Quantitative finance research & learning resources

Research context matters. Records state the access depth and limitations of each review; a citation or predictive result does not establish an implementable strategy.

Interpreting Prediction Market Prices as Probabilities

Wolfers & Zitzewitz, Interpreting Prediction Market Prices as Probabilities, NBER Working Paper 12200, 2006. Abstract/search record; direct page retrieval failed.

Reviewed: 2026-09-12
Access: Abstract / metadata
Status: Wolfers & Zitzewitz, Interpreting Prediction Market Prices as Probabilities, NBER Working Paper 12200, 2006. Abstract/search record; direct page retrieval failed.
Markets and data dates: Binary event contracts; theoretical framework, empirical sample not verified.
Limitations: Teach the assumptions behind reading prices as probabilities. Foundational source; full-text review needed before authoring its derivations.

Unravelling the Probabilistic Forest: Arbitrage in Prediction Markets

Saguillo et al., Unravelling the Probabilistic Forest: Arbitrage in Prediction Markets, arXiv v1, 5 August 2025. Full-text sections. Final proceedings status not verified.

Reviewed: 2026-09-12
Access: Selected full-text sections
Status: Saguillo et al., Unravelling the Probabilistic Forest: Arbitrage in Prediction Markets, arXiv v1, 5 August 2025. Full-text sections. Final proceedings status not verified.
Markets and data dates: Polymarket; markets resolved 1 April 2024–1 April 2025.
Limitations: Taxonomy of within-market and related-market arbitrage; semantic dependence, observational detection, and execution inference. Historical inferred opportunities/profits do not establish current fillability. Treat automated contract matching as a candidate-generation step requiring verification.

Decomposing Crowd Wisdom

Le, Decomposing Crowd Wisdom, preprint submitted February 2026, revised 4 August 2026. Full-text data/methods sections.

Reviewed: 2026-09-12
Access: Selected full-text sections
Status: Le, Decomposing Crowd Wisdom, preprint submitted February 2026, revised 4 August 2026. Full-text data/methods sections.
Markets and data dates: Kalshi: earliest observed trade 1 July 2021; Polymarket: 6 March 2023; analysis cutoff 31 December 2025.
Limitations: Build calibration by domain and time to resolution, with uncertainty and event clustering. Descriptive findings and noisy subgroup estimates cannot be promoted directly into a profitable correction strategy. Author replication materials are linked on the paper page but were not executed.

Optimal Market Making in Prediction Markets

Feil & Nendel, Optimal Market Making in Prediction Markets, preprint, 20 July 2026. Abstract/metadata only.

Reviewed: 2026-09-12
Access: Abstract / metadata
Status: Feil & Nendel, Optimal Market Making in Prediction Markets, preprint, 20 July 2026. Abstract/metadata only.
Markets and data dates: Model of binary event contracts; numerical analysis; no historical sample stated in inspected abstract.
Limitations: Advanced extension from inventory risk to terminal settlement risk, latent beliefs, and stochastic control. Inspect assumptions, proofs, and numerical setup before a replication lesson.

Trading and arbitrage in cryptocurrency markets

Makarov & Schoar, Trading and arbitrage in cryptocurrency markets, 2020 journal article, institutional record.

Reviewed: 2026-09-12
Access: Metadata / summary
Status: Makarov & Schoar, Trading and arbitrage in cryptocurrency markets, 2020 journal article, institutional record.
Markets and data dates: Cryptocurrency exchanges and cross-country price differences; exact sample endpoints not verified here.
Limitations: Establish segmentation, capital mobility, and settlement frictions as core arbitrage concepts. Historical exchange structure may differ materially from current markets; retrieve author manuscript and sample details before implementation.

Crypto carry

Schmeling, Schrimpf & Todorov, Crypto carry, BIS Working Paper 1087, first published 4 April 2023. Institutional summary/abstract.

Reviewed: 2026-09-12
Access: Abstract / metadata
Status: Schmeling, Schrimpf & Todorov, Crypto carry, BIS Working Paper 1087, first published 4 April 2023. Institutional summary/abstract.
Markets and data dates: Bitcoin and Ether spot/futures across crypto and traditional venues; sample dates not disclosed in inspected summary.
Limitations: Cash-and-carry economics, convenience yield, margin and liquidation risk. The legacy PDF URL redirected to the landing page; do not claim full-PDF review or a verified revision date.

