> ## Content Index
> Fetch the complete content index at: https://www.incomestead.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# Strategy Allocation: Build the Incomestead Stack, Then Actually Govern It
- URL: https://www.incomestead.com/blog/strategy-allocation-incomestead-stack/
- Published: 2026-07-11T06:36:25.000Z
- Updated: 2026-08-06T13:19:42.000Z
- Description: I allocate by strategy, not just by asset, then measure the drift weekly. The Incomestead Stack — why an unmeasured target is a wish, and how to govern it.
- Author: Stefano Starkel
- Tags: The Stack, Governance, #stage-1, #pillar

**A diversified pile of assets isn't a portfolio, and a target you set once and never measure against is only a wish.** Strategy allocation means dividing your capital across *strategies* — not just stocks and bonds — to deliberate target weights, and then governing the gap between those targets and where your book actually sits. That gap has a name: [allocation drift](https://www.incomestead.com/blog/allocation-drift/). It grows silently, it compounds, and unlike a vague feeling that you're "probably fine," it's a number you can compute from your own account every week.

TL;DR

- Diversifying *assets* isn't enough — you diversify **strategies** to target weights. That's the **Incomestead Stack**: fill the base before you climb.
- The target is the easy part. The risk is **allocation drift** — the measurable gap between your plan and your book, which grows even when you do nothing.
- Govern drift against **tolerance bands**, not a razor-thin number, and act only when a Layer stays out of band for **two consecutive weeks** — the same persistence test that separates a signal from noise.
- Allocation is one of the two dials worth governing weekly; **margin** is the other — and margin comes first. Both roll into one **Governance Score**.

On this page

- [Why asset diversification isn't enough](#asset-diversification-is-the-answer-to-a-smaller-question)
- [What is the Incomestead Stack?](#what-is-the-incomestead-stack)
- [The target is easy — the drift is the risk](#the-target-is-easy-the-drift-is-the-risk)
- [A worked example: the winner that ate the plan](#a-worked-example-the-winner-that-ate-the-plan)
- [How to govern drift without micromanaging](#how-do-you-govern-drift-without-micromanaging-it)
- [Strategy allocation FAQ](#frequently-asked-questions)

I write this as a practitioner, not a licensed adviser. I run a leveraged, multi-strategy portfolio myself — a one-man family office — and for years my "allocation" was a diagram I'd drawn once and a comfortable belief that I was diversified. What I actually had was forty-odd positions and no idea, week to week, what shape they added up to. My winners had quietly grown into an overweight I never chose. I built a system to tell me the truth about my own book before the market did. This is the part of that system that governs allocation.

## Asset diversification is the answer to a smaller question

[Harry Markowitz won a Nobel Prize](https://www.nobelprize.org/prizes/economic-sciences/1990/markowitz/facts/?ref=incomestead.com) for the insight that a portfolio's risk isn't the sum of its holdings' risks — it's a product of how they move *together*. Diversification, in his framing, is often described as the closest thing to a free lunch in investing: combine assets that don't rise and fall in lockstep and you lower risk without necessarily lowering expected return. That's true, and it's foundational. But "own more than one thing" is where most self-directed investors stop, and it answers a smaller question than the one that actually governs your outcomes.

The larger question is about *strategies* and *time horizons*, not tickers. A book can hold thirty different equities and still be one bet — all long-duration growth, all correlated, all exposed to the same shock. Owning a growth fund, a dividend stock, a gold ETF, and some cash isn't just "four assets"; it's four *jobs*: compounding, income, insurance, and dry powder. When you allocate by job, "am I diversified?" becomes a question you can answer precisely instead of reassure yourself about. It's the same reason [a pile of good stocks isn't a portfolio](https://www.incomestead.com/blog/a-pile-of-good-stocks-isnt-a-portfolio/): eight excellent names that all do one job are a single bet in disguise.

How much does this layer matter? The most-cited study on the subject, [Brinson, Hood and Beebower's 1986 analysis](https://doi.org/10.2469/faj.v42.n4.39?ref=incomestead.com) of large pension plans, found that allocation policy explained about 90% of the *variability* of returns over time. That figure is routinely mangled into "allocation determines 90% of your returns," which is not what it says and not true. It's a statement about the *variance* of returns — how much your results swing — not their level. The honest reading is narrower but still decisive: the structure of your allocation, far more than which specific names you pick, drives how your portfolio behaves. That's precisely why the structure deserves to be governed rather than assumed.

