Putting It All Together
Every framework from Parts I–IV applied to three concrete investor profiles. The synthesis the series has been building toward.
April 8, 2026
Marek Pawlowski
16 min read
Part V of The Mathematics of Diversification — the finale
TL;DR
These are reference portfolios, not personalized advice. They illustrate how the framework’s variables interact — your tax situation, country of residence, employer-sponsored accounts, and personal risk tolerance can shift any weight by 10+ percentage points.
I
The system so far
Across four articles, this series has built a complete mental model for investing. Before we apply it, let’s compress the whole system into a single page.
I
Variance is the enemy of compounding
Geometric returns are always lower than arithmetic ones. The gap scales with volatility. Kelly sizing maximizes geometric growth.
II
Diversification has a ceiling
Unsystematic risk can be diversified away. Systematic risk cannot. Correlation quality matters more than position count.
III
Asset classes have personalities
Real estate is efficient. Stocks are the benchmark. Tech is levered stocks. Bitcoin pays a huge variance tax.
IV
Entry price determines return
Valuation at purchase is one of the largest predictors of forward returns. The premium for buying cheap is real and durable.
V
Time changes the game
Long horizons collapse variance into near-certainty. Short horizons amplify it. Sequence risk makes the early years of withdrawal matter most.
VI
Selling is appropriate
When the horizon shortens, when valuations are extreme, when the thesis breaks, or when you need to fund your life. Almost never because of a crash.
Now the question the series has been implicitly building toward: what do I actually do with this? The answer depends entirely on who you are. Three profiles below cover the most common cases, with specific allocations and rationale drawn from every prior part.
II
Profile one: the young accumulator
Age 25–35, 30+ year horizon
Accumulation phase · No withdrawals · High human capital
| Broad equity index (VT / VWCE) | 70% |
| Real estate (REITs or primary residence) | 10% |
| Bitcoin | 8% |
| Gold | 5% |
| Commodities (broad) | 2% |
| Cash / emergency fund | 5% |
Part I — variance is your friend here. Over 30+ years, the variance of a broad equity index collapses into a near-guaranteed positive geometric return.
Part II — concentration is fine. 75% in a single asset class sounds dangerous until you remember that a broad global equity index is already diversified across thousands of companies, sectors, and geographies.
Part IV — no sequence risk, no glidepath concerns. You’re not withdrawing. A 40% drawdown at age 30 is an opportunity to buy more, not a crisis. Worst-case: a 35–50% peak-to-trough drawdown is historically plausible and recovery can take 5+ years — survivable only if you do not abandon the strategy mid-drawdown.
The 8% Bitcoin allocation is Kelly-bounded sizing — small enough that a total loss is survivable, large enough to matter on the right tail. The 5% gold sleeve is regime-change insurance: it does work precisely when stocks and bonds fall together (2022 inflation shock, 2024–25 geopolitical surge). Commodities at 2% is a token inflation hedge — its long-run Sharpe is weak, so the position earns its row only on the regime-diversification argument. Part III applies here too: at extreme equity valuations (CAPE > 35), this profile tilts contributions toward real estate and gold rather than chasing the index. The ‘go to 25%’ BTC camp underestimates how many Bitcoin paths still end in −90% drawdowns lasting 3+ years; Kelly punishes you for oversizing volatility you cannot see in advance.
III
Profile two: the mid-career builder
Age 40–55, 15–25 year horizon
Peak earning years · Family obligations · Growing portfolio
| Broad equity index | 50% |
| Real estate — primary residence equity (~13%) + REIT or rental (~5%) | 18% |
| Treasury bonds (10Y or duration-matched) | 8% |
| Gold | 7% |
| Bitcoin | 4% |
| Commodities (broad) | 3% |
| Cash / short-term needs | 10% |
Part I — variance still works for you, but less. You have 15–25 years, not 40. Kelly sizing adjusts downward.
Part II — real correlation matters now. Real estate, gold, and Treasuries are the three real diversifiers in this profile. Real estate (~0.3 correlation to equities) handles the normal regime; gold (~0.1 correlation and a different driver) handles the regime change scenario; Treasuries (~−0.15 in growth scares) handle the recession scenario. Each one earns its slot for a different reason.
