Kwant Atlas: The Map of Quantitative Finance

Welcome to Kwant Atlas, the universal, interconnected knowledge base and diagnostic study tool for quantitative finance.

Rather than a linear, week-by-week curriculum, the Atlas is structured around the 8 foundational operational disciplines of quantitative finance. Every discipline connects directly back to first principles: mathematical ground truths, computational systems, empirical failure modes, and canonical literature.

---

🚦 New here? Start here.

The Atlas is a map, not a straight line - but if you are starting from zero and want a route, follow this:

1. Foundations first - this is required, not an appendix. Read the math toolbox before (or alongside) any pillar. The core trio is Linear Algebra → Calculus & Optimization → Probability & Measure Theory; then take the extra foundations each pillar needs (the Math Foundations hub lists the exact consumption order - e.g. Econometrics before Pillar 1, Stochastic Calculus before Pillar 3, Statistics before Pillar 4).
2. Pick one pillar and walk its Reading Path. Every pillar hub has a staged “Reading Path (Zero to X)” that starts at a 01-from-zero-intuition page and climbs to 06-advanced-extensions, plus a “Before this pillar (foundations)” line naming exactly which foundation folders to read first. Recommended first pillar by interest: 1 · Quantitative Research to build strategies, 3 · Derivative Pricing for the mathematics of pricing, 4 · Quantitative Risk to understand risk.
3. Mid-project? Use it as a lookup + debug tool. Don’t read front-to-back - jump straight in via the pillar hubs’ lookup tables and formula indexes, check “what is near this topic” on the interactive graph (see the Knowledge Graph section below), and use the diagnostic matrix below for symptom → cause → fix.

Every page carries runnable code. All Python is stdlib-or-numpy, deterministic, and reproduces the printed output shown. Clone the repo and run it as you read.

---

---

1. Quantitative Research (Alpha Generation)

Analyzing historical and alternative market data to discover predictive trading signals and statistical anomalies.

• Statistical Arbitrage & Pairs Trading: Cointegration, Ornstein–Uhlenbeck spread dynamics, and half-life decay.
• Cross-Sectional & Time-Series Momentum: Factor ranking, CTA trend following, volatility scaling, and momentum crashes.
• Fundamental Multi-Factor Models: Fama–French 5-factor, Barra cross-sectional style factors, and factor crowding.
• Signal Processing & Kalman Filtering: Linear state-space systems, dynamic hedge ratios, and adaptive noise filtering.
• Feature Engineering & Target Labeling: Flaws of fixed-time horizons, the Triple Barrier Method, and Meta-Labeling.
• Backtesting Hygiene & Deflated Sharpe Ratio: The multiple testing problem, selection bias, and Bailey & Lopez de Prado’s DSR.

---

2. Algorithmic and High-Frequency Trading (HFT)

Designing automated execution systems operating across the millisecond to nanosecond frontier.

• Market Microstructure & Order Types: Passive vs aggressive orders, hidden icebergs, pegged orders, and maker-taker fee rebates.
• Low-Latency Systems Architecture: Kernel bypass (Solarflare Onload, DPDK), CPU pinning, NUMA affinity, and zero-copy pipelines.
• Queue Position & Fill Probability: FIFO vs Pro-Rata matching engines, cancellation dynamics, and adverse selection at queue heads.
• Execution Algorithms: VWAP, TWAP, & POV: Institutional order slicing, intraday volume curves, and tracking error optimization.
• Optimal Execution & Almgren–Chriss: Permanent vs temporary market impact, execution risk aversion, and calculus of variations trajectories.
• Hardware Acceleration & FPGA: Wire-speed packet parsing on silicon, sub-50ns tick-to-trade, and hardware safety gates.

---

3. Derivative Pricing and Structuring

The traditional sell-side quant domain: valuing non-linear financial contracts and engineering self-financing dynamic hedges.

• No-Arbitrage Foundations & Binomial Trees: Law of one price, put-call parity, Cox–Ross–Rubinstein discrete replication, and American early exercise.
• Black–Scholes–Merton & Feynman–Kac Bridge: Delta-neutral hedging, parabolic PDE derivation, and risk-neutral conditional expectations.
• The Greeks & Dynamic Hedging: First and higher-order Greeks (Delta, Gamma, Vega, Theta, Vanna, Volga) and the fundamental Gamma-Theta trade-off.
• Implied Volatility Surfaces & Smiles: Numerical root finding, skew/smile dynamics, sticky rules, and Dupire local volatility.
• Advanced Volatility: Heston & SABR Models: Stochastic variance processes, the Feller condition, and swaption smile fitting.
• Interest Rate & Term Structure Models: Yield curve bootstrapping, short-rate dynamics (Vasicek, CIR), and the Hull–White framework.

---

4. Quantitative Risk Management

Measuring, bounding, and mitigating financial exposure to guarantee firm survival across extreme market volatility.

