标题内容1

OLS Regression Analysis Report#

Model Overview| Metric | Value ||--------|-------|| **Dependent Variable** | Dretwd (Daily Returns) || **Model Type** | OLS Regression || **Number of Observations** | 94,389 || **R-squared** | 0.4811 (48.11%) || **Adjusted R-squared** | 0.4810 (48.10%) || **F-statistic** | 4605.83 || **F-statistic p-value** | 0.0000e+00 *** || **AIC** | -510857.31 || **BIC** | -510668.20 |#

Core Variables and Their Significance##

? Extremely Significant Variables (p < 0.001)| Variable | Coefficient | p-value | Interpretation ||----------|-------------|---------|----------------|| **Turnover** | **+0.003320** | < 0.001*** | Higher turnover rate leads to higher daily returns ✅ || **Market_Ret** | **+1.091299** | < 0.001*** | Market returns have strong positive impact on stock returns ✅ || **Reversal** | **-0.019095** | < 0.001*** | Reversal effect - lower returns after short-term decline ❌ || **Beta** | **-0.004430** | < 0.001*** | Higher Beta stocks have lower returns (risk paradox) ❌ || **Momentum** | **-0.003784** | < 0.001*** | Negative momentum effect - contrary to traditional theory ❌ || **BM** | **-0.001345** | < 0.001*** | Higher book-to-market ratio leads to lower returns (value paradox) ❌ || **Illiquidity** | **+0.056956** | < 0.001*** | Higher illiquidity leads to higher returns (liquidity premium) ✅ || **TED_Spread** | **-0.002166** | < 0.001*** | Widening TED spread leads to lower returns (risk aversion) ❌ || **PB** | **-0.000129** | < 0.001*** | Higher P/B ratio leads to lower returns ❌ || **bond_10y** | **+0.000696** | < 0.001*** | Rising 10-year Treasury yields lead to higher stock returns ✅ || **usd_cny** | **+0.001160** | < 0.001*** | RMB appreciation leads to higher returns ✅ |##

⭐ Significant Variables (p < 0.01)| Variable | Coefficient | p-value | Interpretation ||----------|-------------|---------|----------------|| **Mkt_Vol** | **-0.002025** | 0.0025** | Market volatility rises, returns fall ❌ |##

✓ Marginally Significant Variables (p < 0.05)| Variable | Coefficient | p-value | Interpretation ||----------|-------------|---------|----------------|| **Term_Spread** | **+0.000435** | 0.0327* | Term spread increase leads to higher returns ✅ |##

❌ Insignificant Variables (p > 0.05)| Variable | p-value | Interpretation ||----------|---------|----------------|| const | 1.35e-01 | Constant term not significant || PE | 1.85e-01 | P/E ratio not significant || Oil_Ret | 2.55e-01 | Oil price returns not significant || Market_VaR | 5.62e-01 | Market Value at Risk not significant || bond_3m | 1.16e-01 | 3-month Treasury yield not significant || bond_1y | 2.49e-01 | 1-year Treasury yield not significant || Delta_Bond_3m | 5.62e-02 | Change in 3-month bond yield not significant || EPU | 7.07e-02 | Economic Policy Uncertainty only marginally significant |#

Significance Summary| Significance Level | Variable Count | Percentage ||-------------------|----------------|------------|| *** (p < 0.001) | 11 | 52.4% || ** (p < 0.01) | 1 | 4.8% || * (p < 0.05) | 1 | 4.8% || . (p < 0.1) | 2 | 9.5% || Not Significant | 6 | 28.6% |#

Key Findings##

1. Excellent Model Fit- **R² = 48.11%** - Control variables explain nearly half of the return variance- **F-statistic = 4605.83 (p < 0.001)** - Model is highly statistically significant- Large sample size (94,389 observations) provides robust statistical power##

2. Strongest Market Effects- **Market_Ret** has the strongest impact with coefficient = **1.09**- This confirms the strong market beta relationship in Chinese stock market##

3. Liquidity Premium Verification- **Illiquidity** is significantly positive (coefficient = **+0.057**, p < 0.001)- This validates the liquidity premium theory in emerging markets##

