Chapter 8: The fancy aisle
In 2022, the most sophisticated diversification recipe retail money could buy fell 22.8%. The plain S&P 500 fell 18.1% that year, and chapter 7's unfashionable 60/40 lost about 16.1%. Hold those three numbers while you walk this aisle, because every stall here sells some version of "smarter than the classic recipes," and 2022 handed out graded receipts.
The aisle deserves a fair tour rather than a sneer. Every stall sells something real: a genuine research finding, a genuine tax mechanism, a genuine diversification idea, sometimes a genuine asset class. None of it is forbidden, and some of it, at the right size and price for the right person, earns its keep. The discipline is the same at every stall, and it is chapter 7's discipline with the volume turned up: read the price tag, read the receipt, and remember that the burden of proof sits on whatever costs more than $36 a year per $100,000.
What did Yale have that you cannot buy?
David Swensen ran Yale's endowment to famous results, and his 2005 book Unconventional Success translated the thinking for individuals: 30% US stocks, 15% foreign developed stocks, 5% emerging markets, 20% real estate investment trusts, 15% Treasury bonds, and 15% TIPS. The slices sum to 100, every ingredient is an index fund, and the recipe is genuinely thoughtful about inflation, with 35 points in REITs and TIPS standing guard.
The honest critique starts with what made Yale exceptional, because none of it survived the translation. Yale's edge was illiquid assets and access to elite private managers, things a retail account cannot buy at any reasonable price, and Swensen himself said ordinary investors should index instead. Strip the edge away and the retail recipe is an index portfolio with a 20% REIT bet attached. That bet has a receipt: REITs are stocks that hold buildings, and in 2022 they fell alongside the rest of the stock market while the recipe's Treasury sleeve also lost money in the rate shock. The recipe remains a reasonable index portfolio. The name on the stall sells Yale, and Yale is the part not included.
The all-weather stall
Risk parity deserves its idea stated fairly. A 60/40 portfolio holds 40% bonds by dollars, but stocks are so much more volatile that they contribute nearly all of the risk. Risk parity balances the risk contributions instead of the dollars, which requires holding a lot of bonds and then applying leverage, borrowed exposure, so the bond-heavy mix can still grow. The retail version made famous as "All Weather" runs about 30% stocks, 55% Treasuries split between long and intermediate maturities, 7.5% gold, and 7.5% commodities, which sums to 100 and is built to survive any of four economic seasons.
Then 2022 staged the exact storm chapter 3 described: inflation pushed the stock-bond correlation to about +0.50, the highest in AQR's half-century sample, and the asset being leveraged was the asset falling fastest. The flagship retail risk-parity ETF, RPAR, fell 22.8% while the S&P 500 fell 18.1%. The all-weather portfolio had its worst year in precisely the weather it was marketed against, because leverage amplifies whatever the bonds do, in both directions. The idea is honest and the research lineage is serious, and a Bridgewater-branded All Weather ETF launched in March 2025 for anyone still drawn to it. The receipt simply proves the product has weather of its own.
The factor stall
Factor tilts overweight kinds of stocks, small companies and cheaply priced "value" companies above all, because long academic samples in the Fama-French lineage show those groups beating the broad market on average. The premium is real in the data. So is the drought: the 2010s were the worst stretch for value investing since the Depression, a full decade of watching the plain index win while the tilt apologized. Then 2022 arrived as a historic value year and paid the survivors.
That sequence is the entire test of this stall, and it is a test of you rather than of the research. A tilt is a decade-scale commitment, and the well-documented failure mode is behavioral: adopt the tilt after a good run, hold it through three years of drought, abandon it right before the turn, and convert a real premium into a realized loss. Anyone who cannot honestly commit to a tilted fund through ten underperforming years already has their answer at this stall, and the broad index from chapter 7 keeps every factor at market weight in the meantime.
Who actually comes out ahead with direct indexing?
Direct indexing replaces your index fund with the index's individual stocks, held in a separately managed account in your name. The pitch is twofold: customization (drop your employer's stock, apply a screen) and stock-level tax-loss harvesting, selling the individual losers inside a rising index to bank losses that offset gains elsewhere on your tax return. The mechanism works, the fee is 0.10–0.40% a year against 0.03% for the ETF, and Cerulli has projected the segment to pass $800 billion by the end of 2026, a projection worth labeling as one.
The winners are a specific, identifiable group: high earners in taxable accounts who reliably generate capital gains that need offsetting, keep adding fresh cash so the harvesting engine has new lots to work with, and stay long enough to justify the setup. For them the tax math can clear the fee. For a buy-and-hold investor, and for anyone whose money sits in an IRA or 401(k) where losses are worthless on a tax return, direct indexing is complexity at roughly ten times the price, harvesting deductions no one can use. The stall is honest about its mechanism and quiet about its audience, so bring your own tax return as the entry test.
