Week 15: Multi-Asset Portfolios — Risk Parity, All-Weather, and the Four-Tranche Framework

---

1. Why This Is Important

Week 4 gave you 60/40. Week 6 gave you gold. Week 13 gave you the short side. This week the course finally puts those pieces inside one coherent allocation framework — the same framework Bridgewater has run since 1996, the same framework Cliff Asness debuted at AQR in 2004, and the same shape the four-tranche book is designed to imitate at retail scale.

You need this lesson for four reasons.

60/40 is a regime trade in disguise. It worked for forty years

because stocks and bonds were negatively correlated and inflation was falling. When that regime broke in 2022 — both legs down 18% together — the question stopped being "stocks or bonds?" and started being "what is the portfolio that does not depend on a single macro outcome to survive?" Multi-asset construction is the answer the institutional world settled on, and you should know the architecture before you accept or reject it.

Risk parity reframes the allocation question. Traditional

allocation asks "what fraction of my dollars goes to each asset?" Risk parity asks "what fraction of my risk comes from each asset?" In a 60/40 book at typical vols, roughly 90% of the portfolio's variance comes from the equity sleeve, even though only 60% of the dollars are there. Once you see that, you cannot un-see it — the 40% bond sleeve is a rounding error in risk terms.

**The all-weather four-quadrant model is the cleanest mental

picture of macro you will ever own.** Growth × inflation, two axes, four cells, one winning asset class per cell. It is not a complete forecasting tool. It is something better — a checklist that forces you to ask, before you size any position, *which quadrant is this trade a bet on?* Most retail blow-ups come from running an unhedged single-quadrant book and not knowing it.

It is the bridge into the back half of the course. The

four-tranche frame (growth / income / store-of-value / opportunistic) is what we will populate in Weeks 16-30 with sectors, factors, options strategies, and barbell tilts. Without the multi-asset chassis from this week, those later trades have no home to go in.

This is not a recommendation that you literally run risk parity at home. The leverage and rebalancing required are institutional. It is a recommendation that you understand the principles, then apply the shape — not the leverage — to your own four-tranche book.

---

2. What You Need to Know

2.1 The 2022 Break and Why It Reframed Everything

Walk forward from 1981. Treasury yields fall from 15% to 0.5% over forty years. Every time inflation prints surprised softer, bonds rally. Every time growth prints surprised softer, stocks fall and bonds rally. The stock-bond correlation runs around -0.3 for two decades. 60/40 looks like a perpetual motion machine.

Then 2022 happens. The Fed hikes 525 basis points in fifteen months to fight a 9% CPI print. Long Treasuries (TLT) finish the year down 31%. The S&P 500 finishes down 18%. The Bloomberg Aggregate Bond Index finishes down 13% — its worst calendar year since the index launched in 1976. A 60/40 book loses roughly 17%, the worst real return for the strategy since 1937. The negative correlation that made 60/40 work flipped. It is positive again at the time of writing (April 2026), and the institutional consensus is that as long as inflation is the dominant macro risk, stock-bond correlation will stay positive — the same shock that scares stocks now hurts bonds too.

This is the regime change to watch for. Forty years of falling rates and falling inflation handed passive allocators a tailwind. That tailwind is gone. The portfolios designed for it inherited a hidden vulnerability: they were one-quadrant trades.

2.2 The All-Weather Quadrants — Growth × Inflation

Ray Dalio's framework, formalised at Bridgewater in 1996, organises every macro shock into a 2x2 grid. Two axes:

• Growth surprise — is GDP / earnings growth coming in higher

or lower than the market priced in?

• Inflation surprise — is CPI / wage growth coming in higher or

lower than the market priced in?

The word surprise matters. Markets price expected growth and expected inflation already. What moves portfolios is the residual.

Each quadrant has a winning asset class — the one whose payoff is structurally exposed to that macro residual.

The historical anchors:

• Rising growth, rising inflation (top-left): 1970s, 2003-2007,

2021-22 reflation. Winners: commodities, energy, materials, emerging-market equities, gold late-cycle.

• Rising growth, falling inflation (top-right): 1991-1999,

2013-2019. The "Goldilocks" cell. Winners: developed-market equities, growth stocks, credit.

• Falling growth, rising inflation (bottom-left): 1973-74,

1979-80, 2022. Stagflation. The cell where 60/40 dies. Winners: gold, TIPS, short-duration cash, energy.

• Falling growth, falling inflation (bottom-right): 2001-02,

2008-09, 2020. Deflationary recession. Winners: long-duration Treasuries, USD cash, defensive equity (staples, healthcare).

