Minjie Ding

Projects

These projects support the research workflow, from data collection to repeatable analysis and writing systems.

Illustration for the MeShare China macroeconomic data library project

AI Platform

MeShare: China Macroeconomic Data Library

  • A Python library for China's macroeconomic data, featuring bilingual documentation in Chinese and English for data querying, automated charting, and assisted analysis.
  • Compared to common API, MeShare focus more on China's macroeconomic data, covering GDP, prices, PMI, finance, interest rates, bonds, and foreign exchange.
Illustration for the Macro Desk industry macro research drafting skill project

AI Skill

Macro Desk: Skill for Industry Macro Research Drafting

  • An AI skill for automatically drafting macro research reports around China's data releases, covering outlooks, event commentaries, thematic briefings, and macroeconomic weekly reports.
  • Sample reports are included in the project documentation.
Illustration for the Macro Quant Strategies notes and code project

Code

Macro Quant Strategies

  • Notes and code on traditional factor quantitative investing and emerging macro quantitative investing.
  • Starting from CAPM and classical style factors, progressively expanding to a broader factor library and macro allocation strategies.
Illustration for the Monetary Economics and Heterogeneous Agent Model tutorial project

Code

Monetary Economics and Heterogeneous Agent Model

  • Notes and code on monetary economics, covering RBC models, New Keynesian frameworks, numerical methods, the Aiyagari model, and HA models.
  • RBC models are implemented using MATLAB, Dynare, Python, and Julia. Other models are implemented primarily in MATLAB.
Illustration for the Lianjia housing transaction data pipeline project

Data

Lianjia Housing Transaction Data Pipeline

  • A Selenium-based web scraper for Lianjia transaction data, supporting batch scraping by district, resumable downloads, and CAPTCHA interruption handling.
  • Lianjia's publicly listed transactions are dynamically adjusted, causing historical listings to appear and disappear inconsistently across different scraping times.
  • An account gets banned after scraping approximately 15,000 items on average, making the data acquisition cost relatively high. Like-minded peers are welcome to collaborate.
Illustration for the food delivery courier survey project

Data

Discount Rates and MPC among Food Delivery Couriers

  • Collected demographic, financial, discount factor, and consumption propensity data of frontline delivery riders through offline interviews and online questionnaires.
  • Two rounds of surveys have been completed so far, covering over 30,000 respondents, with a 5 RMB cash reward offered for valid responses, yielding over 8,000 data entries.
  • This has successfully materialized abstract economic concepts like discount factors and consumption propensities into concrete figures, holding high research value.