N
Nobu Quantitative Research
Launch Model
About Nobu Quant

Trust the market.
Adapt your exposure systematically.

Nobu (信) in Japanese means trust. At Nobu Quant, we trust the market, trust the numbers, and trust our rigorous, byte-for-byte reproducible backtesting.

We never predict market crashes — we systematically decide how much of the market to own every day. Our 4-layer architecture protects and compounds capital through four common-sense safeguards:

  • 1 Smart Exposure Sizing: Own more when it’s calm, own less when it’s stormy. Step back to cash during choppy turbulence so leverage never works against you.
  • 2 The Debt Market Radar: Corporate bonds spot trouble ~2 weeks before stocks do. We shift into cash before the stock market plunge accelerates.
  • 3 Buying the Panic Bottom: When panic forces institutions to dump stocks on firesale, we use reserved cash to buy the bottom and rocket out on the recovery.
  • 4 The Emergency Brake: An always-on stop-loss seatbelt that bounds rare black-swan shocks and guarantees our capital survives.

The result: more than double to nearly triple the market’s annual compounding rate — 26.9% Compound Annual Growth Rate (CAGR) on Nobu SPY (vs 11.1% for S&P 500) and 46.6% CAGR on Nobu QQQ (vs 16.6% for Nasdaq-100) across 19+ years of historical backtesting.

Nobu SPY Algorithm
26.93% CAGR
Sharpe 0.94 (vs 0.64) • 13.2x Terminal Wealth vs B&H
Nobu QQQ Algorithm
46.56% CAGR
Sharpe 1.20 (vs 0.80) • 85.6x Terminal Wealth vs B&H
Terminal Wealth Multiplier
13.2x – 85.6x
Cumulative outperformance factor vs Buy & Hold over 19+ years of backtesting (2007–2026)
Systematic 4-Layer Architecture

How Nobu Out-Compounds the S&P 500: The 8-Chapter Visual Storyboard

Step through each coordinated layer to see how dynamic volatility exposure, credit early warnings, panic liquidity provision, and stop-loss seatbelts beat both buy-and-hold and 3x leverage.

Architecture Storyboard
The Benchmark Titan

S&P 500 (SPY): America’s 500 Biggest Companies

The S&P 500 holds the 500 largest US companies (Apple, Microsoft, Nvidia, Amazon). Over 90% of professional Wall Street fund managers fail to beat it over 15 years. Beating this giant is the ultimate holy grail in investing.

$1,000 → $7,692 (11.14% CAGR over 19+ Years)
📊 The Benchmark Baseline: 11.14% CAGR over 19+ Years
Empirical Stress Testing

Full-Cycle Crisis Stress Testing: Nobu vs Buy & Hold

Direct, paired comparisons across complete market cycles (6 months pre-peak to 12 months post-trough).

Crisis Regime
6M Peak → 1Y Trough
Full-Cycle Return
Nobu vs Benchmark
Peak-to-Trough Decline
Nobu vs Benchmark
Time to Full Recovery
Nobu vs Benchmark
12-Mo Post-Trough Rally
Nobu vs Benchmark
2008 GFC ?
+31.84% vs −15.53% −57.05% vs −55.19% ~11 mos vs ~3.4 yrs +138.37% vs +44.63%
2011–12 Eurozone ?
+29.81% vs +27.52% −24.97% vs −19.39% ~2.5 mos vs ~4.5 mos +32.40% vs +26.10%
2018 Fed Crunch ?
+56.77% vs +26.04% −17.16% vs −19.78% ~3 mos vs ~4.0 mos +35.60% vs +31.49%
2020 COVID Shock ?
+46.28% vs +37.44% −33.36% vs −33.72% 70 days vs 140 days +70.99% vs +77.50%
2022 Rate Bear ?
+27.05% vs +3.61% −22.77% vs −24.50% 8 days vs ~1.2 yrs +47.82% vs +26.29%
2025 Trade War ?
+52.47% vs +22.84% −16.67% vs −5.68% ~2.5 mos vs ~5 mos +47.82% vs +21.52%
* Methodology: Standardized Full-Cycle Window = 6M pre-peak to 12M post-trough. Recovery time measured independently for each strategy from its own trough to its prior peak. Source: S&P 500 & Nasdaq-100 official market cycle high/low dates.
Interactive Research Engine

Historical Backtesting & Compounding Simulator

Simulate any historical start date and holding horizon across 19+ years of historical backtesting (2007–2026). Watch the portfolio rebase and compound daily against buy-and-hold benchmarks with realistic execution friction.

$
232 Months
Flexible
2-way sync with duration
Nobu SPY Final Value
$101,650
+10,065% Total Return
SPY Buy & Hold Value
$7,692
+669% Total Return
Terminal Wealth Multiple
13.2x
Outperformance vs Benchmark
Annualized Return (CAGR)
26.93%
Benchmark: 11.14%
Nobu SPY Algorithm Buy & Hold Benchmark
Compounded daily • Fills at Open • Realistic Friction
* Deterministic daily compounding ledger with realistic execution friction & transaction fees. Dataset: 4,867 daily verified sessions (2007–2026).
Quantitative Foundations

Core Research Insights: Academic Grounding & Empirical Truths

Our 4-layer architecture is built upon peer-reviewed empirical finance research and rigorous Nobel-prize winning econometric foundations.

Layer 01 • Volatility Drag

Volatility Clustering: Eliminate Geometric Drag

Market volatility is not constant; it clusters in violent temporal regimes. Holding fixed leverage through volatility spikes causes devastating mathematical compounding drag. Scaling leverage dynamically inversely to market turbulence protects accumulated capital.