The Bridgewater claim — and it is empirically defensible — is that if you can hold one asset levered to each cell at *equal risk contribution*, the four streams roughly cancel and you get a portfolio that is no longer betting on any single quadrant.

2.3 Risk Parity — Allocation by Variance, Not Dollars

The technical machinery is one equation. Given $n$ assets with volatilities $\sigma_1, \dots, \sigma_n$, the unlevered risk-parity weight on asset $i$ is

$$ w_i = \frac{1/\sigma_i}{\sum_{j=1}^{n} 1/\sigma_j} $$

That is: each asset gets a dollar weight inversely proportional to its volatility. The high-vol asset (equity at $\sigma \approx 16\%$) gets a small dollar weight; the low-vol asset (long Treasuries at $\sigma \approx 8\%$) gets a large dollar weight; and at the limit, each asset contributes the same variance to the portfolio.

For four assets at typical vols — equity 16%, 10-year Treasury 6%, gold 18%, T-bills 1% — the inverse-vol weights are:

Asset	Vol	1/σ	Raw weight	Risk-parity weight
US equity (SPY)	16%	6.25	0.067	6%
10y Treasury (IEF)	6%	16.67	0.179	18%
Gold (GLD)	18%	5.56	0.060	6%
T-bills (BIL)	1%	100.0	1.072	107% — clipped to 70%

Already a problem. The cash sleeve is so low-vol that pure inverse-vol weighting wants to put more than 100% there. The standard fix at Bridgewater is to drop cash from the unlevered book and *lever the remainder* — typically up to 1.5x to 2x of NAV — so the portfolio hits a target volatility (often 10% per year). The leverage is mostly via Treasury futures, where margin is cheap and basis is tight. The leverage is the price of admission. Without it, an unlevered all-weather book at 60% Treasuries earns only 5-6% per year, which loses to plain 60/40 over most decades.

2.4 The 2022 Risk-Parity Break

For twenty-five years, leveraged risk parity worked. Then 2022 happened the same way to the strategy as to 60/40, but worse — because the leverage amplified the bond loss.

Bridgewater's All-Weather strategy lost roughly 25% in 2022, reportedly its largest calendar drawdown since launch. AQR's Risk Parity Fund (QRPIX) lost roughly 19%. The mechanism is mechanical: leveraged Treasury futures lost 25-30% of notional, and the leverage multiplier turned that into the dominant P&L line. The equity sleeve (small dollar weight, but high vol) lost another chunk. Gold was roughly flat. T-bills earned 2%. There was no quadrant working — inflation was rising, growth was slowing, but the policy response hit bonds harder than the slowdown helped them.

The lesson is not that risk parity is broken. The lesson is that **every diversification scheme that uses leverage to equalise risk contributions assumes the cross-correlations stay where the back-test put them**. When stock-bond correlation flips from -0.3 to +0.5, the diversification math breaks and the leverage that was supposed to help you starts hurting you. This is the same risk that shows up as the volatility tail wagging the dog: the correlation tail wagged the allocation dog.

The institutional response since 2023 has been to add a fifth sleeve — explicit inflation hedges (TIPS, commodities, gold) — at larger weight than the original Bridgewater template, and to hold less leverage in the bond sleeve. The shape is converging on what we call the four-tranche book.

2.5 The Four Tranches at Retail Scale

Strip out the institutional leverage and what remains is a clean four-bucket structure that any retail account can hold using ETFs.

The four tranches:

Growth (40%). US-listed equities. Default: a broad index

(VTI, SPY) plus a quality factor tilt (QUAL) covered in Week 23. This is the engine of long-run real return. It is also the sleeve that loses in the bottom-left and bottom-right quadrants, which is precisely why it is not 60-100% of the book.

Income (30%). Mostly intermediate-duration Treasuries (IEF, 7-10y),

with a small investment-grade credit slice (LQD). This is the sleeve that wins in the bottom-right (deflation) cell and provides the carry that finances the other tranches. After 2022 the institutional consensus is to under-weight duration relative to the historical risk-parity prescription — 30%, not 60%.

Store-of-value (20%). Gold (GLD or IAU) and TIPS (SCHP). The

sleeve that wins in the bottom-left (stagflation) cell. Remember these instruments are belief trades — gold has no coupon, TIPS pay only the inflation print — but the belief has a 1000-year track record and a reliable bid in regime breaks.

