AI Quantitative Research
Exploring financial data through artificial intelligence and quantitative methods, with attention to data quality, hypotheses and model evaluation.
Explore this areaDATA · MODELS · VALIDATION
HONG KONG · EST. 2025
Quantitative thinking.
Intelligent systems.
Bringing together AI quantitative research, asset management and blockchain development at the intersection of data, finance and software engineering.
Explore our focus
JINHENG QUANTITATIVE ASSET MANAGEMENT LIMITED金恆量化資產管理有限公司
47,000Cumulative users
US$100M+Assets under management (USD)
As of 1 August 2026
Figures provided by the company01 / ABOUT JINHENG
At the intersection of
finance and technology.
Founder
Benjamin Gimson
Jinheng Quantitative Asset Management Limited is a Hong Kong company focused on AI quantitative research, asset management and blockchain technology development.
We are interested in turning financial questions into research that can be analysed and tested. Data preparation, model research and system design provide a structured basis for that work.
From a research hypothesis to a portfolio analysis framework or a supporting software system, a clearly defined question—and an understanding of its limits—is the starting point.
02 / OUR FOCUS
Three areas.
A connected perspective.
Research to understand data. Risk analysis to examine assets. Engineering to explore practical applications.
Asset Management
Considering allocation, portfolio structure and risk, with attention to the relationship between investment objectives and the management process.
Explore this areaALLOCATION · PORTFOLIOS · RISK
Blockchain Development
Exploring distributed systems, smart contracts and application integration in the context of specific business and software requirements.
Explore this areaARCHITECTURE · LOGIC · INTEGRATION
01 / AI QUANTITATIVE RESEARCH
Start with data.
Make the research testable.
Useful quantitative research begins with an understanding of the question and the data. Areas of interest include preparing information from different sources, exploring relationships between variables and evaluating models under different conditions.
Historical data provides a basis for observation, with limitations of its own. Research needs to distinguish data used to develop a model from data used to evaluate it, while considering transaction costs, data bias and changing market conditions.
AREAS OF INTEREST
Data and feature research
Examine sources, time alignment, missing values and outliers, and explore variables that help describe market behaviour.
Model research and comparison
Start from a clear hypothesis, compare statistical and machine learning approaches, and consider the conditions in which each may be useful.
Backtesting and robustness
Examine results across periods, parameters and scenarios, with attention to overfitting and out-of-sample performance.
02 / ASSET MANAGEMENT
Understand the relationships.
Recognise the boundaries.
Asset management brings together objectives, time horizons, liquidity and risk. Quantitative analysis offers a structured way to examine the roles that different assets play within a portfolio.
Our focus includes portfolio research and analysis: allocation approaches, risk distribution, correlations and changing scenarios. Alongside individual measures, it is necessary to understand their assumptions and the scope of the underlying data.
AREAS OF INTEREST
Asset allocation research
Compare allocation approaches in the context of time horizons, liquidity needs and relationships between asset classes.
Portfolio risk analysis
Examine volatility, drawdowns, concentration and correlations to understand different sources of portfolio risk.
Scenarios and ongoing review
Consider how changes in market conditions may affect a portfolio, with a framework for continued observation and reassessment.
03 / BLOCKCHAIN DEVELOPMENT
From business requirements
to considered system design.
Blockchain development starts with practical questions: what problem should the system solve, what information needs to be recorded, who can access or update it, and how should it connect to existing workflows?
Our technical focus includes system architecture, smart contract logic and application integration. Design needs to consider traceability, permissions, privacy, operating costs and ongoing maintenance together.
AREAS OF INTEREST
Distributed system design
Explore data flows, state changes, access controls and the division of responsibilities between on-chain and off-chain systems.
Smart contract logic
Define business rules, trigger conditions and exceptions, with attention to testing and understandable contract behaviour.
Application and data integration
Consider how blockchain data connects to software interfaces, databases and business processes, including usability and maintenance.
03 / OUR APPROACH
Begin with a question.
Progress through validation.
A clear framework helps research, analysis and development work towards the same objective.
- 01
Define the question
Clarify the objective, context and constraints. Distinguish the question to be answered from the system to be built.
