Workflow
金融衍生品量化CTA策略
icon
Search documents
期债长周期边际下降
Guo Tou Qi Huo· 2025-08-11 14:31
Report Industry Investment Rating - Stock index: ☆☆☆ [1] - Treasury bond: ☆☆☆ [1] Core Viewpoints - As of the week ending August 8, stock index futures rose, with IH2508 up 1.07%, IF2508 up 1.31%, IC2508 up 2.04%, and IM2508 up 2.79%. The current capital situation and bullish sentiment remain strong. Even if there is a short - term correction due to a marginal weakening of the driving force, the correction space is often limited [1]. - The net value of the financial derivatives quantitative CTA strategy rose 0.42% last week, with profits coming from going long on IC on Monday and shorting T on Wednesday. In the long - term, inflation has little difference from expectations but suppresses treasury bond futures, with PPI slightly lower than expected. In the short - term, the exchange - rate pressure from the US dollar still exists, the capital situation remains relatively loose, the margin balance has increased, and the overall market risk appetite remains high [1]. - In terms of positions, IF and IH have a marginal decline, while IC and IM remain neutral. The overall comprehensive signal is neutral and oscillating. For treasury bond futures, there was a slight rebound in the capital situation at the beginning of the month, but with the recovery of market risk appetite, the stock - bond seesaw rotation is obvious, the position factor weakens marginally, and institutions are still cautious about allocation behavior, with the comprehensive signal being neutral and oscillating [1]. Summary by Related Catalogs Macro - fundamental Medium - and High - Frequency Factor Scores - **Economic Kinetic Energy**: The blast furnace operating rate, PTA operating rate, and Shandong refinery operating rate increased by 1.35%, 1.35%, and 7.06% respectively, while the operating rate of automobile tires and polyester filament downstream looms decreased by 6.95% and 5.93%. The stock index score is 8, and the treasury bond score is 0 [2]. - **Inflation Indicators**: Some inflation indicators such as the vegetable basket product wholesale price index and coking coal index rose, while others like the 1 electrolytic copper price and styrene index fell. The stock index score is 7, and the treasury bond score is 8 [3]. - **Liquidity**: DR007 increased by 0.06%, while DR001, GC001, GC007, SHIBOR overnight, and SHIBOR 1 - week decreased. The stock index score is 9 [4]. - **Index Valuation**: The price - earnings ratio, price - sales ratio, and price - cash - flow ratio increased, while the dividend yield decreased. The stock index score is 10 [5]. - **Market Sentiment - Stock Index**: The margin balance increased by 1.48%, and the short - selling balance increased by 4.19%. The stock index score is 9 [6]. - **Market Sentiment - Bond**: The yield of 10 - year CDB bonds increased by 0.91%, and the S&P 500 volatility index decreased by 25.66%. The treasury bond score is 8 [7]. Strategy Introduction - The variety pool includes stock index futures and treasury bond futures. The short - term model focuses on market style, external factors, and capital situation, while the long - term model focuses on market expectations and macro - economic data. The position is synthesized by considering institutional long and short positions [17]. - The comprehensive signal strength is weighted by the signals of three independent models (0 - 1). Contracts with the top 2 comprehensive signal strengths greater than or equal to 0.6 are considered for long positions, and those with the bottom 2 less than or equal to 0.4 are considered for short positions. Position data within 7 days before delivery is shielded [18]. Last Week's Situation - For the IF, IH, IC, IM, T, and TF main contracts, the trading signals on different days from August 4 to August 8 are shown in the table, with some days having long, short, or no trading signals [20]. 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 factors use the instantaneous forward - rate function proposed by Nelson and Siegel, which decomposes the interest - rate term structure into level, slope, and curvature. Signals are divided into three types: '1' (large spread may decrease), '0' (uncertain spread change trend or oscillation), and '-1' (large spread may increase). The trend regression model is used to filter signals, and trading occurs when there is resonance. In actual operation, a duration - neutral ratio of 1:1.8 is used to adjust the 10 - 5Y spread [21]. - For the TF and T main contracts from August 4 to August 8, the N - S model signals and trend regression model signals are shown in the table, with some days having different signals [24].
