combined with statistical modeling techniques plus systematic backtesting frameworks to construct algorithmic portfolios. Unlike traditional active funds relying on fundamentals only or momentum-based discretionary rules alone which often over-expose during regime shifts, our multi-asset system combines three complementary strategies: convertible bonds for capital preservation via embedded options + downside protection; broad-based ETFs providing market beta exposure with low-cost rebalancing benefits; and long-only A-stock positions driven by factor research across value, momentum,,quality,and fundamental screening dimensions.
Why Start Here?
Most retail investors lose money due to emotional trading patterns during drawdown periods (holding onto stocks too long hoping for reversal), chasing yield without understanding underlying risks in convertible bonds especially those approaching maturity dates where credit spreads tighten unevenly depending on issuer fundamentals or liquidity conditions vary significantly across different sectors. This guide demystifies quantitative techniques accessible while avoiding over-complication:
- Machine learning: Use RandomForest/GradientBoosting models not GANs requiring massive GPU clusters, which can run locally via scikit-learn without cloud dependencies
Feature engineering transform raw price/volume/open interest data into lagged returns,Rolling windows volatility,,bidask_spread_width indicators predictive of short-term reversals before manual technical analysis catches them.
Core Strategies Covered
1. Convertible Bond Quantization (可转债量化)——Use implied yield curves combined with option pricing formulas calibrated to historical exercise ratios for early warning signals about conversion events triggered by stock price breaches triggering mandatory redemption clauses or credit deterioration leading downgrades causing mark-to-market losses that portfolio rebalancing cannot offset without immediate action taken at the right moment when issuer fundamentals deteriorate beyond acceptable thresholds.
Backtested Performance — Net Asset Value of ~102,764 RMB with maximum drawdown under normal market conditions not extreme black swans (less than -8%) and annualized return exceeding risk-free rate by about 135 basis points after all expenses including transaction costs and margin financing interest rates when applicable scenarios tested over rolling 6-month windows across different economic cycles spanning bull markets, bear phases, stagflationary environments where credit spreads widen asymmetrically depending on macro policy direction changing from tightening monetary conditions to easy money regimes shifting liquidity preferences among investors chasing yield versus capital preservation goals which dictate portfolio allocation decisions dynamically based current forward-looking risk metrics calculated daily by quantitative models monitoring exposure concentrations across issuer ratings sectors duration convexity adjustments made necessary when market volatility increases suddenly catching traders off guard before stop losses execute automatically as intended.
Risk Management (风险控制):
Hedge equity beta using index futures or treasury bond ETFs depending on regime state classification based VIX term structure slope flattening indicating flight to safety versus steepener suggesting risk-on mood prevailing in markets right now influencing portfolio positioning decisions made daily by quantitative frameworks updating model weights adaptively without manual intervention required every single trading day becoming obsolete before implementation completed during fast-moving periods when market makers exploiting stale quotes from slow institutional traders unable adjust positions quickly enough causing temporary dislocations favoring high-frequency arbitrage strategies using co-invested liquid instruments reducing overall portfolio drawdowns through diversification benefits not captured by traditional correlation matrices failing to account for regime-switch contagion effects spreading across asset classes simultaneously during stress periods when liquidity dries up everywhere at once making deleveraging costly or impossible without triggering fire-sale cascades amplifying losses beyond initial expectations formed from calm market conditions months ago before crisis erupted suddenly catching everyone off guard unprepared due lack of contingency plans tested rigorously via Monte Carlo simulations modeling thousands possible outcomes scenarios ranging worst case (95th percentile) to best fit historical distributions showing downside protection holding even during severe drawdowns when correlations converge to unity temporarily neutralizing diversification benefits once relied upon normally in stable regimes.