- 02
Establish the evidence
Examine data sources and quality, record key assumptions, and make the scope of the evidence clear.
- 03
Analyse and validate
Use model comparisons, scenario analysis or system testing to assess how well an approach addresses the original question.
- 04
Review and refine
Document findings and limitations, revisit assumptions, and use the results to inform the next stage of research or development.
OUR PRINCIPLES
Technology is the tool.
Understanding is the foundation.
Data needs context
The source, timing and definition of data shape the questions an analysis can answer.
Risk has dimensions
Models, assets and systems have different limitations that need to be understood in context.
Technology needs a purpose
Technical choices should address practical needs and account for long-term use and maintenance.
04 / INDUSTRY PRACTICE
Understanding technology
through industry practice.
These examples draw on third-party official publications for industry reference. They are not Jinheng client projects, investment results or partnerships.
AI QUANTITATIVE RESEARCH
Two Sigma
Data and models in systematic investing
Two Sigma describes a process spanning data preparation, modelling, portfolio construction and execution. Its research uses quantitative methods and techniques such as natural language processing across different types of data.
RESEARCH PERSPECTIVE
View official sourceThis example illustrates the connections between data preparation, model analysis and portfolio decisions.
ASSET MANAGEMENT & RISK
BlackRock · Aladdin Risk
A whole-portfolio view of risk
Aladdin Risk brings together analytics across asset classes. Exposures, scenario analysis and stress testing help evaluate how a portfolio may respond to different market conditions.
RESEARCH PERSPECTIVE
View official sourceThis example illustrates the role of consistent analytics and scenario comparisons in portfolio research.
BLOCKCHAIN TECHNOLOGY
J.P. Morgan · Kinexys
Permissioned blockchain in payments
Kinexys Digital Payments uses a permissioned blockchain as a payment rail and deposit ledger, supporting fund transfers and corporate liquidity management use cases.
RESEARCH PERSPECTIVE
View official sourceThis example shows how blockchain applications can be designed around records, participant permissions and existing financial workflows.
05 / QUESTIONS & ANSWERS
A closer look at Jinheng.
What are Jinheng’s main areas of focus?
The company’s business directions are AI quantitative research, asset management and blockchain technology development. They address data and models, assets and risk, and software and system design respectively.
What role does AI play in quantitative research?
AI can help explore relationships in data, develop research models and compare approaches. Results still need to be understood in the context of data quality, model assumptions and suitable conditions, and remain subject to evaluation.
How should backtesting results be understood?
A backtest simulates historical conditions using particular data and assumptions. Its evaluation needs to consider data bias, transaction costs, parameter choices and out-of-sample performance. Historical results do not guarantee future performance.
How does quantitative research relate to asset management?
Quantitative research can provide tools for allocation, portfolio observation and risk analysis. Asset management also needs to consider objectives, time horizons, liquidity and other specific conditions; a single model or measure cannot capture all of these.
Is blockchain development the same as cryptoasset trading?
They are different activities. Blockchain development involves system architecture, smart contracts, data records and software integration. Its technical uses depend on the requirements of a particular project.
When is blockchain worth considering?
Blockchain may be worth exploring when multiple parties need shared records, consistent state or a traceable history of actions. Technology selection should also consider conventional databases, alongside access permissions, privacy, processing speed and maintenance costs.
Can I open an account or make transactions on this website?
This website introduces the company and its business focus. It does not provide account opening, deposits, trading or investment subscription functions.
What information helps frame a research or technology discussion?
Useful starting information includes the business problem, intended users, current workflow, available data and the area that needs improvement. System constraints, an indicative timeline and criteria for judging whether an outcome is useful help establish a clear scope.
06 / COMPANY INFORMATION
Established in Hong Kong.
Focused on research and technology.
Jinheng Quantitative
Asset Management Limited
金恆量化資產管理有限公司
- Company address
- Room 511, 5/F, Ming Sang Industrial Building
19–21 Hing Yip Street, Kwun Tong
Kowloon, Hong Kong - Business Registration No.
- 77789843
- D-U-N-S® Number
- 765316217
- Date of incorporation
- 4 March 2025