## What is the Incomestead Stack?

The idea of a wealth pyramid — a broad base of stable holdings supporting progressively higher-risk tiers above it — is a familiar one; the Oxford Club popularized a widely-read version, and family offices have run the shape for generations. I use my own version of it, and because our model isn't a pyramid and isn't theirs, it has its own name: the **Incomestead Stack**. The rule of the Stack is simple and old: *fill the base before you climb.* You don't add a speculative tier until the foundational ones are built, because the base is what lets you survive long enough for the upper tiers to pay off.

LAYERJOB

Cryptoasymmetric upside

Short-Horizon Growthtactical compounding

Long-Horizon Growthpatient compounding

Income Enginecash it pays you

Precious Metalsinsurance

Pillarsdurable blue-chip

Bedrockthe foundation

The Incomestead Stack — Layers by job, foundation at the base. Cash sits alongside as dry powder. Weights are yours to set; the Stack is the structure, not a prescription.

Each strategy is a **Layer** of the Stack, and each Layer has a job. **Bedrock** is the foundation — the broad, boring, low-cost core that most of the book rests on. **Pillars** are the durable, individually-chosen blue chips you'd hold through a decade. The **Income Engine** is the Layer that pays you cash — dividends and realised options premium — the "gets you paid while you wait" layer, provided you [count its income honestly](https://www.incomestead.com/blog/realized-vs-unrealized-options-income/). **Long-Horizon** and **Short-Horizon Growth** split compounding by patience: one you'll leave alone for years, one you manage more actively. **Precious Metals** is insurance that tends to zag when equities zig. **Crypto** is the small, capped, asymmetric-upside tier at the top — the one that only belongs there once everything below it is built. **Cash** sits alongside as dry powder and buffer.

I'm deliberately not printing "correct" percentages here, because there isn't a universal one — the right weights depend on your circumstances, and handing you a number would be exactly the personalized prescription this isn't. What matters for governance is that you *write down* a target weight for each Layer. A Stack with named Layers and no target weights is a nice diagram. A Stack with target weights is a policy you can be held to.

## The target is easy. The drift is the risk.

Here's the part almost nobody governs. The day you set your targets, your book matches them. Then the market goes to work. Your growth Layer has a good quarter and swells past its share. Your Bedrock, by simple arithmetic, shrinks as a percentage even though you never sold a share. A position you loved doubles and quietly becomes a tenth of everything you own. None of this required a decision. It's the default behaviour of a portfolio left alone — and it moves you off the risk profile you actually chose, toward one you didn't.

**Allocation drift is the measurable distance between your target Stack and your real one.** It's the gap between the paper plan and the live book. And "measurable" is the whole point: it isn't a vibe or a quarterly guess, it's a per-Layer number — this Layer is 6% over target, that one is 4% under — that you can compute from your own positions in a few minutes. The trouble is that computing it by hand across forty positions, honestly, every week, is exactly the chore a busy and capable investor stops doing the moment life gets loud — which quietly makes you the single point of failure in your own book. So the drift accrues unwatched, and you find out what shape your book is really in at the worst possible time — when one concentrated winner turns and takes a year of gains with it.

## A worked example: the winner that ate the plan

Take an illustrative $1,000,000 book. Say your written target for the Long-Horizon Growth Layer is 25% — $250,000 — and inside it sits a position you're right about. Over a year that one growth Layer runs 60%, to roughly $400,000, while the rest of your book slips about 8% — from $750,000 to around $690,000\. Your total is still up, to about $1,090,000, a respectable 9%, so on the surface nothing feels wrong. But the Layer is now **36.7% of everything**, more than eleven points over its 25% target. You didn't add a dollar to it. You didn't make a decision. Being *right* did this to you.

The ungoverned investor experiences none of this as it happens — the account balance is up, which feels like nothing but good news, so nothing prompts a look. The gap only becomes visible after the turn, when the Layer that grew to a third of the book gives it all back and he realises he'd been carrying a concentration bet he never chose — [when one winner becomes your whole portfolio](https://www.incomestead.com/blog/concentrated-position-risk/). The governed investor sees the Layer cross its band in week one, sees it *stay* across in week two, and trims from a position of choice while it's still a gain and still his call. Same book, same winner — the difference is entirely whether anyone was measuring the drift.

> You didn't add a dollar to it. You didn't make a decision. Being *right* did this to you — and only a measured drift number would have shown it before the turn. 

