✦Research Report Catalog✦
✦Brief Description✦ _
Sort by release time
Northeast Securities
Factor stock selection series four: Realized high-order moment factors and improvements under high-frequency data
Release date: 2023-03-01 Keywords: stock, high-frequency data, high-order moments
Main content: This article serves as a preliminary study on the study of high-frequency factors. For the research on high-frequency factors in the return distribution category, we found that the constructed and improved factors have better stock selection performance. Among them, the subRK_5min_week factor we constructed and improved has an excess return of 13.03% in 2022.
Huaxi Securities
Financial Engineering Research Report: Industry Effective Analyst Expectation Factors and Industry Rotation Strategy
Release date: 2023-03-02 Keywords: stocks, analysts, industry rotation
Main content: We start with the data of analysts' expectations, synthesize the indicators related to analysts' expectations of individual stocks into industry factors, and classify them according to their essential meanings, including changes in analysts' expectations, market confidence, acceleration of report coverage, and institutional coverage Acceleration, exceeding expectations in financial statements, and exceeding expectations in performance forecasts. And further based on the expansion method, all effective factors are aggregated to construct analysts' expected industry rotation strategy, with an annualized return of 18.36% and an annualized excess of 12.86%.
Guosheng Securities
Quantitative Special Report: Convertible Bond Pricing Model and Application
Release date: 2023-03-02 Keywords: convertible bonds, pricing model, CCB model
Main content: Based on the existing research results, this paper studies and demonstrates the "Callable Convertible Bond Pricing Model with Redemption Protection Period". We call it the CCB model for short. Compared with the traditional model, this model has many advantages. The "undervaluation strategy" constructed based on this factor can achieve an annualized return of 20.6%, with fluctuations and drawdowns of only 12.6% and 11.7%, respectively, an excess return of 8.5%, and an information ratio of 2.28.
Huafu Securities
Topic on Market Sentiment Indicators (5): Top and Bottom Signals of Industry Indexes: Proportion of Net New Highs ((NH-NL)%)
Release Date: 2023-03-03 Keywords: Stocks, Industry, Proportion of Net New Highs
Main content: The article describes the sentiment of the industry index by constructing the net new high ratio ((NH-NL)%) indicator, which continues our emphasis on new highs and lows in prices, and its core logic is still based on the anchoring of behavioral finance effect. Based on the net new high percentage ((NH-NL)%) signal, we constructed a simple reversal strategy, including elements of buy, sell, position, stop loss and move.
Orient Securities
Eighty-eight of the factor stock selection series: an enhanced plan based on partial stock fund index
Release Date: 2023-03-06 Keywords: Stocks, Partial Equity Fund Index, Index Enhancement
Main content: For the two partial stock fund indexes 885001.WI and 930950.CSI, this report detects their stock positions, and builds an enhanced portfolio on this basis. Among them, the annualized excess returns of 930950.CSI under the enhanced combination of fundamental factors and volume price factors are 4.4% and 9.6% respectively.
Guosheng Securities
"Volume Price Gold Rush" Stock Selection Factor Series Research (3): How to build an effective stock selection factor based on RSI technical indicators?
Release date: 2023-03-07 Keywords: stock, RSI, high frequency
Main content: This article conducts an in-depth exploration of the RSI indicator, discusses how to expand it to the field of cross-sectional stock selection, and constructs effective stock selection factors. We tried to increase the data frequency, and constructed high-frequency RSI factors based on the minute-by-minute rise and fall data. The effect of stock selection has been significantly improved, and the information ratio of 10-group long-short hedging is close to 2. Furthermore, the daily RSI index is weighted by using the information of the trading volume, and the trading volume matching RSI factor with a more stable stock selection effect is obtained.
pacific ocean
Interest Rate Quantification Series: Research on the Correlation Between Low-Frequency Factor Momentum and Macro Interest Rate
Release date: 2023-03-07 Keywords: macro, interest rate, factor momentum
Main content: This paper starts from the perspective of time section and studies the relationship between the dynamic attributes of macro factors and the rise and fall of interest rates, aiming to find a momentum model with greater certainty with the rise and fall of interest rates. The article finds that there is a negative correlation delay effect between the year-on-year growth rate of export delivery value and the differential momentum of the PMI raw material purchase price index, and there is a positive positive correlation delay effect between the business climate index and the inertia momentum of the GDP deflator.