• Value at Risk & Expected Shortfall (CVaR): Coherent risk measure axioms, the subadditivity flaw of VaR, and Cornish–Fisher expansions.
• Parametric, Historical, & Monte Carlo VaR: Variance-covariance methods, filtered historical simulation (FHS), and Kupiec backtest batteries.
• Extreme Value Theory & Fat Tails: Breakdown of normality, Peaks-Over-Threshold (POT), Generalized Pareto Distributions (GPD), and the Hill tail index.
• Stress Testing & Reverse Stress Testing: Historical crisis replay (1987, 1998, 2008, 2020), macro factor shocks, and reverse stress testing.
• Credit Risk & the Merton Structural Model: Equity as a call option on firm assets, distance-to-default, and structural default probability.
• Liquidity Risk & Margin Spirals: Bid-ask spread hair-cuts, market depth exhaustion, and the Brunnermeier–Pedersen margin spiral.

---

5. Portfolio Construction and Optimization

Applying mathematical frameworks to allocate capital across assets, maximizing risk-adjusted return under real friction.

• Modern Portfolio Theory & Mean–Variance Frontiers: Markowitz quadratic programs, the tangency portfolio, and the “estimation error maximizer” paradox.
• Covariance Shrinkage & RMT Denoising: The curse of dimensionality ($N>T$), Ledoit–Wolf analytical shrinkage, and Marchenko–Pastur eigenvalue filtering.
• Black–Litterman Bayesian Asset Allocation: Reverse optimization for equilibrium implied returns, blending quant views with market priors.
• Risk Parity & Equal Risk Contribution (ERC): Marginal risk contribution, why 60/40 is 90% equity risk, and leverage in risk-balanced portfolios.
• Hierarchical Risk Parity (HRP): Correlation distance metrics, tree graph clustering, and matrix-inversion-free allocation.
• Transaction Costs & Turnover Constraints: Penalizing quadratic market impact, L1/L2 regularization, and sparse rebalancing frontiers.

---

6. Market Making and Liquidity Provision

Designing automated systems that quote continuous two-sided liquidity, profiting from the bid-ask spread while managing inventory and adverse selection.

• Limit Order Book Mechanics & L3 Data: Level 1/2/3 data feeds, order reconstruction engines, and Order Flow Imbalance (OFI).
• The Avellaneda–Stoikov Model: Hamilton–Jacobi–Bellman (HJB) formulation, reservation prices, inventory penalty, and optimal quotes.
• Adverse Selection & Glosten–Milgrom: Informed vs uninformed noise traders, Bayesian price updating, and Kyle’s Lambda price impact.
• Spread Decomposition & the Roll Model: Return autocovariance, bid-ask bounce, and separating inventory costs from adverse selection.
• Inventory Management & Quote Skewing: Asymmetric quote positioning, mean-reverting inventory targets, and overnight risk.
• Toxic Order Flow & VPIN: The Lee–Ready algorithm, Volume-Synchronized Probability of Toxicity (VPIN), and flash crash early warning metrics.

---

7. Machine Learning and Alternative Data

Extracting non-linear signals and structural patterns from alternative, unstructured, and high-dimensional datasets.

• Financial ML Pitfalls & Low SNR: Why standard ML fails, the microscopic Signal-to-Noise Ratio, non-stationarity, and subtle data leakage traps.
• Tree-Based Factor Ranking & Purged CV: LightGBM/XGBoost factor synthesis, Mean Decrease Accuracy (MDA), and Purged/Embargoed Cross-Validation.
• Financial NLP & Earnings Transcripts: SEC 10-K delta analysis, FinBERT fine-tuning, earnings call Q&A sentiment, and LLM extraction.
• Alternative Data Pipelines & Evaluation: Credit card streams, web scraping, geolocation, point-in-time hygiene, and alpha decay.
• Regime Classification: HMM & GMM: Unsupervised regime detection, Baum–Welch Expectation–Maximization, and Viterbi state path decoding.
• Deep Learning for Sequences: LSTMs, Temporal Convolutional Networks (TCN), and Temporal Fusion Transformers for tick and bar series.

---

8. Quantitative Development (Quant Engineering)

The software and systems engineering backbone: translating mathematical models into ultra-low-latency production infrastructure.

• High-Performance C++ for Trading: Zero-allocation paradigms, CPU cache locality (L1/L2/L3), cacheline false sharing, and SIMD vectorization.
• q: Column-oriented architectures, kdb+/q vector primitives, DuckDB/ClickHouse pipelines, and point-in-time as-of joins.
• Event-Driven Backtesting Engines: Vectorized vs event loops, realistic fill modeling, order state machines, and deterministic historical replay.
• FIX Protocol & Exchange Connectivity: Tag-value FIX, session recovery, binary ITCH/OUCH protocols, and order state lifecycles.
• Concurrency & Lockless Programming: The LMAX Disruptor pattern, SPSC ring buffers, memory fences, and atomic synchronization.
• Production Risk Guards & Kill Switches: Wire-speed pre-trade risk checks, fat-finger caps, rate throttles, and automated kill switches.