4. Traditional Finance Theory Paradoxes###

? Risk Paradox (Beta)- **Expected**: Higher Beta → Higher returns (Risk-return tradeoff)- **Observed**: Beta is **negative** and significant (coefficient = **-0.0044**, p < 0.001)- **Interpretation**: In Chinese market, high-risk stocks underperform, possibly due to speculative bubbles and retail investor behavior###

? Momentum Paradox- **Expected**: Past winners continue to win (Positive momentum)- **Observed**: Momentum is **negative** and significant (coefficient = **-0.0038**, p < 0.001)- **Interpretation**: Contrary to international evidence, Chinese stocks exhibit reversal patterns, suggesting high-frequency trading and mean reversion###

? Value Paradox (Book-to-Market)- **Expected**: High BM (value stocks) → Higher returns- **Observed**: BM is **negative** and significant (coefficient = **-0.0013**, p < 0.001)- **Interpretation**: Growth stocks outperform value stocks in Chinese market, reflecting preference for growth narratives and state-owned enterprise characteristics###

? Valuation Paradox (P/B Ratio)- **Expected**: Lower P/B → Higher returns- **Observed**: PB is **negative** and significant (coefficient = **-0.00013**, p < 0.001)- **Interpretation**: Market rewards growth valuation over traditional value metrics##

5. Exchange Rate Effects- **usd_cny** (USD/CNY rate) is significantly positive (coefficient = **+0.0012**, p < 0.001)- **RMB appreciation** (higher USD/CNY) leads to higher returns- This suggests capital inflows and positive sentiment during currency appreciation##

6. Macroeconomic Risk Indicators###

TED Spread (Credit Risk)- **TED_Spread** is significantly **negative** (coefficient = **-0.0022**, p < 0.001)- Widening TED spread (increasing credit risk) leads to lower returns- Reflects risk aversion periods and global contagion effects###

Term Structure- **Term_Spread** is positively significant (coefficient = **+0.00043**, p < 0.05)- Steepening yield curve (higher term spread) leads to higher returns- Consistent with economic growth expectations###

Bond Yields- **10-year Treasury yield** is significantly positive (coefficient = **+0.00070**, p < 0.001)- Rising long-term rates lead to higher stock returns- Suggests expectations of economic growth rather than pure discount rate effects##

7. Trading Activity Effects###

Turnover- **Turnover** is strongly **positive** (coefficient = **+0.0033**, p < 0.001)- Higher trading volume leads to higher returns- Reflects investor interest and market efficiency###

Reversal- **Reversal** is strongly **negative** (coefficient = **-0.019**, p < 0.001)- Strong mean reversion effect- Suggests market overreaction and correction mechanisms##

8. Market Volatility Impact- **Mkt_Vol** is significantly **negative** (coefficient = **-0.0020**, p < 0.01)- Higher market volatility leads to lower returns- Consistent with volatility risk premium theory#

Implications for Investment Strategy##

What Works in Chinese Market:- ✅ Follow market trends (Market_Ret = 1.09)- ✅ Target illiquid stocks (Liquidity premium)- ✅ Monitor exchange rate movements (RMB appreciation)- ✅ Focus on high-turnover stocks (Active trading)##

What Doesn't Work (Contrary to Conventional Wisdom):- ❌ Don't rely on beta for return prediction (Risk paradox)- ❌ Don't expect momentum to persist (Reversal dominates)- ❌ Don't blindly invest in value stocks (Growth outperforms)- ❌ Don't assume higher volatility always means higher returns#

Limitations and Further Research##

Current Limitations:- Sample limited to 35 stocks (may not be representative)- Cross-sectional analysis without robustness checks- No consideration of time-varying coefficients- Missing values in key variables (Market_VaR: 21.37%)##

Recommended Further Analysis:- Expand sample to broader market indices- Test for structural breaks and regime changes- Include interaction effects- Perform robustness checks with alternative model specifications- Analyze sub-periods to identify regime-specific effects---**Note**: This analysis is based on data from 2011-01-04 to 2022-03-31. Significance levels: *** p < 0.001, ** p < 0.01, * p < 0.05, † p < 0.1