Huachuang Securities
CSI 300 index enhancement strategy based on canslim and FESC
Release date: 2023-03-07 Keywords: stock, index enhancement
Main content: The article tests the stock selection effect of canslim2.0 in various industries, and screens the industries where the canslim strategy is applicable; and tests the stock selection effect of canslim2.0 in various industries, and screens the industries where the canslim strategy is applicable. During the backtest period, the annualized rate of return of the strategy was 10.2%, the annualized rate of return of the benchmark (CSI 300) was 4.7%, the annualized excess return of the strategy relative to the CSI 300 was 5.5%, and the maximum drawdown of the strategy was 49.1%.
Tebon Securities
Debon Metalworking Small Market Value Topic 3: Large-capacity National Securities 2000 Enhancement Strategy
Release Date: 2023-03-07 Keywords: Stocks, National Securities 2000, Small Market Value
Main content: The article tests the enhanced effect of the machine learning model in the China Securities 2000 Index. The machine learning residual factor, inversion factor and compound factor grouping backtest all have robust monotonicity. Synthesize the machine learning compound factor and the scale factor into a new stock selection factor, construct the national security 2000 enhanced strategy, and the excess annualized rate of return of the synthetic factor long group is 15.4%.
pacific ocean
Special Research Report on Financial Engineering: ETF Selection Strategy - Based on the Overcompetitive Hypothesis
Release date: 2023-03-08 Keywords: fund, ETF, excessive competition
Main content: The ETF market can be divided into two categories from the perspective of fund targets, one is a broad-based category based on an index, and the other is a specialized category based on dimensions such as industry or style. One of the differences between the two types of ETFs is the target buyer group. The needs of buyers of broad base classes are primarily to replicate the index. Buyers of specializations are based on their belief in a concept. Based on this, the construction of specialized ETFs will cater to the needs of investors and choose tracks with high attention. Our research found that avoiding investing in newly created ETFs within a year can significantly improve portfolio performance.
Sealand Securities
Market Watch Series: Exploring the timing strategy behind the calendar effect
Release date: 2023-03-08 Keywords: stock, calendar effect, timing
Main content: The article believes that there is a calendar effect with significant effect and stable cycle. It is found that the ten-year effect within a month is close to the time node of the social financing effect, and the rhythm is similar, and the common one of the two strengthens the "half-month effect" in the calendar effect. The article constructs timing strategies around the characteristics of the calendar effect. Among them, the excess of the industry rotation preferred calendar long strategy increased from 6.74% to 12.50%.
Zheshang Securities
Financial Engineering Depth: Index Enhancement Strategies Based on Constituent Stock Timing
Release date: 2023-03-08 Keywords: stock, index enhancement, timing
Main content: This paper applies the AlphaCY timing system launched in the previous research to the stock market, and proposes a method for constructing index growth strategies without factor timing by using a stock selection method different from factor scoring. And the application of China Securities 1000 Index has been verified. Among them, the CSI 1000 enhanced equal-weight portfolio achieved a cumulative excess return of 84% from 2017 to 2022, with an average annualized excess return of 11.34%.
Warburg Securities
Special Research on Indexed Investment: How to Choose a Dividend ETF? ——Research based on hierarchical clustering method
Release date: 2023-03-09 Keywords: fund, ETF, dividend
Main content: The article classifies 9 indexes into 4 categories through hierarchical clustering method. The first category is the dividend strategy index that is biased towards the broader market, the second category is the mainstream dividend strategy index with the purest exposure to dividend factors, the third category is the dividend strategy index weighted by market capitalization, and the fourth category is the dividend + quality mixed strategy index , and gives different investment suggestions according to different types.
Huatai Futures
Quantitative Special Report: How CTA Breakthrough
Release date: 2023-03-09 Keywords: futures, CTA
Main content: CTA funds represented by trend strategies have retreated since March 2022. We believe that 1) transaction congestion has intensified; 2) effective market volatility has decreased; 3) strategy applicability has weakened; These four factors are significant contributors to poor CTA performance.
Huaan Securities
Quantitative Fundamentals Series Report Nine: Quantitative Practice of Comparable Company Law: Reshaping Value Factors
Release Date: 2023-03-10 Keywords: Stocks, Quantification of Fundamentals, Value Factors
Main content: Define comparable companies from the five dimensions of life cycle, company size, profitability, asset structure, and operating efficiency, and calculate the similarity between comparable companies through cosine similarity. Under the framework of comparable companies, the value factor is improved, and the stock selection effect of the synthesized large-category related value factor has been significantly improved. Compared with the original value factor, ICIR has increased from 2.47 to 3.75, and the IC monthly winning rate has increased from 74.4% to 86.8%. Long and short The annualized return increased by 3.3%.