---

🔬 Math Foundations

Read this before (or alongside) the pillars - it is the shared foundation the whole Atlas is built on, not an optional appendix. Every pillar’s 01-from-zero-intuition page assumes the foundations its prerequisites name; the Math Foundations hub gives the exact consumption order (which foundation to read before which pillar). The core trio everyone needs:

• 📐 Linear Algebra & Matrices: Vector spaces, spectral theory, positive semi-definiteness, and SVD.
• 📈 Calculus & Constrained Optimization: Gradients, Hessians, Taylor expansions (Greeks), and KKT conditions.
• 🎲 Probability & Measure Theory: Probability spaces, filtrations, conditional expectations, and martingales.

Then the foundations each pillar draws on: Stochastic Calculus (Pillars 3, 6), Econometrics & Time Series (Pillars 1, 6), Statistics & Inference (Pillars 4, 7), Bayesian Statistics (Pillar 5), Numerical Methods (Pillars 2, 3, 8), and Ergodicity & Statistical Mechanics (Pillar 5, Kelly).

---

📊 Accounting & Finance (Cross-Cutting Area)

The company underneath every security. This area sits alongside the 8 pillars and the Foundations toolbox, covering how a business is reported, valued, financed, quality-checked, and sourced as data - the shared substrate for both quantitative factor projects and discretionary fundamental work.

→ Open the Accounting & Finance area hub

The 8 topic-folders, in recommended learning order:

1. Financial Statements & Accounting - the accounting equation, double-entry, and the three statements.
2. Core Financial Ratios - profitability, multiples, liquidity, leverage, and the ratio lookup table.
3. Equity Valuation - DCF, cost of capital, multiples, comps, and the margin of safety.
4. Fundamental Analysis & Screening - Graham-style criteria and mechanical screen construction.
5. Accounting Quality & Red Flags - accruals, earnings management, and the red-flag checklist.
6. Capital Structure & Corporate Finance - Modigliani–Miller, the tax shield, distress, and agency costs.
7. Quantitative Fundamental Investing - value, profitability, investment, and accrual factors.
8. Data Sources & Corporate Data - EDGAR/XBRL, vendors, and point-in-time data hygiene.

---

🛠️ First-Principles Diagnostic Matrix: “Why Is My Strategy Failing?”

When a quantitative strategy underperforms, drawdowns blow out, or live execution bleeds cash, use this first-principles troubleshooting matrix to isolate the mathematical or structural flaw:

Observed Symptom in Production / Test	Underlying First-Principles Failure Mode	Diagnostic & Root Cause	Remedy Note
Strategy shows 3.5 Sharpe in backtest, but instantly loses money live	Backtest Overfitting & Selection Bias	Testing thousands of parameter permutations without controlling for sample length or non-normality.	Backtesting Hygiene (DSR)
Pairs trading spread diverges indefinitely into a 5-sigma loss	Cointegration Structural Break	The underlying linear relationship $z_t=P_A-\beta P_B$ broke down due to corporate or macro shifts.	Statistical Arbitrage
Portfolio weights swing violently ($+300\%$ to $-200\%$) on small updates	Inverted Covariance Noise Maximization ($N>T$)	Unconstrained sample covariance matrix inversion ${\Sigma }^{-1}$ magnifies empirical noise eigenvalues.	Covariance Shrinkage & RMT
Market maker fills 100 consecutive buys right before price crashes	Toxic Order Flow & Adverse Selection	Passive limit orders at inside spread were swept by informed traders; quote skewing failed to react.	Adverse Selection (Glosten–Milgrom)
Delta-hedged option portfolio bleeds cash in fast volatile markets	Discrete Hedging Error & Jump Risk	Black–Scholes continuous rebalancing assumption violated ($dt>0$); unhedged Gamma loss $\frac{1}{2}\Gamma (\Delta S{)}^2\approx \frac{1}{2}\Gamma S^2{\sigma }^2\Delta t$.	Black–Scholes (Greeks & Hedging)
Machine learning model achieves 95% accuracy in-sample, 0% live	Information Leakage in Cross-Validation	Standard K-Fold CV leaked auto-regressive returns across folds; standard differencing destroyed memory.	Financial ML Pitfalls
Trading engine latency spikes from 1 $\mu$s to 2 ms intermittently	OS Syscall / Dynamic Memory Heap Lock	Fast path called malloc or hit OS page fault, triggering kernel context switch and heap contention.	High-Performance C++
Fund suffers catastrophic liquidation during market stress	Funding & Market Liquidity Spiral	Prime broker hiked margin haircuts; forced liquidations depressed market prices in feedback spiral.	Liquidity Risk & Margin Spirals

---

🌐 Interactive Graph & Exploration

• 🕸️ Open Fullscreen Interactive D3 Graph: Explore all 8 operational clusters, foundational nodes, cross-pillar bridges, and filter by hard-skill intensity (Math, Code, Intuition).
• 📖 Glossary & Symbol Index: every term and symbol, A–Z, with the canonical page to jump to - the “find a word without knowing the pillar” entry point.
• 🩺 Diagnostic Index (Symptom → Cause → Fix): the global debug matrix - find your symptom, get the first-principles root cause and the page that fixes it.
• 📁 Obsidian Integration: If cloning locally (git clone https://github.com/kwantklubben/kwant-atlas), open the folder directly in Obsidian to view pre-configured branded color clusters and tag groups.