Human Capital ETF · Qiaomai Liu

Part III · Portfolio Construction and Rebalancing

From Evidence to Allocation Policy

Fourth edition ·

Preface.Begin with Evidence

A framework becomes an operating system only when it changes a real allocation. Begin with evidence, not an ideal identity: the last four weeks of calendars, recovery, spending, work, learning, output, and unfinished commitments.

One page is enough for the first cycle. Do not begin by inventing a perfect percentage for each position. Begin by seeing where resources already went, what persisted, and what risk is concentrated in one place.

The first rebalance begins with an honest inventory, not a new ambition.

Section 1.Inventory the Allocation Ledger

Use the same ledger as Part I. Record what exists before deciding what should exist.

Allocable Inputs
Time, usable energy, attention, and allocable cash observed over the last four weeks.
Embodied Stocks
Health and capabilities retained in the person.
Enabling Stocks
A financial buffer, tools, and systems under meaningful control.
Externalized Stocks
Controlled artifacts and documented methods that can be reused.
Return Flows
Income, feedback, and opportunities entering the system.
Outcomes
Present resilience, autonomy, and optionality.
Constraints and Exposures
Health debt, financial obligations, family responsibilities, contractual boundaries, and dependence on one employer, skill, platform, credential, country, income source, or identity.

A constraint is not a personal failure. It is a condition the policy must respect. A portfolio that ignores health, debt, caregiving, law, or employment obligations is not ambitious; it is mis-specified.

Do not force return flows, constraints, or exposures into Core, Growth, Distribution, or Meta. The positions classify what an allocation is doing; the ledger classifies what kind of thing is being observed.

What do I control, what only passes through me, and what can be taken away by one decision outside my control?

Section 2.Asset Status and Risk Flags

An activity can be useful without yet producing an asset. Test status first; then assess quality and risk.

Asset status

  1. What persists after the immediate task is complete?
  2. Where does it reside: in the person, in a controlled system, or in someone else's system?
  3. Who owns or meaningfully controls it?
  4. Can it be deployed or used again without being recreated from zero?

Risk and quality flags

Portability
How much value remains outside the present employer, industry, platform, or country?
Maintenance
What recurring time, energy, money, practice, or access keeps it useful?
Decay
How quickly could nonuse, impaired health, or changing tools and institutions reduce its value?
Concentration
Which employer, system, person, credential, or context can impair it?
Evidence of Usefulness
Where has it improved delivery, created a reusable result, or generated verified feedback?

A course may be an input. A capability demonstrated in repeated work may be an embodied stock. A documented tool may be an externalized stock. A paycheck is a return flow; the financial buffer retained from it is an enabling stock. Firm-specific capability can qualify as an asset while still carrying high concentration risk.

Employment and ownership boundary. Employer-confidential information and employer-owned intellectual property do not become personal assets merely because a worker helped create them. Externalize the transferable method, not protected data, code, documents, designs, or customer information.

Relationship boundary. Professional trust, reputation, and reciprocal access can matter to the portfolio, but other people and relationships are not owned assets. Record them as co-maintained social-capital signals, not owned stocks; record the opportunities they produce as return flows.

Section 3.The One-Page Allocation Policy

Human Capital ETF provides no universal ratio. A useful policy is specific to the worker's present stage, binding constraint, obligations, and evidence. Write only one marginal priority for the period.

An allocation policy is meaningful only when it excludes something.

Human Capital ETF · One-Page Policy

Print or copy this template. Complete each line in plain language.

1. Policy period and current stage
Blank line to complete.
2. Binding constraint and Core floor
Blank line to complete.
3. One marginal allocation priority
Blank line to complete.
4. Inputs committed and where they come from
Blank line to complete.
5. Evidence expected by the review date
Blank line to complete.
6. Maintenance, concentration, stop, and exit risks
Blank line to complete.
7. Not pursued during this period
Blank line to complete.
8. Review date
Blank line to complete.

The policy should enter the calendar and budget. A declared priority with no protected time, no resource limit, and no stop rule is a preference, not an allocation. Opportunity cost belongs in the policy: name what will receive less because this priority receives more.