Huaan Securities
Quantitative Fundamentals Series Report 8: Looking for the "comfort zone" of stock selection strategy and industry rotation strategy
Release date: 2023-03-10 Keywords: stocks, quantification of fundamentals, timing
Main content: Starting from the basic logic, we construct a comprehensive analysis based on macro and market (economic fundamentals, market cycle), meso (style switching, industry switching, capital switching), micro (industry differentiation, business climate divergence, investor structure) ) multi-dimensional index system, through the index voting method to synthesize strategy timing signals, and comprehensively evaluate which type of strategy the current market environment is in the "comfort zone".
CICC
Quantifying multi-factor series (11): How to capture the four-quadrant style rotation?
Release date: 2023-03-11 Keywords: stock, style rotation, four-quadrant
Main content: The article constructs a two-dimensional composite score index based on 15 indicators in three dimensions, and further builds a four-quadrant style position adjustment rotation strategy through the coordinate method. The annualized return of the rotation strategy is 9.30%, and the annualized excess return is 5.06%.
Orient Securities
Interest Rate Timing Based on Neural Network Model - Macro Fixed Income Quantitative Research Series (9)
Release date: 2023-03-13 Keywords: fixed income, bonds, neural network
Main content: This report introduces the relevant model of the neural network, and conducts training based on the daily frequency volume price factor (feature), so as to predict the rise and fall of interest rates and the rise and fall of treasury bond futures in the next N days, so as to form a daily long-short signal . For the 10-year treasury bond futures model, the annualized return reaches 5.98%, the return-to-risk ratio reaches 2.36, and the average trading days is 4.79 days/time, among which, the annualized return of the long side is 3.81%, and the annualized rate of the short side is 2.26%.
Zheshang Securities
Financial Engineering Special Topic: Research on High-Frequency Factor Series - The Influence of Intraday Trading Volume Proportion on Slippage Cost
Release Date: 2023-03-14 Keywords: Stocks, Intraday Trading Volume, Slippage
Main content: This paper takes the high-frequency factor of the proportion of trading volume as the research object, and takes the opening price of the day as the pending order price, and studies the correlation between the slippage cost of intraday minute-level trading and this factor. The study found that the proportion of trading volume has a significant linear relationship with the cost of slippage, and it is relatively stable below 20%; the accuracy of the cycle step is improved, and the volatility of marginal slippage cost is approximately proportional to the proportion of trading volume.
Sinolink Securities
Beta Hunter Series II: Entropy Pool Model: How to Incorporate Pure Active Viewpoint into Quantitative Allocation Model
Release date: 2023-03-14 Keywords: stock, entropy pool model, active view
Main content: The article introduces a new model framework (entropy pool model) to make the allocation model more applicable, and gives examples to illustrate that the comparative views used by subjective investors and macro external factor views can be put into the allocation model, and used The non-parametric method improves the estimation accuracy and obtains more stable results. Among them, the allocation strategy of using macro factor allocation signals to construct sorting views and superimposing external factors reached 23.82% per year, and the Sharpe ratio was 1.624.
Everbright Securities
Revisiting Institutional Research: Fine Segmentation of Institutional Research——The Ninth Series of Quantitative Stock Selection Reports
Release date: 2023-03-15 Keywords: stock, institutional research
Main content: The article conducts an in-depth study of institutional investors’ research activities on listed companies, and believes that institutional research information should not be adjusted through daily event-driven methods, and institutional research is superimposed on performance forecasts, brokerage gold stocks, and sudden investigations. Significant benefits; a scoring system for institutional research has been constructed. After testing, the annualized rate of return of the model is 33.91%, and the Sharpe ratio is 0.92.
Warburg Securities
Financial Engineering Special Report: How to Build the Impact Factors and Monitoring Model of CTA Strategy?
Release date: 2023-03-17 Keywords: futures, CTA, monitoring model
Main content: This article draws on the multi-factor research ideas of the stock market. We split the influencing factors of the CTA strategy into independent factors, and study the rotation rules of the factors, and sort the various CTA factors in different environments. Scoring, and selecting the factors with the highest scores to build a rotation model.
