Sanad research · Foundational paper

Growth on one side.
Defense on the other.

A deliberate barbell for long-horizon investors: one bounded Nasdaq-100 growth pole, five defensive sleeves, and no pretense that the middle is automatically safe.

Read the full paper ↗Get started →

Historical windowJan 2008–Jul 2026

Historical proxyMonthly rebalanced

LeverageNone

Decision rightSanad investment committee

01The thesis

Survival and participation are complements.

The goal is not to make every sleeve moderate. It is to seek growth where we mean to take risk, then hold enough liquidity and macro defense to survive being wrong.

Dalio contributes

Economic balance

Assets respond differently to growth and inflation surprises. A portfolio should not quietly make one dominant macroeconomic bet.

Taleb changes

The model of risk

Fat tails, nonlinear damage, and estimation error make fragility more important than a neat average. Liquidity has option value when disorder arrives.

Sanad concludes

Govern the barbell

Bound the concentrated growth pole at 30%, refuse leverage, and version every policy change. This is robust and barbell-shaped—not inherently antifragile.

02The flagship

One aggressive pole. Five defensive jobs.

Thirty percent is an explicit Nasdaq-100 growth bet. Fifty-five percent sits in Treasuries and short TIPS. The remaining fifteen percent holds real assets for inflation and monetary disorder.

Sanad Flagship target allocation: 30% Nasdaq-100 growth, 25% Treasury bills, 20% intermediate Treasuries, 10% short TIPS, 10% gold, and 5% commodities — Policy weights are governance decisions, not forecasts of future returns.

SleeveRoleWeight

Nasdaq-100 growthQQQMAggressive growth pole30%

Treasury billsSGOVLiquidity reserve25%

Intermediate TreasuriesVGITDisinflation defense20%

Short TIPSVTIPInflation-linked defense10%

GoldGLDMMonetary-reserve diversification10%

Broad commoditiesCOMTInflation surprise5%

The QQQ proxy contributed roughly 73% of estimated portfolio variance. Capital balance is not the same as risk balance.

03The growth sleeve

QQQM is the growth pole—not the diversification.

The concentration is deliberate and bounded. The remaining 70% supplies liquidity and macro defense; it does not make the Nasdaq-100 sleeve safe.

Current-policy diagnostics only. QQQ is the long-history proxy for QQQM; historical results are not forecasts.

Capital weight30%explicit concentration ceiling

Estimated variance contribution73%full-sample ex-post estimate

Actual-fund policy CAGR8.7%Nov 2020–Jul 2026 · short history

“Middle-risk trap” is a product-design critique, not a theorem. QQQM is concentrated, valuation-sensitive, and capable of decade-long recovery periods.

04The evidence

The common sample favored the barbell. We do not assume the future will.

The QQQ proxy version beat 60/40 on return and recorded less drawdown in this window. That is evidence to examine, not a premium we can promise—especially because the window starts after the dot-com collapse.

Jan 2008–Jul 2026Sanad Flagship60/40 benchmark

Annualized return7.4%6.5%

Annualized volatility7.0%10.3%

Maximum month-end drawdown−16.2%−32.1%

Maximum daily-close drawdown−20.2%−37.2%

Expected shortfall (95%)4.0%6.7%

$10,000 ending value$37,415$32,368

Growth of ten thousand dollars from 2008 through July 2026 for the Sanad Flagship, a 60/40 benchmark, and Treasury bills — Nominal monthly proxy before client taxes, slippage, and intra-month losses. The result is sample- and path-dependent.

05Correlation risk

Diversification is a moving target.

Defensive assets did not maintain one stable relationship with equities. That is why no sleeve gets to carry the entire burden of protection.

Rolling 36-month correlations of the QQQ growth proxy with Treasuries, short TIPS, gold, and commodities from 2008 through July 2026 — Rolling 36-month monthly-return correlations. A negative correlation is helpful, but it is not a permanent property.

Equity versusLowMedianLatest

Treasuries−0.60−0.21+0.33

Short TIPS−0.14+0.19+0.39

Gold−0.33+0.14−0.08

Commodities−0.25+0.30−0.19

Latest values are sample observations, not forecasts. Treasury diversification weakened materially near the end of this window.

06Implementation drag

A policy is only as good as its execution.

We model fund expenses and trading friction instead of treating the gross backtest as investable. Drift bands reduced sample turnover, but they are an operational rule—not a return promise.

Weighted fund expenses: ~0.11% / year
Band trigger events: 18
Annual one-way turnover: 5.21%
$10k ending value at 50 bp friction: $34,103

In the same sample, forced monthly rebalancing produced 223 events and 14.19% annual one-way turnover. The comparison excludes taxes and market impact.

Ending value of ten thousand dollars for the Sanad Flagship under zero, ten, twenty-five, and fifty basis points of annual implementation friction — Illustrative annual friction applied to the historical monthly proxy. Actual fees, spreads, taxes, and market impact vary.

07Tail behavior

Drawdowns remain material.

Diversification is not immunity. In a crisis, correlations can change and liquid assets can fall together. The design aims to reduce fragility, not manufacture certainty.

PeriodSanad60/40

2008Global financial crisis−11.0%−22.9%

2020Pandemic shock+19.4%+12.7%

2022Inflation and rate shock−13.5%−15.2%

Historical drawdown chart comparing the Sanad Flagship with a 60/40 benchmark and Treasury bills — Peak-to-trough loss measured at month-end. Intramonth losses may have been worse.

Omitted regime · QQQ history

−83%. Thirteen years to recover.

QQQ fell 82.96% from March 2000 to October 2002 and did not regain its prior adjusted-price peak until February 2015. The common portfolio sample begins in 2008, so this counterexample must remain visible beside every higher-return claim.

QQQ adjusted-close drawdown from inception through 2003, including an 82.96 percent dot-com collapse — QQQ is the long-history proxy for QQQM. Adjusted closes are not an official index-return series.

08Path dependence

The average hides the journey.

A 4,000-path block bootstrap preserves some historical clustering and illustrates outcome dispersion. It does not assign probabilities to the future or invent regimes absent from the sample.

Over ten-year resamples, annualized returns ranged from 3.2% to 11.2% between the 5th and 95th percentiles. Maximum drawdown ranged from −23.4% to −4.3%.

Ten-year block-bootstrap ranges showing annualized return and maximum drawdown for the Sanad Flagship — 5th–50th–95th percentiles from 4,000 resamples of 12-month historical blocks.

09Governance

The portfolio is a policy, not a preference screen.

Clients do not edit securities or weights. The backend owns the approved strategy, and every material change creates a new immutable version.

01Observe
Monitor drift, liquidity, data quality, and account restrictions.
02Research
Document the evidence, tradeoffs, operational impact, and failure modes.
03Approve
Record committee approval, effective date, and version identifier.
04Execute
Use cash first, respect drift bands, and avoid leverage or forced turnover.
05Reconcile
Confirm fills, positions, cash, and exceptions against the custodian.

Read the full argument

Methods, mathematics, limitations, and sources.

The paper explains the regime framework, fat-tail mathematics, portfolio construction, historical proxy, bootstrap analysis, implementation rules, and the claims this evidence cannot support. The companion workbook exposes the monthly series, rolling correlations, cost cases, sources, and formula checks.

Open the white paper ↗Download the research workbook ↓

Important

This material is research, not individualized investment advice, a recommendation, an offer, or a guarantee. Historical proxy results are hypothetical and may not reflect fees, taxes, spreads, market impact, implementation constraints, or intramonth losses. Historical relationships can change. Any live implementation requires legal, regulatory, suitability, tax, liquidity, and operational review.