About Me

Here is Weili Song (宋伟力).

I am currently a PhD candidate in Management Science and Engineering at Hunan University, focusing on Quantitative Finance and Financial Risk Management. I am passionate about quantitative investment research aimed at live trading, dedicated to closely combining theory with practice and continuously exploring innovative investment strategies.

I also work part-time as a quantitative researcher at a private equity fund, where I have been part of a two-person team that developed a stock index enhancement product from the data stage to full implementation. Since 2022, I have attended over 200 roadshows by private equity funds and regularly engage in offline discussions with industry peers, giving me a deep understanding of quantitative private equity funds’ stock index enhancement products.

Collaboration Opportunities

I warmly welcome various external collaboration opportunities, including but not limited to:

Industry exchange: Sharing professional experiences and discussing industry trends.
FOF fund part-time work: Providing expert analysis on quantitative stock selection products.
Investor cooperation: Able to provide detailed live trading performance.
Academic collaboration: As a doctoral student, I am also actively seeking academic research collaboration opportunities.

If you are interested in any of the above areas, please feel free to contact me via WeChat or email. My email address is [email protected]. I am currently based in Beijing and welcome face-to-face meetings.

Live Trading Product

In collaboration with my partner at the private equity fund (a team of two), we dedicated ourselves to researching quantitative stock selection strategies from 2022 to 2023. We officially launched our CSI 500 Index enhancement product for live trading on October 22, 2023. Over the past 10 months of operation, our product has achieved the following results:

A significant excess return of 21.90% relative to the CSI 500 Index.
Far surpassing the industry average (the average excess return of CSI 500 enhancement products from large-scale quantitative private equity funds over the past year was only 8%).
Achieved an excellent excess Sharpe ratio of 2.398.

During the quantitative stock market crash in February 2024, our performance:

Maximum excess drawdown was only 7%, (the average maximum excess drawdown of CSI 500 enhancement products from large-scale quantitative private equity funds over the past year was as high as 11%).
Recovered from the excess drawdown in just 10 trading days (the average recovery time for CSI 500 enhancement products from large-scale quantitative private equity funds was as long as 5 months).

Characteristics of our trading strategy:

Stock selection pool covers the entire market, with an average holding of 300-500 stocks.
Employs intraday medium to low-frequency signals, without the participation of intraday high-frequency round-trip trading signals, not falling under the category of high-frequency trading closely monitored by regulatory authorities.
Annualized two-way turnover rate of approximately 30-50 times, with large capital capacity.

Research Interests

My current research focus is in the field of Quantitative Finance, dedicated to applying advanced technologies to enhance mid-frequency prediction capabilities and optimize risk management, thereby improving live trading performance. My main research directions include:

Computational Finance: Exploring new factor mining methods on price and volume data.
Financial Machine Learning: Researching and applying advanced non-linear models to combine factors, improving the predictive power of signals.
Portfolio Management: Exploring better portfolio optimization methods, including developing more effective optimization algorithms and mining more effective risk factors to achieve superior risk-adjusted returns.

News and Updates

August 2024: Co-authored and submitted a research paper titled “AlphaForge: A Framework to Mine and Dynamically Combine Formulaic Alpha Factors”. This paper proposes an innovative two-stage framework for formulaic Alpha factor generation, providing a new methodology for factor mining and dynamic combination in the field of quantitative investment.
October 2023: Ranked 88th/2664 (Silver Medal) in the Kaggle Large Language Model Science Exam competition; Ranked 23rd/2064 (Silver Medal) in the CommonLit - Evaluate Student Summaries competition.
August 2023: Introduced a portfolio optimization framework using Gurobi to control risk exposure and enhance small-scale fund performance, increasing annualized excess returns to 20% while keeping drawdowns under 3%.
June 2023: Conducted research on factor generation and combination, achieving 20% annualized excess returns and excess drawdowns within 3% in backtests on the recent 2-year test set. This research laid the foundation for the subsequent “AlphaForge” project.
April 2023: Successfully implemented a low-frequency alpha factor extraction using BERT-based sentiment classification, achieving a 10% excess return on sentiment signals alone.
December 2022: Placed 6th/532 in the 5th “UBIQUANT CHALLENGE” Quantitative League.
August 2022: Achieved 50th/4874 (Solo Silver Medal) in the Kaggle AMEX Credit Default Prediction competition.