Section 4.Observation and Rebalancing

Rebalancing is not constant change. It separates observation from intervention so that a temporary fluctuation does not become a new life strategy.

Monthly Observation
Record drift, recovery, spending, project evidence, output, and concentrated exposure. Observe without trading every fluctuation.
Quarterly Rebalance
Add, increase, maintain, reduce, migrate, or exit projects and commitments according to accumulated evidence.
Annual Policy Review
Reconsider life stage, health, career direction, family responsibility, constraints, portfolio purpose, and the rules themselves.

Evidence can trigger an earlier review: sustained deterioration in recovery; a financial buffer crossing its minimum threshold; a capability succeeding in repeated delivery; verified Distribution feedback; a change in employment conditions; or firm-specific exposure becoming too large for the worker to survive its loss.

Add
A missing mechanism deserves an initial allocation.
Increase
Stronger evidence justifies more resources.
Maintain
The allocation is functioning and needs no reactive change.
Reduce
Cost, risk, or declining usefulness no longer justifies the present allocation.
Exit
The thesis has failed, the constraint has changed, or the project no longer belongs in the policy.

What changed in the evidence—not merely in my mood?

Section 5.Migration, Conversion, Returns, and Reinvestment

Migration
The same capability changes its primary function. Python may migrate from Growth to Core after repeated professional deployment.
Conversion
Embodied capability produces a distinct externalized stock that the worker meaningfully controls. Core engineering knowledge does not leave Core when it produces an essay or method; it creates another asset.
Return
An asset generates income, feedback, or opportunity. The flow should not be double-counted as the asset that produced it.
Reinvestment
A return is allocated to recovery, a financial buffer, maintenance, or the next capability.

These distinctions preserve the ledger. An essay is not simultaneously the skill that produced it, the reputation it may build, the feedback it receives, and the income that might follow. One causal chain can contain different stocks, flows, functions, and decisions.

Migration changes function. Conversion creates another stock. Returns enter as flows. Reinvestment begins the next allocation.

Section 6.A Worked Engineering Case

In my first year as an engineer, a repeated fastener-replacement workflow was slow and error-prone. The allocation to completing assigned projects served Core: present delivery had to remain reliable. Learning enough Python to automate the pattern began in Growth.

The first script worked. I later rewrote it, packaged it for the company's internal software library, documented its use, and trained colleagues. Their requests exposed new needs and led to another iteration. The loop was simple: problem, tool, distribution, feedback, revision.

The ledger prevents a false claim of ownership. The internal executable, documentation, and protected workflow remained inside the employer's boundary; they were not automatically my personal externalized stocks. What remained embodied in me was the more portable capability to decompose an engineering problem, automate a repeated process, document a method, teach its use, and learn from feedback.

A non-confidential account of that method can later be converted into a separate Distribution asset. Feedback, trust, and new requests are returns. If Python becomes reliable in repeated professional delivery, the capability migrates from Growth toward Core. Meta decides whether that evidence justifies the next allocation.

This case is condensed from the author's account of his first year at work. Product details, code, and employer-confidential information are intentionally excluded.

Section 7.The First Portfolio Cycle

A cycle does not need to advance every position. If Core is impaired, protecting or restoring Core may be the entire policy for that period. Meta exists to make that trade-off explicit.

Start with six decisions.

  1. Review the last four weeks of actual allocation.
  2. Name the binding constraint and the minimum Core floor.
  3. Choose one marginal priority. It may be Core, Growth, Distribution, or Meta.
  4. Assign specific inputs and name their opportunity cost.
  5. Write the evidence, stop condition, and review date before beginning.
  6. Decline one competing activity that does not support this policy.

The first ninety days can then run as one complete cycle.

Month One
Complete the ledger, identify the binding constraint, and write the one-page policy.
Month Two
Protect the Core floor and run the chosen marginal allocation in real conditions.
Month Three
Compare evidence with the policy. Maintain, increase, reduce, exit, or migrate only what the record supports.

Do not attempt to design an entire life in ninety days. The purpose is to make Human Capital ETF operate once, leave a record, and improve the next allocation.

The goal is not to optimize every hour. It is to make the next allocation more deliberate and the future less fragile.