金融工程周报：期指持仓量有所分化-20260126

1. Report Industry Investment Ratings - Index futures: ☆☆☆ [1] - Treasury bond futures: ☆☆☆ [1] 2. Core Views of the Report - As of the week ending January 23, index futures showed divergence. IH2601 decreased by 1.48%, IF2601 decreased by 0.18%, IC2601 increased by 5.49%, and IM2601 increased by 4.14%. Market sentiment declined slightly, with large-cap stocks under pressure, while risk appetite for small- and mid-cap stocks remained high. The possibility of a style shift is gradually increasing as the capital structure evolves [1]. - From the high-frequency macro fundamental factor scores, for index futures, the inflation indicator scored 7 points, the liquidity indicator scored 8 points, the valuation indicator scored 12 points, and the market sentiment indicator scored 9 points. For Treasury bond futures, the inflation indicator scored 7 points, the liquidity indicator scored 9 points, and the market sentiment indicator scored 5 points [1]. - The net value of the financial derivatives quantitative CTA strategy increased by 0.29% last week, with the profit coming from shorting and closing TF on Thursday. In the long term, the supply and demand sides showed divergence. Social retail sales and real estate investment were slightly below expectations, but the decline in index futures was not significant. In the short term, high - frequency real estate sales indicators were still weak, the RMB remained strong against the US dollar, and the capital situation remained relatively loose. The short - term divergence narrowed, and the contribution weight decreased. In terms of open interest, risk appetite remained high but declined marginally compared to the previous week, with a slight decrease in IC and IM. The current overall composite signal is still above the neutral range. For Treasury bond futures, the capital situation remained loose at the beginning of the year, and the short - term trend rebounded. The stock - bond seesaw effect was still significant, and the bond market was less sensitive to fundamental feedback. TF and T showed divergence in the open - interest factor, and the composite signal was neutrally oscillating, with TF relatively stronger [1]. 3. Summary by Relevant Catalogs 3.1 Macro Fundamental Medium - and High - Frequency Factor Scores - Economic momentum: Different indicators such as blast furnace operating rate, PTA operating rate, etc., showed various weekly changes, historical quantiles, and correlations with index and Treasury bond futures. The index futures score was 7, and the Treasury bond futures score was 8 [2]. 3.2 Inflation Indicators - Various inflation - related indicators like the vegetable basket product wholesale price index, coking coal index, etc., had different weekly changes, historical quantiles, and correlations with index and Treasury bond futures. Both index futures and Treasury bond futures scored 7 [3]. 3.3 Liquidity - Liquidity - related indicators such as DR007, DR001, etc., showed different weekly changes, historical quantiles, and correlations with index and Treasury bond futures. The index futures score was 8 [4]. 3.4 Index Valuation - Index valuation indicators including PE, PS, etc., had specific weekly changes, historical quantiles, and correlations with index futures. The index futures score was 11 [5]. 3.5 Market Sentiment: Index Futures - Market sentiment indicators for index futures such as margin trading balance, northbound trading net inflow, etc., showed various weekly changes, historical quantiles, and correlations with index futures. The index futures score was 9 [6]. 3.6 Market Sentiment: Bond Futures - Market sentiment indicators for bond futures such as the 10 - year CDB bond yield, VIX, etc., showed different weekly changes, historical quantiles, and correlations with Treasury bond futures. The Treasury bond futures score was 5 [7]. 3.7 Strategy Introduction - The strategy's product pool includes index futures and Treasury bond futures. The short - term model focuses on market style, external factors, and capital - related high - frequency financial data, while the long - term model focuses on market expectations and macroeconomic data. Open interest is synthesized based on institutional long - and short - position open interest [16]. 3.8 Forecast Signals as of Last Friday - The composite signal strength is weighted by three independent models (0 - 1). The principle is to go long on the top 2 contracts with a composite signal strength of at least 0.6 and go short on the bottom 2 contracts with a composite signal strength of at most 0.4. Open - interest signals are blocked 7 days before delivery. The stop - loss point is set at a daily decline of more than 1%, and funds are allocated equally. Consecutive two - day same - direction signals are blocked [17][18]. 3.9 Last Week's Situation - From January 19 to January 23, 2026, the signals of IF, IH, IC, IM, T, and TF main contracts showed different values. TF had a - 1 signal on January 22 [19]. 3.10 Treasury Bond Futures Cross - Variety Arbitrage Strategy - The cross - variety arbitrage strategy is based on the signal resonance of the fundamental three - factor model and the trend regression model. The fundamental factor uses the instantaneous forward - rate function proposed by Nelson and Siegel, which decomposes the interest - rate term structure into three parts. The model uses PCA, factor rotation, and logistic regression. Signals are divided into three types: '1', '0', and '-1'. The trend regression model filters signals, and trading occurs when there is resonance. In practice, a 1:1.8 ratio is used for the 10 - 5Y spread adjustment [20]. 3.11 Market Quotes and Trading Signals - For TF and T main contracts, from January 19 to January 23, 2026, the N - S model signals and trend regression model signals showed different values [23].