The newest stall: private assets in your 401(k)
This stall did not exist when the sibling guides went to press, so date everything in it. Executive Order 14330, signed August 7, 2025, directed regulators to open 401(k)-style menus to private equity, private credit, real estate, digital assets, and lifetime-income products. The Department of Labor rescinded its earlier caution and proposed a safe-harbor rule on March 30, 2026, listing six factors a plan must weigh; comments closed June 1, 2026, and the final rule is pending as this guide goes to press. Money will follow the rule, and the pitches have already started, because a trillion-dollar industry just gained legal scaffolding into the largest pool of patient money in the country.
The DOL's six factors are also the buyer's complete concern list, which is a rare gift from a regulator.
- Performance. What is the net-of-fee record inside a daily-priced retirement wrapper, as opposed to the institutional record in the brochure?
- Fees. Private funds carry a 2-and-20 heritage; the target-date fund this would replace averages 0.27%.
- Liquidity. Your 401(k) trades daily; the underlying assets do not. Gates and windows decide who exits when it matters.
- Valuation. Appraisal-based prices move smoothly by construction, so some of the advertised calm is the measuring stick, not the asset.
- Benchmarks. A product no standard index describes is a product whose underperformance is hard to ever prove.
- Complexity. If the structure cannot be explained in two sentences, the explanation gap is borne by you.
The pitch reached Hugo before the final rule did: a lunch presentation offering his practice's 401(k) a private-credit sleeve, "institutional yields with low volatility." Hugo had the six factors printed and used them as a script, out loud and in order. Net-of-fee record in a daily wrapper: under two years. Fees: a stack several times his plan's index funds. Liquidity: quarterly windows, with gates. Valuation: appraisal-based, which answered the low-volatility claim by itself. Benchmark: the presenter offered "cash plus 5%," which is a goal wearing a benchmark costume. Complexity: Hugo asked the presenter to explain the structure in two sentences, counted three disclaimers, and stopped counting. He declined politely and specifically: not forbidden, not proven at this price, and not before a record exists. The presenter wrote down "revisit next year," which was fair.
The crypto shelf
Spot bitcoin ETFs have traded since January 2024, which collapsed the practical question into a portfolio question: one ticker, any account, what size. BlackRock's published sizing work answers with 1–2% and supplies the logic: at about 2%, bitcoin already contributes risk comparable to holding a mega-cap stock position, because its volatility is several times the market's. Size, in other words, is set by risk contribution rather than conviction.
The record sets the rehearsal. Bitcoin fell 77% from its 2021 peak into 2022, and inside the post-ETF era it has already drawn down roughly half into early 2026. A sleeve sized at 2% that halves costs the portfolio about one percentage point, an entry in your annual review. The same asset at 15% turns the same routine event into a plan-breaking year. Anyone taking the sleeve should write it as a named exception, sized so a halving happens quietly, and rebalance it on chapter 10's schedule like everything else it owns.
Nothing in this aisle is forbidden. Everything in it must beat its own fee and your own patience, the receipts get read before the brochures, and the burden of proof never moves off the seller.
Line eight is the honesty line: "Exceptions I allow myself: ____ (each one named, each one sized)." Hugo's reads: "Exceptions: a 2% sleeve I call what it is; no private assets until a net-of-fee record exists at this price." Writing "none" is a complete and respectable answer.
Key takeaways
- Every stall sells something real, and 2022 graded the claims: RPAR fell 22.8% against the S&P 500's 18.1% and the classic 60/40's roughly 16.1%, because leveraged bonds amplified the rate shock.
- Swensen's retail recipe (30/15/5/20/15/15) is a sensible index portfolio with a 20% REIT bet; the Yale edge, illiquid assets and manager access, is exactly the part retail cannot buy, and REITs fell with stocks in 2022.
- Factor premia are real in the Fama-French data, the 2010s value drought was the worst since the Depression, and 2022 paid the survivors; the tilt is a decade-scale commitment that loses money when abandoned mid-drought.
- Direct indexing at 0.10–0.40% (versus 0.03% for the ETF) genuinely helps high earners with gains to offset and fresh cash; Cerulli's $800 billion by end-2026 is a projection, and in an IRA the product is complexity at ten times the price.
- Executive Order 14330 (August 7, 2025) and the DOL's proposed safe-harbor rule (March 30, 2026, final rule pending as of June 2026) are opening 401(k)s to private assets; the rule's six factors, performance, fees, liquidity, valuation, benchmarks, complexity, double as your complete question list.
- A crypto sleeve, if taken, is sized by risk contribution: BlackRock calls 1–2% reasonable, and the record (−77% in 2021–2022, roughly half post-ETF into early 2026) says size it to halve quietly.
Sources: White House Executive Order 14330 · DOL proposed safe-harbor rule · BlackRock on sizing bitcoin · AQR on value investing · AQR on the stock-bond correlation · Morningstar target-date research · Swensen's Unconventional Success and Cerulli's direct-indexing projection, cited by name.