• Key Academic Literature:
• Moreira & Muir (2017), "Volatility-Managed Portfolios", Journal of Finance, 72(4), pp. 1611–1644 (850+ citations).
• Bollerslev (1986), "Generalized Autoregressive Conditional Heteroskedasticity", Journal of Econometrics (42,000+ citations).
Receipt: +17.7% CAGR Vol-Dial vs SPXL −97.8% GFC Trough Collapse
Layer 02 • Credit Radar

Debt Seniority: ~17-Day Macro Early Warning

Corporate debt is senior to equity in capital structures. Bond markets reprice insolvency and liquidity stress ~14–17 days before stock market selloffs accelerate. Tracking high-yield credit momentum provides an empirical early warning radar to shift into Treasury cash.

• Key Academic Literature:
• Campbell, Hilscher & Szilagyi (2008), "In Search of Distress Risk", Journal of Finance, 63(6), pp. 2899–2939 (2,100+ citations).
• Gilchrist & Zakrajšek (2012), "Credit Spreads and Business Cycle Fluctuations", American Economic Review, 102(4), pp. 1692–1720 (2,600+ citations).
Receipt: +27.1% in 2022 Bear vs SPY −18.6% Drawdown
Layer 03 • Panic Liquidity

Firesale Extraction: The Panic Bounce

Deep 15%–30% market crashes trigger institutional mandate constraints, margin calls, and non-economic fire-sales. The Panic Bounce reserves cash specifically to supply liquidity into these dislocations, accumulating assets at generational discounts to ride explosive V-shaped rebounds.

• Key Academic Literature:
• Brunnermeier & Pedersen (2009), "Market Liquidity and Funding Liquidity", Review of Financial Studies, 22(6), pp. 2201–2238 (4,800+ citations).
• Coval & Stafford (2007), "Asset Fire Sales (and Purchases) in Equity Markets", Journal of Financial Economics, 86(2), pp. 479–512 (1,700+ citations).
Receipt: +138% GFC & +71% COVID Post-Trough Rebound Rally
Layer 04 • Tail Risk Stop

Circuit Breaker: Guaranteed Capital Survival

While dynamic exposure models mitigate ordinary volatility, rare black-swan dislocations can produce severe gap-downs. An always-on portfolio stop-loss acts as a non-negotiable circuit breaker, bounding tail losses and ensuring 100% capital survival to compound into the future.

• Key Academic Literature:
• Kaminski & Lo (2014), "When Do Stop-Loss Rules Stop Losses?", Journal of Financial & Quantitative Analysis, 49(1), pp. 137–154 (420+ citations).
• Kouwenberg & Zwinkels (2014), "Forecasting the Equity Premium with Stop-Loss Rules", Journal of Banking & Finance (300+ citations).
Receipt (19+ Years Backtest): Nobu SPY +10,065% Total (+26.9% CAGR) with Stop vs +7,023% without • Nobu QQQ +165,366% Total (+46.6% CAGR) with Stop vs +118,450% without
Rigor & Standards

Institutional Methodology & Engineering Moat

Engineered for full transparency, byte-for-byte reproducibility, realistic friction modeling, and frictionless execution.

01 • Transparency

Byte-for-Byte Reproducibility

Researchers and developers can re-run the research pipeline from raw daily adjusted market data and reproduce the exact equity curves down to the cent. Zero discretionary overrides.

02 • Timing Boundary

Zero-Lookahead Execution

Signals evaluated at Market Close $t−1$; orders execute strictly at Market Open $t$. Every series is lag-shifted and verified through unit-tested timing assertions.

03 • Realism

Friction & US Tax Engine

Idle cash held in SHV T-Bills. Rebalances account for 0.3bp spreads + SEC fees. Realistic US tax simulation models April 15 lump-sum payments and loss carryforward.

04 • Execution

Frictionless Simplicity

Trades once daily at the open in deep, liquid large-cap index ETFs. No high-frequency churning, no complicated options Greeks, and zero intraday screen monitoring.

Research Products

Product Offerings & Research Access

From free open-source backtest datasets to live signals, full Python codebases, and autonomous quant agent nodes.

Open Access

Verified Backtest Data

Free open download

Complete 19+ year transaction-level backtest ledger and daily equity curves in CSV/JSON format for independent verification.

  • • 4,867 Daily Equity Series (2007–2026)
  • • ~4,500 Trade Fill Ledgers
  • • Verifiable CSV/JSON Downloads
Live Signals

Strategy Weights Feed

$5 / month

Daily pre-market target exposure weights and credit gate alerts broadcasted via Telegram and webhook.

  • • Daily Pre-Market Telegram Feed
  • • Target Weights for SPY & QQQ
  • • Credit Early-Warning Alerts
Full Parameters
Full Source Code • Single-User License

Python Research Engine

$200 one-time

Complete, inspectable Python 3.11 backtesting engine with full strategy parameters. Protected against resale via cryptographic buyer watermarking.

  • • Nobu SPY Full Python Code & Parameters ($200)
  • • +$50 for Auto-Tuning Optimization Module
  • • +$100 for Dual SPY+QQQ Research Bundle
  • • Traceable single-user cryptographic license
Autonomous Node • Alpaca Ready

Nobu Autonomous Trading Node

$1,000 one-time

Turnkey, self-hosted Docker execution node. Zero-touch pre-market signal calculation, automated Alpaca execution, and private Telegram bot controls.

  • • Autonomous Head Trader Execution Node
  • • 1-Line Alpaca Paper → Live Trading Swap
  • • Private Telegram Bot Command & P&L Desk
  • • Alpha Mining & Parameter Sweep Harnesses