Group 1: Report Industry Investment Rating - Not provided in the content Group 2: Core Views of the Report - The report focuses on constructing a convertible bond multi - factor system from a style dimension, aiming to establish a framework covering 115 factors and five types of style factors (valuation, momentum, volatility, liquidity, and volume - price) based on a "behavior - valuation - volatility" three - dimensional logic, and obtain excess returns while keeping the investment portfolio market - neutral through non - linear combination optimization, providing quantitative strategy support for asset allocation [1] - In the environment of low interest rates and asset shortage, the shift of funds to the "fixed income +" strategy drives the structural prosperity of the convertible bond market. The market has entered a stage of "structural differentiation - complex pricing - refined strategies", and the multi - factor system has significant applicability in the convertible bond market [2] - The style factor framework provides a path for convertible bond research. Different convertible bonds can be regarded as recombinations of style factors, and depicting convertible bonds from the style dimension helps understand market structure and rotation rules and provides a framework for constructing a multi - factor bond - selection system [3] Group 3: Summary According to Relevant Catalogs 1. Introduction - In 2025, driven by the equity market, the convertible bond market continued to strengthen, showing characteristics of active trading, stable stock, and structural differentiation. As of November 4, 2025, the average daily trading volume in the convertible bond market was about 66 billion yuan, with a high - volatility and high - central - value feature. The market had 415 convertible bonds in circulation, with a total scale of about 595.7 billion yuan. The price distribution was biased towards the medium - high price range, indicating an increase in the performance of the underlying stocks and market risk appetite [12] - Quantitative methods are more applicable in the convertible bond market. The T + 0 mechanism and high - frequency trading structure provide rich price - volume information, and the stock - bond hybrid characteristics of convertible bonds enable the multi - factor system to be applied in five dimensions: valuation, momentum, volatility, liquidity, and volume - price correlation [13] 2. Recent Expansion of the Convertible Bond Market 2.1 Convertible Bonds: "Hybrid Assets" with Both Stock and Bond Attributes - Convertible bonds can be converted into the issuer's stocks under specific conditions, with both "bond" and "stock" characteristics. Their price is composed of the pure bond value and the option value of conversion. The market has expanded rapidly, and its concentrated and active trading provides a basis for multi - factor model testing [15][16] - Compared with stocks, convertible bonds have bond - based downside protection, stock - based upside potential, medium - level volatility between stocks and bonds, and more flexible trading rules. Quantitative methods are highly applicable in the convertible bond market due to high - frequency data support, effective behavioral factors, Alpha - providing stock - bond linkage factors, and the advantage of trading systems [17][19] 2.2 Necessity of Strategies Driven by the Expansion of "Fixed Income +" under Low Interest Rates - In the environment of low interest rates and asset shortage, the shift of funds to the "fixed income +" strategy drives the prosperity of the convertible bond market, creating a situation of strong demand, tight supply, and a rising pricing center, which provides a long - term foundation for quantitative and systematic strategies [18] - As of Q3 2025, the scale of public funds has increased steadily, with a pattern of "expansion of equity products and contraction of bond funds". The "fixed income +" products, especially secondary bond funds, have expanded significantly. The demand for convertible bond allocation has increased, while the supply has slowed down. The market has formed a pattern of "high valuation - high position - low supply", and convertible bonds have shown stronger resilience in the volatile market [20][21][23] 3. Convertible Bonds and Equities from the Perspective of Style Factors 3.1 Style Factors: Systematic Depiction of the Equity Market from the Barra System - Style factors are core dimensions for depicting the common characteristics and systematic differences of assets in the multi - factor model system. The Barra model decomposes asset returns into style factor returns and idiosyncratic returns, and in the Barra framework, style factors in the equity market include valuation, growth, momentum, volatility, scale, leverage, and liquidity, which jointly form the "style map" of the equity market and provide a path for convertible bond research [28][29][32] 3.2 Style - Based Structure of the Convertible Bond Market: Division into Stock - Oriented, Balanced, and Bond - Oriented Types - Convertible bonds can be divided into stock - oriented, balanced, and bond - oriented types based on style factors. Stock - oriented convertible bonds are dominated by stock characteristics, with high elasticity and large fluctuations; balanced convertible bonds have a balanced risk - return profile, with both stock and bond features; bond - oriented convertible bonds are dominated by bond characteristics, with strong defensive properties. This division provides a basis for factor stratification and strategy construction [33] 3.3 The Stock - Dominant Nature of the Convertible Bond Market under the Slow - Bull Expectation - The convertible bond market has shifted from being bond - dominated to stock - dominant. The high correlation between the convertible bond index and the CSI 1000 and CSI 2000 indices indicates that the market is currently in a stock - driven stage. The reasons include the increase in the concentration of high - priced convertible bonds, the change in the capital structure, and the support of the macro - liquidity and interest - rate environment [35][36][37] 3.4 Introduction to the Multi - Factor Convertible Bond System: From Five Style Factors to the Systematic Back - Testing Framework - A multi - dimensional system covering 115 daily - frequency factors is constructed based on the price - volume characteristics and clause structure of the convertible bond market, including valuation, momentum, volatility, liquidity, and volume - price correlation factors. These factors form a relatively complete convertible bond quantitative framework [41][42] - Daily - frequency data is chosen as the core sample dimension for constructing the convertible bond multi - factor system. It can capture short - term market changes, maintain signal effectiveness, and balance signal sensitivity and execution feasibility [44][45] 4. Convertible Bond Multi - Factor System and Back - Testing Results 4.1 Historical Performance of Five Types of Style Factors - Based on the back - testing results from 2021 to 2025, the five types of style factors can be divided into three categories: the leading group includes momentum and volatility factors with high annualized excess returns; the stable group includes the liquidity factor; the medium group includes the five - factor equally - weighted composite factor, valuation factor, and volume - price correlation factor [47] - The excellent performance of the momentum factor is due to its ability to capture the "trend effect" in the convertible bond market. The volatility factor has high risk - adjusted returns and good risk control, which may be related to risk - pricing compensation and avoiding the "volatility trap" [48] 4.2 Portfolio Optimization Logic - Single - factor investment in convertible bonds has shortcomings such as high return volatility, insufficient factor synergy, significant trading - cost erosion, and style - deviation risk. A non - linear optimization framework is used for portfolio construction, with the goal of maximizing risk - adjusted returns under multiple constraints such as market value, industry, style, and individual bond weights [51][53][54] - Back - testing results show that the liquidity factor performs best under market neutrality since 2021, followed by volume - price and momentum factors. After optimization, the excess returns of most style factors decline significantly, indicating that high returns in the convertible bond market often come from style deviation and high turnover [56] 4.3 Follow - up Optimization Logic - The follow - up optimization should change the way of synthesizing large - category factors from "equally - weighted synthesis" to "weighted synthesis based on historical performance". Specific methods include weighted synthesis based on risk indicators, weighted synthesis based on return indicators, and direct optimization by eliminating ineffective or redundant sub - factors [58][59] 5. Follow - up Strategy Optimization 5.1 Event - Driven: Seizing the Certainty Opportunities in Clause Games - The event - driven strategy uses issuers' active actions such as downward - revision of conversion prices and share repurchases to obtain excess returns. It is necessary to establish a systematic event database and real - time monitoring mechanism [60][61][62] 5.2 Mispricing: Exploiting the Cognitive Bias of Option Value - The mispricing strategy is based on the market's mis - evaluation of the option value of convertible bonds. It involves constructing a theoretical value model, identifying pricing deviations, and constructing a market - neutral portfolio to earn value - regression returns [63]