Mid-career diversifiers: each covers a different regime
| Asset | Correlation to equities | Regime it handles |
|---|---|---|
| Real estate | ~0.3 | Normal regime |
| Gold | ~0.1 | Regime change / inflation shock |
| Treasuries | ~-0.15 (in growth scares) | Recession scenario |
Part IV — sequence risk is starting to matter. The 8% Treasuries and 10% cash are sequence-risk insurance that gets deployed to buy equities after a crash. Bonds are duration-matched to the horizon — see Part IV on why that neutralizes most rate risk. Worst-case drawdown for this profile: 20–35% peak-to-trough; recovery 3–5 years.
Bitcoin trimmed from 8% to 4% — not because the thesis is weaker, but because your ruin tolerance has narrowed. Gold at 7% steps up from the young profile’s 5%: the older you get, the more you value an asset that earns its keep when the equity-and-bond pair fails together. Commodities stays small (3%) — its standalone Sharpe doesn’t justify more.
IV
Profile three: the near-retiree / early retiree
Age 60+, withdrawing or about to
Capital preservation · Sequence risk · 20–30 year remaining horizon
| Broad equity index | 35% |
| Treasury bonds (laddered, duration-matched) | 22% |
| Real estate (primary + yield-generating) | 18% |
| Cash & T-bills (2–3 years of expenses) | 13% |
| Gold | 8% |
| Commodities (broad) | 2% |
| Bitcoin | 2% |
Part I — variance is now the primary enemy. You need the money soon. Every percentage point of volatility translates into geometric drag at exactly the moment you can least afford it.
Part II — diversification’s most important job is tail protection. The 22% Treasuries, 13% cash & T-bills, and 8% gold exist so that no single regime can force you to sell stocks at the bottom to fund living expenses. Treasuries cover the growth-scare regime; gold covers the regime where Treasuries and equities fall together (2022 was the warning shot); cash covers the immediate withdrawal need.
Part IV — sequence risk is the whole game. The 2–3 years of cash is not an emergency fund — it is a sequence-risk buffer. This buffer is the structural lever sequence-risk research consistently identifies — the difference between Trinity-style 95% success scenarios and forced selling at the bottom. Treasuries are laddered and duration-matchedto the next 5–10 years of expected spending — Part IV’s point about duration neutralizing rate risk in practice. Worst-case drawdown for this profile: 12–22%; deep stagflation can still extend recovery beyond 5 years, which is exactly why the 8% gold is no longer optional.
Bitcoin at 2% — essentially symbolic. At this life stage, a 90% Bitcoin drawdown should not change your retirement outcome. Commodities at 2% follows the same logic: a token position in case of a 1970s-style supply-driven inflation regime, sized so its weak Sharpe doesn’t hurt you in normal times.
V
The pattern across profiles
Compare the three allocations side by side and a clear pattern emerges. As the horizon shortens and withdrawals approach, equity exposure declines, defensive assets grow, and the allocation to high-variance bets shrinks sharply.
Glidepath: how each sleeve evolves across life stages
Equities and Bitcoin taper; bonds, cash, and gold step up
Young accumulator (25-35) allocation
Equities-heavy; small Kelly-bounded Bitcoin sleeve
Mid-career builder (40-55) allocation
Real diversifiers appear; Bitcoin trimmed as ruin tolerance narrows
Near-retiree (60+) allocation
Defensive assets dominate; Bitcoin essentially symbolic
Allocation evolution by life stage
| Asset class | Young (25–35) | Mid-career (40–55) | Near-retiree (60+) |
|---|---|---|---|
| Equities | 70% | 50% | 35% |
| Real estate | 10% | 18% | 18% |
| Treasury bonds | 0% | 8% | 22% |
| Gold | 5% | 7% | 8% |
| Bitcoin | 8% | 4% | 2% |
| Commodities | 2% | 3% | 2% |
| Cash & T-bills | 5% | 10% | 13% |
What’s striking is what doesn’t change. Real estate stays roughly constant in the 10–18% band. Gold moves modestly upward (5% → 8%) as ruin tolerance narrows — the older you are, the more you value an asset that does well when stocks and bonds fail together. Cash grows. The big shifts are between equities and Treasuries — and in the steady tapering of the Bitcoin allocation. Commodities stays a small token across all three profiles.
Good portfolio construction is not a secret recipe. It is a thermostat — responsive to the same few inputs, adjusting steadily as your life changes.
VI
What's not in any of these portfolios
Individual stock picks.Concentrating in individual names is an expressive choice, not a mathematical one. It belongs in a separate “play money” bucket — sized outside the 100% above and capped (e.g., <5% of net worth) so a total loss does not affect your long-term outcome.
Sector bets and thematic funds. No dedicated tech allocation, no AI fund, no clean energy theme. The broad equity index already holds them in market-cap weights.