Opportunistic (10%). Cash (BIL/SGOV), short-duration T-bills,

and the options-premium / barbell trades from Weeks 25-30. This is the sleeve that lets you add when something gets cheap. Dry powder is not a cost — it is the asymmetric option that funds the other three sleeves' rebalances.

Tranche	Weight	Vehicles	Quadrant served
Growth	40%	VTI, SPY, QQQ, QUAL	Top-right (Goldilocks)
Income	30%	IEF, LQD, GOVT	Bottom-right (deflation)
Store-of-value	20%	GLD, IAU, SCHP	Bottom-left (stagflation)
Opportunistic	10%	BIL, SGOV, options	Top-left + dry powder

This is not a risk-parity book — the dollar weights are higher than the equal-risk-contribution math would prescribe for equity. It is deliberately closer to a 60/40 starting point, *adjusted for the post-2022 regime*. The store-of-value sleeve is materially larger than what Bridgewater's original 1996 template ran (5-15%), and the income sleeve is materially smaller. The shape is the barbell — actual safety on one end, asymmetric upside on the other — applied at the asset-allocation layer.

2.6 Backtest — All-Weather vs 60/40 vs 100% Equity, 1928-2024

Run the four-tranche book back against Damodaran's 1928-2024 annual data, with gold filled at 4% real per year before 1971 (the gold-standard window where the price was fixed) and at the actual London PM-fix annual return after. Annual rebalance. The results:

Metric	100% equity	60/40	All-weather (40/30/20/10)
Geometric annual return (nominal)	9.6%	7.4%	6.7%
Geometric annual real return	6.6%	4.4%	3.7%
Annualised volatility	19.6%	12.0%	9.4%
Sharpe (excess over T-bills)	0.30	0.30	0.32
Max real drawdown	-75% (1929-32)	-53% (1929-32)	-38% (1973-74)
Worst calendar year	-44% (1931)	-28% (1931)	-19% (1981)
Decades with positive real return	8 of 10	9 of 10	10 of 10

The all-weather book gives up about 90 bps of real CAGR versus 60/40 in exchange for two things: every decade is positive in real terms, and the worst real drawdown is 15 percentage points smaller. That trade is worth it for an investor who lives off the portfolio. It is not worth it for a 35-year-old with a 30-year horizon and a paycheque, who should be closer to a barbell shape — heavier equity, smaller but real safety sleeve, larger opportunistic tail.

The interactive below lets you slide the four weights and watch the historical wealth path, max drawdown, Sharpe, and drawdown duration update in real time. Try the institutional 35/40/15/10 template, the 40/30/20/10 retail template, and a barbell at 60/10/20/10 and compare them on the same chart.

2.7 What This Lesson Is Not

A few things this lesson is not telling you to do.

• Not telling you to run leverage. The Bridgewater leverage is

feasible only with futures access, professional risk management, and tolerance for 25%+ drawdowns when the correlations break. The retail four-tranche shape uses no leverage.

• Not telling you to abandon 60/40. For a long-horizon

accumulator, 60/40 is still defensible. The all-weather shape is the right chassis for someone whose horizon is ten years or less, or whose risk budget cannot survive a 50%+ equity drawdown.

• Not telling you to equal-weight quadrants forever. The shape

should tilt with the regime. We anticipate the 40-year passive regime breaking; the four-tranche book is a vehicle that can tilt without abandoning the chassis. Weeks 16-22 will populate the tilts.

---

3. Common Misconceptions

"Risk parity is just 50/50 stocks and bonds." It is not. Risk

parity weights inversely by volatility, which gives a small equity weight and a large bond weight, and then levers the whole thing to a target vol. A 50/50 mix is closer to risk-parity than 60/40 but it is still equity-dominated in variance terms.

"Diversification means owning more stocks." Diversification is

about the cross-correlation between return streams, not the number of tickers. A portfolio of 100 tech stocks is one return stream. A portfolio of one S&P fund + one Treasury fund + gold is three.

"The 60/40 portfolio is dead." It is not dead. It is in a

harder regime than the one it was designed for. In the 1970s 60/40 had a real-return drawdown of -33% and recovered. The regime-vulnerability is real but the portfolio still works.

"All-weather earns more than 60/40." It does not. Over 1928-2024

it earned less in compound terms. Its claim is on lower drawdowns and a smoother return path, not higher returns.

"Gold is dead weight in a portfolio." Gold's contribution is

not its standalone return — it is its covariance with stress. At a 20% allocation it added 80 bps of Sharpe to an all-weather book over 1971-2024 even though it earned less than equity.

"Risk parity protects you from any drawdown." It protects you

from quadrant-specific shocks. It does not protect you from a correlation regime change. 2022 was the latter.