Everbright Securities
Alpha in Fund Flow Data——Factor Combing and Backtesting Part 2
Release date: 2023-03-17 Keywords: stocks, capital flow
Main content: This series aims to provide investors with a factor reference manual. This report is the second in the series and constructs factors based on capital flow data. Among them, the total single type and special period trading factors performed better, and the weekly RankIC of the momentum-adjusted large single net inflow factor in the liquidity 1500 stock pool was 4.32%.
Guotai Junan
How to copy options to create an investment weapon - the principle and practice of option copying
Release Date: 2023-03-21 Keywords: Option, Option Copy
Main content: This report introduces the option replication solution, including the basic principles of replication and potential sources of risk, and conducts actual combat tests in the domestic market. The replication effect of the vanilla option is similar to that of the spread option, and the replication effect of the binary option is weak.
Orient Futures
Fundamental Quantitative Special Report: Entropy-Based Multi-Asset Timing Model
Release date: 2023-03-23 Keywords: index, timing, entropy
Main content: The article attempts to test the relationship between the entropy of financial time series and the price fluctuation of financial assets, including stock index, national debt, commodity and foreign exchange asset targets, calculate the sample entropy index of each asset, set the opening and closing rules through slope and average line, and study It is believed that the winning rate of the combination tested by the Shanghai Composite Index is as high as 59.6%, and the annual average return is recorded as 392.4 in terms of index points.
Everbright Securities
Capturing the "smart money" among individual investors - the ninth report of the quantitative stock selection series
Release date: 2023-03-24 Keywords: stocks, smart money, individual investors
Main content: The article takes individual investors who have entered the top ten tradable shareholders of listed companies as the research object, further quantitatively analyzes "smart", tracks "smart" "individual investors", tracks new stocks and builds a portfolio. Among them, the excess annualized rate of return of excellent short-term traders tracking combination strategy is 16.39%, and the monthly winning rate is 66.94%.
Caixin Securities
Metalworking special industry rotation series 1: Quantity and energy reminder of turnover rate
Release date: 2023-03-24 Keywords: stocks, industry rotation, turnover rate
Main content: The article constructs a portfolio with equal weight by screening the industries with "heavy volume at the bottom", "shrinking volume and rising", and eliminating "heavy volume at the top". In the long run, the annualized return of the strategy is 4pct.+ higher than that of the equal weighted portfolio, Sharpe Nearly double the ratio, while reducing the maximum drawdown.
Sinolink Securities
Beta Hunter Series III: All-round identification and rotation strategies that exceed industry expectations
Release date: 2023-03-24 Keywords: Stocks, industry exceeding expectations, rotation strategy
Main content: In this report, we will measure the unexpected characteristics at the industry level from the two dimensions of performance exceeding expectations and text exceeding expectations, and construct an industry rotation factor. The unexpected enhanced rotation factor synthesized in the article has an average IC value of 9.70%, an annualized rate of return of long and short positions of 20.43%, and a Sharpe ratio of 1.15.
CITIC Construction Investment
Quantitative Special Topic· Commodity Futures Market Risk Factor Model (1) - Overview of Risk Model
Release date: 2023-03-26 Keywords: futures, risk factors
Main content: This series of reports discusses, analyzes and studies the risk model of the domestic commodity market. This article is the first in the series, and mainly gives a general description of the functions of risk models, the classification of risk models, and the evolution of risk models.
Founder Securities
The Tenth Research Series of Multi-factor Stock Selection: Decomposition of Factors Driving Individual Stock Price Changes and Factors of "Hidden Flowers in the Forest"
Release date: 2023-03-27 Keywords: stock, multiple factors, price change
Main content: The article constructs the factors of "morning mist in the morning", "ancient trees in the afternoon" and "sleep road at night", and finally synthesized into the factor of "flowers hidden in the forest". Stocks (flowers) with better meanings are often very There is less information at the market level (hidden in the forest), Rank IC is -9.34%, Rank ICIR is -5.69, the annualized return rate of the long-short combination is 32.39%, the information ratio is 4.46, and the factor monthly winning rate is 88.43%.
Huaxi Securities
Industry Prosperity Rotation Research IV: Dynamic Improvement and Factor Disassembly of the Industry Prosperity Driving Model
Release date: 2023-03-27 Keywords: industry, prosperity
Main content: In this report, we classify and delete the mesoscale factors in the past to reduce the impact of collinearity on the model. At the same time, we abandon the transformation of the model by ad hoc sparsity methods such as lasso, and face high-dimensional data directly. Influence. Compared with the average return of the industry, the top five industries selected according to the model recommendation have an annualized excess return of 16.90%, an information ratio of 1.54, and a monthly winning rate of 65%.
