Bilateral Autotrading Framework for Stock Return Forecasting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Quantitative trading methods often focus on achieving high Pearson Correlation Coefficient (CORR) while ignoring indicators that may produce high profits, leading to missed opportunities in stock prediction.
Innovation Solution
The Bilateral Autotrading Framework (BAF) based on Bilateral Loss is used to forecast the cross-sectional rank of stock returns, employing a weighted Bilateral Correlation Coefficient (BCORR) to detect profitable indicators with low CORR, and optimizing stock positions using Sharpe-oriented optimization to reduce risk and improve returns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If quantitative trading methods focus on achieving high Pearson Correlation Coefficient (CORR), then measurement precision is improved, but profitable indicators with low CORR are ignored, leading to loss of information
Solution Approach 1:
The patent transforms the single-parameter CORR metric into a multi-dimensional evaluation system by introducing BCORR with asymmetric weighting. This parameter transformation allows the system to simultaneously consider both high CORR indicators and low CORR indicators that may be profitable, converting the trade-off into a comprehensive assessment that captures previously overlooked profitable patterns
Solution Approach 2:
The patent adds a new dimension to the evaluation by introducing asymmetric weighting based on prediction direction (upward vs downward movements). Instead of treating all CORR values equally, the system creates a weighted BCORR metric that amplifies the importance of certain directional predictions, effectively adding a weighting dimension to the traditional CORR measurement
2Measurement precision
If traditional quantitative trading focuses on high CORR indicators, then prediction accuracy is improved, but adaptability to different market conditions deteriorates
Solution Approach 1:
The patent implements dynamic adaptability by adjusting the weighting parameters in BCORR based on market conditions. The asymmetric weighting can be modified to emphasize different directional predictions depending on the current market environment, allowing the system to adapt to varying market conditions while maintaining prediction accuracy through the core BCORR framework
Solution Approach 2:
The system changes parameters dynamically by adjusting the weighting factors in the BCORR calculation based on observed market behavior. When certain types of movements become more prevalent or profitable, the weighting parameters are adjusted to prioritize those directions, enabling the system to adapt to different market regimes
Data Source
AI summary
In a method for quantitative investment using a bilateral autotrading framework (BAF), a processor receives a market dataset comprising stock prices, constructs a time series input by applying cross-sectional rank forecasting to the market dataset, generates a bilateral indicator based on the time series input, predicts a rank of stock return based on the bilateral indicator, executing, by one or more processors, an adjustment of a position in a stock based on the rank of stock return.