"Cash is a drag." Cash funds the next rebalance. A 10% cash

sleeve that lets you add 5% to equity at a 30% drawdown is not drag — it is the option that pays for itself in the next bear market.

"You need leverage to run all-weather." You need leverage to

run Bridgewater's version of all-weather at 10% target vol. You do not need it to run the unlevered four-tranche shape, which targets 9-10% vol naturally because of how the dollar weights land.

"Sharpe is the same so it does not matter." The Sharpe is

similar; the shape is not. Two portfolios with the same Sharpe but different max drawdowns are not equivalent for a retiree, an endowment, or anyone using options on the equity sleeve for tax efficiency.

"TIPS solved inflation." TIPS solve measured inflation

(the BLS CPI). They do not solve regime-shift inflation (the 1970s) and they do not solve sovereign debasement. Gold is the second leg of the store-of-value tranche for a reason.

---

4. Q&A Section

Q: Should I literally run 40/30/20/10 in my IRA tomorrow? A: As a starting point, yes — it is a defensible default and will not hurt you over a 10-year horizon. But the right long-run weight depends on your horizon and your other income sources. A 30-year-old with a stable paycheque should probably tilt the equity sleeve up toward 60-70%; a 65-year-old drawing 4% per year should probably stay at 40-50% equity. The chassis is what is fixed; the tilts are personal.

Q: How do I rebalance a four-tranche book? A: Once a year, in January, mechanically. If any sleeve is more than 5 percentage points off target, trade it back. Week 11's interactive shows that band rebalancing beats calendar rebalancing slightly on turnover but the gap is small — what matters is doing it at all.

Q: Is risk parity practical for a $50,000 account? A: The dollar-weight shape is. The leverage is not. To get Bridgewater's 10% target vol unlevered you would need 80% bonds, and that earns 5% nominal — not enough. Skip the leverage, run the four-tranche shape unlevered, target 9-10% vol naturally.

Q: Why 20% in store-of-value, not the historical 5%? A: Because we are no longer in the 1996-2020 disinflationary regime. Bridgewater's original 5-7% gold sleeve was sized for a world where inflation was a tail risk, not a base case. April 2026 is not that world. The regime change anchors the chassis adjustment; the bigger gold weight follows from it.

Q: What about international equity? You only mention US. A: For a US-domiciled investor, after fees, withholding taxes, currency hedging costs, and capital controls, US-listed equities are the only equities reliably investable. Owning EFA or VWO is an option but a small one, and most of the international exposure you actually want is already inside SPY's international revenue lines.

**Q: Should I add private equity or venture in the opportunistic sleeve?** A: Not at retail. Liquidity, fees, and the 10-year lockup destroy the "opportunistic" property. The retail opportunistic sleeve should stay liquid — cash, T-bills, and the options structures from Weeks 25-30. A barbell only works if both ends are liquid.

Q: What if I disagree with the quadrant model? A: Run the unlevered version and see how it performs in a decade you think the model gets wrong. The 1970s and 2022 are the two periods where the model gets criticised. In both, the all-weather book had a smaller real drawdown than 60/40 and a positive real return for the decade. The model is not perfect; it is more robust than the single-quadrant alternatives.

Q: How does this connect to factor investing in Week 23? A: The four tranches are the asset-class chassis. Factor tilts — quality, value, momentum, low-vol — go inside the growth tranche. You do not run quality at the expense of bonds; you run quality at the expense of plain S&P inside the 40% equity sleeve.

Q: Is the all-weather chassis tax-efficient? A: Less than 100% equity. Treasury and TIPS coupons are taxable as ordinary income at the federal level (state-tax-exempt for Treasuries). Gold ETFs (GLD, IAU) are taxed as collectibles at 28% long-term. Hold the income and store-of-value tranches in tax- advantaged accounts (IRA, 401k, HSA) and reserve the taxable account for the equity sleeve, where long-term cap gains and qualified dividends apply.

**Q: What is the single biggest mistake retail investors make with this framework?** A: Looking at the last twelve months. The all-weather chassis is a multi-decade architecture; in any given twelve-month window it will underperform whichever single quadrant is winning. The investor who checks performance monthly and chases the leader is not running all-weather — they are running momentum on quadrants, badly.

**Q: How does this relate to the dashboard at the top of the website?** A: The dashboard tracks a four-tranche book at 50/20/20/10 with a 2-year duration income sleeve and a barbell options overlay on the opportunistic sleeve. It is one specific instantiation of this chassis, not the chassis itself.