## How do you govern drift without micromanaging it?

Governing allocation is not rebalancing on a nervous whim, and it's not chasing your book back to the exact target every week — that just churns costs and taxes for noise. It's a small, repeatable discipline built on three ideas.

**Bands, not points.** Give every Layer a tolerance band around its target — a floor and a ceiling — not a single number. Inside the band, you do nothing; the book is allowed to breathe. It's only when a Layer crosses its band that a decision is even on the table. This is what turns "am I perfectly on target?" (an anxious, unanswerable question) into "is any Layer out of band?" (a calm, binary one).

**Persistence, not a single reading.** A Layer that pops out of band because one holding had a big week is often just noise. A Layer that sits out of band for two consecutive weekly checks is a trend. So I don't act on the first breach — I use the same **two-week persistence rule** I apply to margin: a breach that survives two consecutive weekly reads is a real signal worth acting on; a one-week blip usually isn't. It's the discipline that keeps you from trading against your own noise.

**Rebalance to the edge, not the bullseye.** When a Layer does earn an action, the tidy move isn't to drag it all the way back to the exact target — it's to bring it to the near edge of where it belongs. Trim an overweight Layer back toward the midpoint between its target and its upper band; top up an underweight one toward the midpoint between its target and its lower band. You end up comfortably inside the band with the smallest necessary transaction, rather than making a large, precise, expensive move that the next week's market will undo anyway.

An honest aside, because governance means surfacing inconvenient truths rather than smoothing them over: when you write down your own targets, *add them up.* It is startlingly common for a set of hand-built Layer targets to sum to something other than 100% — a few points of double-counting here, an aspirational overweight there. A real governor flags that your targets are internally inconsistent instead of silently "normalizing" them behind your back, because the inconsistency is information about your plan, not a rounding error to hide. The number should be your decision, made in the open.

## Allocation is one dial. Margin is the other — and it comes first.

Drift is one of exactly two things worth governing on your book every single week. The other is [how much leverage you're carrying](https://www.incomestead.com/blog/using-margin-safely/) — your margin utilization, held inside the Clear / Harvest / Freeze / Forced zones. They're the same discipline pointed at two different risks: measure it, put it in a band or a zone, act only on a persisted breach.

When both need attention at once, the order isn't a matter of taste. **Margin comes first, always**, because a margin problem is existential — a book near forced-liquidation can be sold out from under you by your broker (under [FINRA's margin rules](https://www.finra.org/rules-guidance/key-topics/margin-accounts?ref=incomestead.com), without a call, at its choice of what to sell) before any allocation question matters — while drift is a risk that unfolds over quarters. Fix the thing that can end the game this week, then fix the thing that erodes you over the year. That precedence is built into how I govern, not left to whichever anxiety is loudest on a given morning.

Both dials are the two halves of [portfolio governance](https://www.incomestead.com/blog/portfolio-governance/) — the standing layer above your strategy — and they roll up into one number I watch: the [Governance Score](https://www.incomestead.com/blog/governance-score/), a single 0–100 read of how well you're *governing*, not how good your picks are. Allocation discipline is one of its components; margin safety is the gate that caps it, because no amount of tidy allocation redeems a book that's one bad week from a forced sale. A brilliant portfolio left ungoverned scores low; a modest one governed well scores high. That's the honest priority, made into a number.

## What governing your allocation actually looks like

Put together, it's not complicated — it's just relentless in a way a human doing it by hand rarely manages:

- **A written Stack** — named strategy Layers with target weights, decided calmly in advance, that add up.
- **A weekly read** of your real per-Layer weights against those targets, so drift is caught while it's small.
- **Tolerance bands**, so the book can breathe and only genuine breaches raise a flag.
- **Persistence tracking**, so a one-week wobble isn't confused with a two-week trend.
- **Edge-not-bullseye rebalancing** when a breach persists — the smallest move that brings a Layer home.

The drift check — run it on your own book this week

1. **Sort into Layers.** Group every position into its Stack Layer (Bedrock, Pillars, Income Engine, Long-/Short-Horizon Growth, Precious Metals, Crypto, Cash).
2. **Compute each share.** Layer value ÷ total book = its actual weight. Subtract your written target = that Layer's drift, in points.
3. **Flag the bands.** Any Layer outside its tolerance band is a breach. Everything inside the band: leave it alone.
4. **Persist, then trim to the edge.** Act only if a Layer stays out of band two weeks running — and move it just inside the band, not to the exact target.