Private equity, hedge funds, structured products. No PE, no hedge funds, no actively managed crypto funds, no structured notes. Each one fails the “earns its row’’ test from Part II — either the correlation overlap is too high (most hedge funds are levered equity beta), or the fee load eats the diversification premium (PE), or the downside is non-linear in ways the math can’t cover (structured notes). Gold and commodities do earn their rows on regime-diversification grounds and are now in all three profiles in modest size — what kept them out of older versions of this framework was the simplicity bias, not the math.
Market timing overlays. Part III showed valuation matters, but Part IV showed timing is hard. Rebalancing is the timing you need: review annually or when any sleeve drifts more than 5 percentage points from target. Rebalance with new contributions first (no tax cost), then in tax-advantaged accounts, then realize gains only as a last resort.
VII
Common questions
What allocation should a 25–35 year old hold?
70% broad equity index (VT / VWCE), 10% real estate, 8% Bitcoin, 5% gold, 2% commodities, 5% cash. The long horizon lets variance work in your favor; Kelly-bounded Bitcoin sizing keeps a total loss survivable; the small gold sleeve is regime-change insurance for the rare years when stocks and bonds fall together.
How much Bitcoin should a near-retiree hold?
Around 2% — essentially symbolic. At this life stage, a 90% Bitcoin drawdown should not change your retirement outcome, so the position must be small enough that its failure is irrelevant.
How much gold belongs in a portfolio?
Gold allocation and correlations
| Profile | Gold weight | Why size scales up |
|---|---|---|
| Young (25-35) | 5% | Regime-change insurance, small sleeve |
| Mid-career (40-55) | 7% | Ruin tolerance narrowing |
| Near-retiree (60+) | 8% | Tail risk hurts more on short horizon |
5% for a young accumulator, 7% mid-career, 8% near-retiree. Gold earns its slot through low correlation to equities (~0.10), low correlation to Bitcoin (~0.20), and a different return driver (real interest rates and central-bank flows) — it shines precisely when stocks and bonds fail together, e.g., the 2022 inflation shock. Sizes scale up with age because regime-change tail risk hurts more when your horizon is shorter.
Why hold Treasury bonds in 2025 after the 2022 crash?
Treasuries earn their slot in the mid-career and near-retiree profiles for two reasons. First, the 2022 drawdown happened precisely because real yields had been negative — the entry was expensive on Part III’s metric. With 10Y TIPS yields back above 2% in 2024–25, the forward setup is structurally cheap. Second, duration-matching the bonds to your spending horizon neutralizes most rate risk (Part IV). Treasuries are sizeable in the near-retiree profile (22%) because their negative correlation with equities in growth scares is exactly what sequence risk needs; gold sits alongside them at 8% to cover the inflation regime where that negative correlation breaks.
What is a sequence-risk cash buffer?
Two to three years of living expenses in cash and T-bills, held specifically so a major equity drawdown cannot force you to sell stocks at the bottom to fund withdrawals. It is not an emergency fund; it is structural insurance against the early years of retirement.
Why do these portfolios contain no individual stocks?
Concentrating in individual names is an expressive choice, not a mathematical one. It belongs in a separate “play money” bucket sized outside the 100% allocation and capped (e.g., <5% of net worth) so a total loss does not affect your long-term outcome.
Should I add a dedicated tech or AI fund?
No. The broad equity index already holds tech, AI, and clean energy names in market-cap weights. Layering a thematic fund on top is a concentrated bet against the index — an expressive choice, not a mathematical improvement.
The series in one sentence
Investing is the discipline of making decisions today whose correctness depends on time you cannot yet see — and the math gives you the tools to make those decisions as rationally as possible given that uncertainty.
Parts I through IV built the framework: variance as the invisible tax, diversification as the one free lunch that has a ceiling, asset personalities as the inputs to portfolio construction, valuation as the largest predictor of forward returns, and time as the variable that changes everything. Part V shows how these fit together for real people at real points in their lives.
No framework survives contact with the future intact. But the point of a framework is not to predict outcomes — it is to give you a principled way to respond when the unpredictable happens.
The last most dangerous sentence in investing: "I’ll figure out my portfolio later." Later is never more rational than now.
Up Next
Trading 101: Beginner's Guide
Every concept, tool, and skill you need to trade — mapped before you take a single position.
Continue reading
Explore More
Dive deeper into order flow analysis with our platform features or structured learning paths in the Trading Academy.