金工策略周报-20260208

1. Report Industry Investment Rating No relevant information provided. 2. Core Viewpoints of the Report - Last week, all Treasury bond futures contracts closed higher, except for the two - year - term contract which closed down 0.05%. The 30 - year, ten - year, and five - year contracts rose 0.58%, 0.11%, and 0.05% respectively. The market risk preference weakened, activating the hedging property of Treasury bond futures. The downward trend of Treasury bond futures is not likely to reverse when the long - term bull market logic of the stock market remains unchanged and the coupon yield of Treasury bonds is not attractive. Only when the expected return of equity or risk assets declines marginally, the short - term hedging trading property of the bond market becomes more obvious [5]. - Last week, the domestic commodity market generally declined, with silver having the largest decline. Gold and silver dropped significantly due to external hedging and cooling speculation. The term structure and price - volume trend factors, which performed well before, were weak last week. The basis factors rose by more than 1.5% on average, the warrant factors closed slightly flat, and other factors generally declined by more than 2%. The volatility of commodity factor returns is increasing, and investors are advised to focus on commodity factors with long - term expected return capabilities [16][19][20]. 3. Summary by Relevant Catalogs 3.1 Treasury Bond Futures Quantitative Strategy - Market Review: The 30 - year, ten - year, and five - year Treasury bond futures contracts closed higher last week, while the two - year contract closed down. The basis of each variety was differentiated. The CTD bond of the ten - year bond was 250018, with a basis of about 0.05 yuan on the 6th, which was the same as the historical average. The CTD bond of the 30 - year bond was 210005, with a basis of 0.19 yuan on the 6th, higher than the historical average [5]. - Daily - frequency Timing Strategy for Treasury Bond Futures: For the ten - year Treasury bond, from 2021/01/01 to the present, the annualized return, Sharpe ratio, and maximum drawdown of the single - leverage portfolio were 2.45%, 1.16, and 1.89% respectively. Since the report was released (2025/11/01 to the present), the corresponding figures were 4.24%, 2.95, and 0.59% respectively [5]. - Single - side Strategy Performance: The construction method can be found in "Sorting and Optimization of Daily - frequency Timing Factors for Treasury Bond Futures". The large - category factors include basis, daily technical, intraday price - volume, high - frequency capital flow, member positions, and risk assets. The signals are generated by equal - weighting and averaging within large - category factors, and the positive or negative sign of the average is used as the long - short signal. The back - testing details are that the strategy uses the VWAP of the first ten minutes of the next day's opening as the trading price and buys with single leverage [11][14]. 3.2 Commodity CTA Factor and Strategy Performance - Commodity Factor Performance: Last week, the domestic commodity market generally declined, with silver having the largest decline. The term structure and price - volume trend factors, which performed well before, were weak, while the basis factors rose by more than 1.5% on average, the warrant factors closed slightly flat, and other factors generally declined by more than 2%. The volatility of commodity factor returns is increasing [16][19][20]. - Tracking Strategy Performance: - CWFT strategy: Annualized return of 9.2%, Sharpe ratio of 1.58, Calmar of 1.04, maximum drawdown of - 8.81%, recent one - week return of - 0.92%, and year - to - date return of 0.56% [17]. - C_frontnext & Short Trend strategy: Annualized return of 11.1%, Sharpe ratio of 1.70, Calmar of 1.65, maximum drawdown of - 6.72%, recent one - week return of - 0.47%, and year - to - date return of - 0.17% [17]. - Long CWFT & Short CWFT strategy: Annualized return of 12.0%, Sharpe ratio of 1.35, Calmar of 0.92, maximum drawdown of - 13.07%, recent one - week return of - 1.58%, and year - to - date return of 0.73% [17]. - CS XGBoost strategy: Annualized return of 5.4%, Sharpe ratio of 0.88, Calmar of 0.26, maximum drawdown of - 21.19%, recent one - week return of 2.60%, and year - to - date return of - 2.86% [17]. - RuleBased TS Sharp - combine strategy: Annualized return of 11.2%, Sharpe ratio of 1.47, Calmar of 1.36, maximum drawdown of - 8.26%, recent one - week return of - 1.64%, and year - to - date return of - 2.20% [17]. - RuleBased TS XGB - combine strategy: Annualized return of 11.3%, Sharpe ratio of 1.97, Calmar of 2.52, maximum drawdown of - 4.49%, recent one - week return of 1.24%, and year - to - date return of - 1.88% [17]. - CS strategies, EW combine strategy: Annualized return of 12.5%, Sharpe ratio of 1.78, Calmar of 1.69, maximum drawdown of - 7.38%, recent one - week return of - 0.72%, and year - to - date return of 0.82% [17]. - Strategy Details: - CWFT combination: A simple composite combination of Carry, Warrant, Futurespot, and Trend factors, with equal weights within the same large category and a 5:2:2:1 weight for different large categories [23]. - C_frontnext & Short Trend combination: It aims to hedge the negative returns of short - term price fluctuations without changing the position direction of long - term spread factors, with C_frontnext as the main factor and Short Trend as the secondary factor [23]. - Long CWFT & Short CWFT combination: The four types of C, W, Fs, and T factors are equally weighted and compounded within the category and then compounded with a 5:2:2:1 ratio, with the long - cycle CWFT factor as the main factor and the short - cycle CWFT factor as the secondary factor [23]. - CS XGB combination: A time - series factor XGB combination based on all market varieties, trained with samples from 20091231 - 20191231 [23]. - RuleBased TS Sharp - weighted combination: A time - series factor strategy based on rule - based long - short signals, obtained by overlaying rules on the factor library and coarse - grained window parameters, and then a Sharpe - weighted combination is obtained after in - sample screening [23]. - RuleBased TS XGB combination: A time - series factor strategy based on rule - based long - short signals, obtained by overlaying rules on the factor library and coarse - grained window parameters, and then an XGBoost combination is obtained after in - sample screening [23]. - Strategy Performance Comparison: Among the six strategies, CS XGBoost performed the best last week with a return of 2.60%, and Long CWFT & Short CWFT performed the best year - to - date with a return of 0.73%. The equal - weighted composite strategy of the above cross - sectional strategies has an annualized return of 12.5%, a Sharpe ratio of 1.78, a Calmar of 1.69, a maximum drawdown of - 7.38%, a recent one - week return of - 0.72%, and a year - to - date return of 0.82% [40].