No written targets yet? That's the first gap to close — a target you never measure against is only a wish.

I built [Incomestead](https://www.incomestead.com/) because I wanted this to happen to my own book without my having to remember to do it. Every week it reads my positions, sorts them into their Layers, measures each Layer's drift against its target and band, tells me whether a breach has *persisted*, and — when one has — lays out the specific buy and sell amounts to bring the book back inside its bands. It never trades. It never touches the money. It reports what's true and what needs deciding, and then it stops.

Where to go next on the Stack — in this order

1. [Why eight excellent stocks still may not be a portfolio](https://www.incomestead.com/blog/a-pile-of-good-stocks-isnt-a-portfolio/) — the premise underneath the Stack. A pile of good picks and a designed structure are different objects, and the difference is not quality.
2. [The test for whether your book is one bet in disguise](https://www.incomestead.com/blog/am-i-diversified-enough/) — owning a lot of things feels like diversification and frequently is not. This is how to tell, on your own holdings.
3. [What happens when one winner quietly becomes the book](https://www.incomestead.com/blog/concentrated-position-risk/) — concentration does not announce itself — it compounds. Here is how to measure it before it is the only position that matters.
4. [And the number you can compute every week](https://www.incomestead.com/blog/allocation-drift/) — drift moves your risk off the level you chose, while you do nothing, and it is not a feeling: it is arithmetic you can run on your own book. The exact method, and the rule for when it is worth acting on.

## Frequently asked questions

### What is strategy allocation, and how is it different from asset allocation?

Asset allocation divides your capital across asset classes — stocks, bonds, cash. Strategy allocation goes a layer deeper and divides it across *strategies with distinct jobs*: a compounding core, an income Layer, an insurance Layer, a speculative tier, and so on. Two portfolios can have identical stock-versus-bond splits and behave completely differently because their strategies and time horizons differ. Allocating by strategy is what makes "am I really diversified?" a question you can answer precisely.

### What is the Oxford Wealth Pyramid?

It's a popular framework, widely associated with the Oxford Club, that arranges a portfolio in tiers — a broad, stable base supporting progressively smaller and higher-risk layers above it — with the rule that you build the base before adding the speculative top. The Incomestead Stack is my own strategy-Layer version of that same old shape; I cite the pyramid only as the familiar reference, not as a model I'm reproducing.

### What is allocation drift and why does it matter?

Allocation drift is the measurable gap between your target weights and your portfolio's actual weights. It grows on its own as some holdings outperform others, moving you off the risk profile you chose without any decision on your part. It matters because unwatched drift quietly turns a diversified plan into a concentrated bet — you can be "diversified" on paper and dangerously overweight one winner in reality. It is the concrete form of the question [am I diversified enough?](https://www.incomestead.com/blog/am-i-diversified-enough/#tool-am-i-diversified-enough)

### How often should I rebalance my portfolio?

Check on a weekly cadence, but act rarely. The point of checking often isn't to trade often — it's to catch a genuine breach early. Use tolerance bands so the book can drift a little without action, and only rebalance a Layer that stays out of its band for two consecutive weeks. When you do act, move it to the near edge of its band rather than the exact target, which keeps transactions small. This is education about a governance process, not a personalized rebalancing schedule.

### How do I know if my portfolio has drifted off-target?

You compute it: for each strategy Layer, compare its current share of the book to your written target, and flag any Layer outside its tolerance band. If you don't have written target weights, that's the first gap to close — a target you never measure against is only a wish. [The free Allocation Reality Check](https://www.incomestead.com/blog/allocation-drift/#tool-allocation-reality-check) does this read for you from your own numbers.

*A note on honesty: I'm a practitioner, not a licensed adviser, and this is education about how strategy allocation and drift work — not personalized investment advice, and not a recommendation of any particular allocation for you. The right weights depend on your own circumstances and are your decision. Incomestead is a deterministic governance tool, not an investment adviser; it measures and reports, and never trades or touches your money. And if you want the other side of it, that is public too —* [where I publish the counter-arguments to this whole approach](https://www.incomestead.com/what-we-believe/)*.*

Do this next

Incomestead is still being built — be governed from your first week. The free [Allocation Reality Check](https://www.incomestead.com/blog/allocation-drift/#tool-allocation-reality-check) is already live: it reads your real Layer weights against your target, on your own numbers, in your browser.

[Join the founding cohort →](https://www.incomestead.com/#capture)

> **Incomestead recommends. You decide.**