The Structured Path Through HRP Theory and Practice

From first principles to working implementations

Most analysts encounter Hierarchical Risk Parity as a finished idea — a dendrogram, a weight vector, a chart showing lower drawdowns. Our team starts somewhere different: at the moment a traditional covariance-based optimizer fails, and the question becomes what to do instead. This curriculum walks through that problem honestly, from the mathematical foundations of correlation structure to the mechanics of hierarchical clustering, through to portfolio construction and backtesting. Each section builds on the previous one, so the logic accumulates rather than repeating itself. You will not be handed conclusions. You will be walked through the reasoning, the notation, and the practical decisions that practitioners face at each step. The sequence is deliberate, and skipping ahead is not recommended — the later modules assume the earlier ones.

Start with Theory

Five Modules That Build the Full Picture

Each module addresses one layer of the HRP construction process, in the order a practitioner would actually encounter it when building from scratch.

01

Module One — Risk Fundamentals and Why They Break

We open by examining what goes wrong with mean-variance optimization under realistic conditions. Unstable covariance matrices, estimation error, and concentration risk are introduced not as abstract critiques but as concrete failure modes that motivate everything that follows. This module establishes the vocabulary and the problem statement that the rest of the curriculum solves.
02

Module Two — Correlation Structure and Clustering Algorithms

With the problem defined, this module introduces the tools used to detect structure in a correlation matrix. Hierarchical agglomerative clustering, linkage methods, and dendrogram construction are covered in sequence. The mathematics are shown explicitly, and each algorithmic step is connected to a decision a practitioner would recognize from real allocation work.

03

Module Three — HRP Portfolio Construction in Full Detail

This is the core module. Using the cluster structure identified in Module Two, we walk through the recursive bisection algorithm that assigns weights without inverting the covariance matrix. Every step is derived, not summarized. By the end, a reader with the prerequisites should be able to reconstruct the full procedure from the notation alone.

04

Module Four — Backtesting Methodology and Honest Evaluation

A method is only as credible as the way it is tested. This module covers how to structure a walk-forward backtest for HRP, what statistics to examine, and — critically — what the results cannot tell you. We address look-ahead bias, regime sensitivity, and the difference between in-sample fit and out-of-sample robustness with the same rigor applied to the construction steps.
05

Module Five — Advanced Extensions: HERC and Nested Clustering

The final module extends the base HRP framework to Hierarchical Equal Risk Contribution and nested clustering approaches. These methods address limitations in the standard construction and introduce additional flexibility for multi-asset universes. Prerequisites from all prior modules are assumed, and the notation builds directly on what has already been established.

Learning Formats

Three Ways to Engage With the Material

Our team has structured the HRP curriculum across three complementary formats, because different readers absorb quantitative content differently. Some work best reading through derivations at their own pace. Others need a live session to ask the question that unlocks a concept. A third group learns most effectively by working alongside peers on a shared analytical problem. All three formats cover the same rigorous content — the difference is how the material is encountered and reinforced over time.

01

Self-Paced Video Lectures

Learn at your own schedule
The video lecture series follows the five-module sequence exactly. Each session presents one concept — a clustering algorithm, a bisection step, a backtesting decision — and works through it with annotated notation and numerical examples. Our team deliberately keeps each session focused on a single idea rather than surveying a broad topic, because that is how quantitative understanding actually accumulates. Viewers can pause, revisit, and cross-reference the written materials at any point. The lectures assume the prerequisites listed in the preparation guide, and they do not skip the mathematical steps that most introductory treatments leave out. Supplementary notebooks accompany each session for readers who want to trace the computation directly.
  • Modular sessions aligned to the five-module curriculum structure for clear progression
  • Full derivations shown on screen — no steps summarized away or left as exercises
  • Downloadable companion notebooks trace each numerical example in reproducible form
  • Closed captions available in English and Arabic for accessibility across the region
02

Live Virtual Workshops

Interactive sessions with the research team

Workshops run as small cohorts on a scheduled basis, led directly by members of our quantitative research team. The format is deliberately conversational — participants are expected to have worked through the relevant module materials in advance, so the session can focus on questions, edge cases, and the kind of practitioner judgment that does not fit neatly into a written derivation. Each workshop covers one module in depth over a structured session, with time reserved for open discussion. Our team has found that the most productive workshop moments come when participants bring a specific point of confusion or a real analytical context where the theory does not immediately apply. That friction is not a problem to manage — it is the most productive part of the session.

  • Small cohort format ensures every participant can ask questions directly during the session
  • Led by named researchers from the team, not by junior facilitators reading from slides
  • Participants are expected to have reviewed module materials before each workshop session
03

Collaborative Research Projects

Apply the framework to a real problem

For readers who want to move beyond following a derivation and into producing one, the collaborative research track pairs participants with a small working group and a defined analytical question. Groups select an asset universe, apply the full HRP construction procedure, and document their methodology, results, and interpretation in a structured research note format. Our team provides structured feedback at two checkpoints — once after the clustering and construction phase, and once after the backtesting and evaluation phase. The goal is not a polished publication but a rigorous analytical document that a practitioner could hand to a risk committee and defend. This format is particularly well suited to analysts who already work in institutional contexts and want to apply HRP to a universe they know.

  • Small working groups of two to four participants per project for focused collaboration
  • Structured feedback from the research team at two defined checkpoints during the project
  • Final deliverable follows a practitioner research note format, not an academic essay structure

Enrollment Tiers for Every Analytical Context

Individual
Contact for access details

Designed for independent analysts and researchers who want to work through the full HRP curriculum at their own pace with access to all written and video materials.

