EPT + HiFuPi Asset Model
Research Engine
Expanded Portfolio Theory with the HiFuPi asset model
EPT is the computational framework. HiFuPi is the asset model. Together they provide a research platform for active ETF design, backtesting, and portfolio exploration.
Expanded Portfolio Theory
Modern Portfolio Theory is powerful, but it comes with an embedded set of assumptions. Those assumptions shape the type of portfolios the model is naturally built to solve. Expanded Portfolio Theory starts from a broader computational design space.
The central idea is simple:
When the assumptions embedded in Modern Portfolio Theory are imposed on EPT, the solution is exactly the same. When those assumptions are not imposed, EPT can solve portfolio problems across a much wider range of investment situations.
That makes EPT useful for active ETF design because real portfolio problems often include features that are difficult to express cleanly in a traditional optimization setup: holding limits, turnover limits, tax-aware constraints, liquidity screens, path rules, leverage limits, concentration rules, universe changes, and conditional reallocation logic.
Why EPT matters for active ETFs
Broader constraint handling
EPT can be used to explore portfolios when the design problem includes practical constraints that are central to ETF construction.
Backtest-ready structure
The framework can be expressed as repeated time-slice decisions, making it naturally suited for historical testing and transparent portfolio review.
MPT-compatible when needed
EPT will collapse back into the traditional MPT result when MPT assumptions are intentionally imposed.
HiFuPi Asset Model
HiFuPi is the asset model supporting the active ETF design process. It is intended to help organize the investment universe, historical data, portfolio attributes, and decision inputs needed for repeated portfolio construction.
What HiFuPi & EPT supports
- Asset universe definition
- Any periodic reallocation testing
- Liquidity and tradability contraints
- Backtest input preparation
- Portfolio weights and turnover review
- Scenario comparisons across ETF concepts
Design objective
Create a structured bridge between raw market data, computational portfolio methods, and explainable active ETF design.
Research positioning
The 5thN approach is not simply to choose a benchmark and make small human adjustments around it. The objective is to build a repeatable research process that can define an active ETF concept, test the rules, review the historical behavior, and provide transparent supporting analytics.
Any historical or backtested result should be reviewed carefully. Backtested results are hypothetical and may be sensitive to data choices, constraints, universe construction, assumptions, and implementation rules.