Dongguan Securities
Special Report on Options in Financial Engineering Research Series: Introduction to Snowball Options of OTC Derivatives (1)
Release date: 2023-03-28 Keywords: options, snowball
Main content: The income of the snowball option issuer comes from the delta hedging income of the CSI 500 stock index futures and the fixed-income products purchased with idle margin. The stock index futures delta hedging method is to sell high and buy low. At the same time, CSI 500 stock index futures can also bring additional income with an annualized discount of 5%-15%. Because the stock index futures contract must converge to the spot price on the delivery day.
Xiangcai Securities
Multi-factor Quantitative Stock Selection Series IV: Research and Testing of New Technology Factors
Release date: 2023-03-28 Keywords: stocks, technical factors
Main content: This report continues to explore new technical factors from the aspects of turnover rate, illiquidity and the combination of volume and price information. After adding the enhanced strategy of the CSI 500 Index, the improved strategy will increase from 2021.12 to 2023.02. The excess return will increase from 5.43% to 7.21%, the information ratio will increase from 1.34 to 1.77, and the Calmar ratio will increase from 1.42 to 2.12.
Orient Securities
Factor 89: Analyst Sentiment Adjustment Score ASAS
Release date: 2023-03-29 Keywords: stocks, analysts, sentiment
Main Content: Analyst Sentiment Adjustment Score Factor, which captures analysts' views on stocks through their research report title and summary text sequences, and combines earnings forecast adjustment values as labels to train the model. Use a two-layer Transformer and a one-dimensional convolutional network to extract features, and calculate the average emotional score of a stock in the past three months as a quantitative stock selection factor. The average RankIC of the full sample is 0.04, and the ICIR is 2.0.
Minsheng Securities
Quantitative Analysis Report: Prospects for the Application of Large Language Model (LLM) in Quantitative Finance
Release date: 2023-03-30 Keywords: big language model, quantization
Main content: At present, the large language model (LLM) has become the focus of the market. This report will look forward to the application of LLM in the field of quantitative finance from various angles. The representative model of LLM, GPT-4, is a neural network that predicts the context (words) based on the pre-trained given context. Through the autoregressive language model, GPT realizes the function from word inference to complete answer.
Orient Futures
National Bond Futures Quantification Series IV: Discussion on Strategy Framework Based on Multiple Deep Learning Models
Release date: 2023-03-30 Keywords: futures, deep learning
Main content: In this article, we discuss the application of deep learning models under the target return volatility, explore the multi-variety quantitative strategy configuration framework, and obtain the optimal configuration based on position prediction based on different models.
Orient Securities
Quantitative strategy configuration series (1): Achilles' heel of CTA strategy
Release Date: 2023-03-30 Keywords: Futures, CTA
Main content: Based on the selected indicator parameters and thresholds, we simply built an early warning signal mechanism to better predict changes in the future market environment. Through the early warning mechanism, we dynamically configure the time series momentum strategy (CTA trend following strategy) and time series reversal strategy (CTA trend reversal strategy). The test results show that the Sharpe ratio of the strategy has increased from 0.86 to 1.32.
Minsheng Securities
Quantitative special report: Quantitative capture of industry Davis double-click opportunities driven by macroeconomics
Release Date: 2023-03-30 Keywords: Industry, Macro, Prosperity
Main content: The article introduces prosperity degree and valuation as intermediate variables between macro variables and asset prices, and converts the forecast target of macro indicators from yield to prosperity degree and valuation forecast, and then uses industry prosperity degree and industry valuation Value changes drive industry price changes. The absolute annualized return of the strategy is 28.0%, and the excess annualized return is 17.6%, showing excellent performance of the strategy.
CITIC Securities
Fundamental Quantification Series - Topic on Data Strategy of Central Enterprises (1): Scientific Quantification of Analytical Perspectives of Central Enterprises, and Dynamic Tracking of Multidimensional Data Strategies
Release date: 2023-03-31 Keywords: stocks, fundamentals, central enterprise data
Main content: Based on policy, ESG, valuation, central enterprise assessment and other five dimensions, combined with the industry configuration framework to construct a monthly central enterprise scoring data strategy, in which the policy dividend index and ESG rating system have obvious long-short combination benefits in retrospect .
- End -