  • Full access to all five curriculum modules
  • Self-paced video lecture series included
  • Downloadable companion notebooks for each module
  • Written derivations and reference materials
  • Access to the annotated bibliography and source list
  • Email support from the research team for technical questions
Get in touch
Popular
Professional
Contact for a quote

Built for working analysts and institutional researchers who want structured workshop access alongside the full self-paced materials and direct team engagement.

    Everything included in the Individual tier
  • Access to the scheduled live virtual workshop series
  • Priority response from the research team on technical queries
  • Module completion certificates for professional records
  • Early access to updated and extended module content
  • Quarterly live Q and A session with the lead research team
  • Access to the practitioner discussion forum
Enquire now

Institutional

Contact to discuss terms
Structured for teams, asset management firms, and academic institutions that want cohort access, collaborative research tracks, and tailored analytical support.
    Everything included in the Professional tier
  • Cohort enrollment for analyst teams of any size
  • Access to the collaborative research project track
    Dedicated research team contact for the institution
  • Custom workshop scheduling to fit institutional calendars
  • Research note feedback at two structured checkpoints
    Institutional license for internal distribution of materials
  • Annual review session with the lead quantitative researcher
Reach out

Preparing to Get the Most From the Curriculum

The HRP curriculum is mathematically demanding. These practical recommendations help readers arrive at each module ready to engage with the content rather than catching up on prerequisites mid-session.

Build Your Mathematical Foundation First

Prerequisites

The curriculum assumes familiarity with linear algebra — matrix operations, eigenvalues, and covariance structures — as well as basic probability and statistics. Readers who find Module One notation unfamiliar are encouraged to spend time with a standard linear algebra reference before continuing. Our team has found that gaps in this foundation are the most common reason readers stall in Module Two, and that a focused review of two to three weeks resolves most of them before they become obstacles in the clustering material.

Review linear algebra Revisit covariance basics Check Python familiarity

Set a Realistic Weekly Study Schedule

Study Planning

The five modules are sequenced to build on each other, and rushing through a module to reach the next one tends to create confusion that compounds rather than resolves. Our team recommends allocating dedicated, uninterrupted time for each session — treating it the way a researcher would treat a reading group, with notes and questions prepared in advance. Readers who work through one module per week, including the companion notebooks, typically find the material accumulates steadily without requiring repeated revision of earlier sections.

Schedule weekly sessions Prepare questions in advance Work through notebooks

Engage With the Notation, Not Around It

Study Method

A common instinct when encountering unfamiliar mathematical notation is to read past it and focus on the surrounding prose explanation. In this curriculum, that approach will not work — the notation carries meaning that the prose cannot fully substitute for. Our team recommends writing out each derivation by hand at least once, even if the steps feel mechanical. That process of transcription forces engagement with each symbol and its role in the argument, and it reliably surfaces the specific points of confusion that need to be resolved before the next step makes sense.

Transcribe key derivations Annotate each symbol Flag confusion points

Use the Companion Notebooks Actively

Practical Work

Each module is accompanied by a computational notebook that traces the numerical steps of the derivation in executable form. These are not supplementary extras — they are a core part of how the curriculum is designed to work. Running the code, modifying parameters, and observing how outputs change is one of the most direct ways to build genuine intuition for how clustering algorithms and bisection procedures behave across different asset universes. Readers who treat the notebooks as optional consistently report weaker intuition in the backtesting module, where that computational familiarity becomes essential.

Run each notebook fully Modify sample parameters Compare output changes

Bring a Real Asset Universe to the Material

Application

Abstract familiarity with HRP construction is significantly reinforced when a reader applies each step to an asset universe they already know — a set of regional equity indices, a multi-asset class mix familiar from their analytical work, or a historical dataset they have used before. This does not require access to proprietary data. Publicly available series are entirely sufficient. What matters is that the clustering results and weight assignments connect to a context where the reader can evaluate whether the output makes analytical sense, rather than accepting it as a correct answer to an unfamiliar problem.

Select a familiar universe Apply clustering steps Interpret weight outputs
Read our research background

Ready to Work Through the HRP Framework With Our Team?

The curriculum is designed for analysts who want to understand Hierarchical Risk Parity at the level of its construction, not just its conclusions. Our team has structured every module to build genuine analytical understanding, one rigorous step at a time. Whether you are an independent researcher or part of an institutional analytical team, the material meets you where you are and takes you through the full procedure without shortcuts.
1

Review the Prerequisites and Confirm Your Readiness

Check the preparation guide against your current familiarity with linear algebra, Python, and portfolio theory. A short self-assessment helps identify any gaps worth addressing before Module One. Our team is available to answer questions about whether the curriculum is the right fit for your current level.

2

Choose the Enrollment Tier That Fits Your Context

Individual, Professional, and Institutional tiers offer different levels of access and support. Review what each tier includes and reach out to our team with any questions about which structure suits your analytical goals and working schedule before enrolling.

3

Contact Our Team to Complete Enrollment

Enrollment is handled directly through our team rather than through an automated system, because we want to confirm that each participant has the prerequisites in place and understands what the curriculum covers. Reach out through the contact page and we will guide you through the next steps.

4

Begin Module One at Your Own Pace

Once enrolled, access to the materials is immediate. Our team recommends starting with the Module One reading before the first video session, so the notation is already partially familiar when the derivation begins. The companion notebook for Module One can be run in parallel from the first session.

Contact Our Team to Enroll Today

Our team reviews every enrollment inquiry personally and responds within two business days. We will confirm your prerequisites, answer questions about the module structure, and guide you through the enrollment process for the tier that best fits your analytical context and goals. There is no automated queue — your message reaches the research team directly.

No automated enrollment process
Direct access to research team
English and Arabic support
Two-day response commitment