金融工程周报:股债持仓量有所分化-20250707
Guo Tou Qi Huo· 2025-07-07 12:01
1. Report Industry Investment Ratings - Stock index: ☆☆☆ [1] - Treasury bonds: ☆☆☆ [1] 2. Core Views of the Report - As of the week ending July 4, the stock index continued to rise, with the steel sector leading the gains and the computer sector leading the losses. The margin trading balance increased by 18.928 billion yuan. The stock market is currently supported by loose liquidity and risk appetite, and the overall trading volume continues to recover. Attention should be paid to the change in the US dollar trend, as the stock index may face correction pressure [1]. - From the high - frequency macro - fundamental factor scores, for stock index futures, the inflation indicator scored 7 points, the liquidity indicator scored 5 points, the valuation indicator scored 11 points, and the market sentiment indicator scored 8 points. For bond futures, the inflation indicator scored 8 points, the liquidity indicator scored 10 points, and the market sentiment indicator scored 8 points. In terms of the term structure, the basis in July gradually converged but remained at a relatively high level and was still affected by the dividend season, with the convergence of the discount being slow [1]. - The net value of the financial derivatives quantitative CTA strategy increased by 0.85% last week, mainly due to going long on T on Wednesday and intraday long on IC on Friday. In the long - term, both the Caixin and official PMIs improved marginally, which significantly boosted IC and IM, while T declined to some extent. In the short - term, as the US dollar oscillated downward, the pressure on spreads and exchange rates decreased, and the liquidity supported both stocks and bonds, with its influence weight continuously increasing [1]. - In terms of open interest, the overall market risk appetite first rose and then fell. The signal strength of IC and IM remained high, while that of IF and IH decreased slightly compared to last week. The comprehensive signal was neutral and oscillating. For bond futures, the liquidity at the beginning of the month was still relatively loose, and the open interest factor showed a sign of marginal weakening after recovery, indicating that institutions were still waiting for the right time to allocate after adjustment. The comprehensive signal was also neutral and oscillating [1]. 3. Summary by Relevant Catalogs 3.1 Macro - fundamental High - frequency Factor Scores 3.1.1 Economic Momentum - The blast furnace operating rate increased by 1.80% to 44.97, with a historical quantile of 0.34, an index moving correlation of - 0.06, and a treasury bond moving correlation of - 0.88. The PTA operating rate in China increased by 1.80% to 79.13, with a historical quantile of 0.34, an index moving correlation of 0.01, and a treasury bond moving correlation of - 0.53. The refinery operating rate in Shandong decreased by 2.32% to 44.97, with a historical quantile of 0.01, an index moving correlation of - 0.56, and a treasury bond moving correlation of - 0.34. The operating rate of all - steel tires increased by 1.05% to 63.75, with a historical quantile of 0.43, an index moving correlation of 0.15, and a treasury bond moving correlation of - 0.52. The operating rate of downstream looms for polyester filament in the Jiangsu and Zhejiang regions decreased by 6.98% to 58.09, with a historical quantile of 0.57, an index moving correlation of - 0.44, and a treasury bond moving correlation of 0.81. The stock index futures score was 7, and the bond futures score was 0 [2]. 3.1.2 Inflation Indicators - Multiple inflation - related indicators such as the vegetable basket product wholesale price index, the coking coal index, and the price of electrolytic copper showed varying degrees of weekly changes. The stock index futures inflation score was 7, and the bond futures inflation score was 8 [3]. 3.1.3 Liquidity - Liquidity - related indicators such as DR007, DR001, and the US dollar index all changed last week. The stock index futures liquidity score was 5 [4]. 3.1.4 Index Valuation - Index valuation indicators such as PE, PS, and dividend yield showed different changes. The stock index futures valuation score was 10 [5]. 3.1.5 Market Sentiment - For the stock market, indicators such as margin trading balance and northbound capital inflows showed different trends. The stock index futures market sentiment score was 8. For the bond market, indicators such as the yield of 10 - year CDB bonds and the S&P 500 volatility index also showed different trends, and the bond futures market sentiment score was 8 [6][7]. 3.2 Strategy Introduction - The product pool includes stock index futures and treasury bond futures. The goal is to use a multi - strategy model to select and allocate contracts in the financial futures market to achieve stable net value growth. The short - term model focuses on market style, external factors, and liquidity, while the long - term model focuses on market expectations and macro - economic data. Open interest is synthesized by considering institutional long and short positions [17]. 3.3 Forecast Signals and Historical Data - As of last Friday, the short - term model, open interest indicator, long - term model, and comprehensive signals for IF, IH, IC, IM, T, and TF were provided. The comprehensive signal strength is weighted by three independent models (0 - 1). The trading rules include principles for going long and short, signal shielding during the delivery period, stop - loss points, and capital allocation [18]. - The trading signals of IF, IH, IC, IM, T, and TF from June 30 to July 4, 2025, were presented, along with the net value tracking data [20]. 3.4 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 factors use the instantaneous forward - rate function proposed by Nelson and Siegel, which decomposes the interest - rate term structure into three parts: level, slope, and curvature. The model uses PCA, factor rotation, and logistic regression to construct the three - factor model, with signals divided into three categories. The trend regression model is used to filter signals, and trading occurs when there is resonance. In actual operation, a duration - neutral ratio of 1:1.8 is used to adjust the 10 - 5Y spread [21]. - The trading signals of the TF and T main contracts from June 30 to July 4, 2025, were provided [24].