Designing funding rates for perpetual futures in cryptocurrency markets

Kim and Park study funding-rule design and replication for perpetual futures using arbitrage pricing and path-dependent infinite-horizon backward stochastic differential equations.

Reviewed: 2026-09-12
Access: Abstract / metadata; primary arXiv submission history checked
Status: arXiv preprint; journal publication not verified
Markets and data dates: Theoretical perpetual futures in cryptocurrency markets; no empirical trading sample verified in the inspected abstract.
Limitations: Abstract and version metadata only. Full assumptions and proofs were not audited. A theoretically anchoring rule does not establish the value of a trading strategy under execution costs, finite capital or liquidation constraints.

Automated Market Making and Loss-Versus-Rebalancing

Milionis et al., Automated Market Making and Loss-Versus-Rebalancing, arXiv record first submitted August 2022. Abstract/metadata only; final publication/version details require pinning.

Reviewed: 2026-09-12
Access: Abstract / metadata
Status: Milionis et al., Automated Market Making and Loss-Versus-Rebalancing, arXiv record first submitted August 2022. Abstract/metadata only; final publication/version details require pinning.
Markets and data dates: Constant-function AMMs; empirical sample dates not verified.
Limitations: Distinguish LP holding-period comparisons from adverse-selection cost relative to rebalancing. Full-text fee, price-process, and reference-portfolio assumptions must be reviewed before reproducing formulas.

Stablecoins and safe asset prices

Ahmed & Aldasoro, Stablecoins and safe asset prices, BIS Working Paper 1270, initial page date 28 May 2025. Current institutional summary/abstract inspected.

Reviewed: 2026-09-12
Access: Abstract / metadata
Status: Ahmed & Aldasoro, Stablecoins and safe asset prices, BIS Working Paper 1270, initial page date 28 May 2025. Current institutional summary/abstract inspected.
Markets and data dates: Dollar-backed stablecoin flows and US Treasury yields; current page specifies daily January 2021–March 2026 data.
Limitations: Link stablecoin reserves to rates and liquidity. The landing page combines an original publication date with updated results; freeze an exact paper revision before lesson authoring. Identification depends on the instrument and local-projection design.

Decomposing Hedge Funds’ U.S. Treasury Exposures

Monin, Decomposing Hedge Funds’ U.S. Treasury Exposures, FEDS Note, 22 June 2026. Full-text sections.

Reviewed: 2026-09-12
Access: Selected full-text sections
Status: Monin, Decomposing Hedge Funds’ U.S. Treasury Exposures, FEDS Note, 22 June 2026. Full-text sections.
Markets and data dates: US hedge-fund Treasury, repo, and derivatives exposures; monthly data through September 2025.
Limitations: Compare cash-futures basis, swap spreads, and curve trades. Form PF does not directly reveal trade-level positioning; decomposition is approximate and underlying confidential records are not a public replication dataset.

The Limits of Arbitrage

Shleifer & Vishny, The Limits of Arbitrage, Journal of Finance, 1997. Publisher search excerpt only.

Reviewed: 2026-09-12
Access: Metadata / summary
Status: Shleifer & Vishny, The Limits of Arbitrage, Journal of Finance, 1997. Publisher search excerpt only.
Markets and data dates: Theoretical delegated arbitrage/capital constraints; empirical sample not applicable to the proposed conceptual use.
Limitations: Explain why divergence can force exit before convergence. Retrieve accessible full text before teaching the specific formal model; distinguish a motivating framework from current empirical evidence.

Empirical Asset Pricing via Machine Learning

Gu, Kelly & Xiu, Empirical Asset Pricing via Machine Learning, NBER WP 2018; journal publication, Review of Financial Studies, 2020. Abstract/metadata.

Reviewed: 2026-09-12
Access: Abstract / metadata
Status: Gu, Kelly & Xiu, Empirical Asset Pricing via Machine Learning, NBER WP 2018; journal publication, Review of Financial Studies, 2020. Abstract/metadata.
Markets and data dates: Equity return prediction; exact data dates and licensed inputs not verified in this review.
Limitations: Baselines, nonlinearities, feature sets, and economic evaluation. Validate data availability, chronology, turnover and costs before offering a faithful replication.

Deep Learning for Financial Time Series: A Large-Scale Benchmark of Risk-Adjusted Performance

Deep Learning for Financial Time Series: A Large-Scale Benchmark of Risk-Adjusted Performance, preprint, 2 March 2026. Full-text selected methods/results.

Reviewed: 2026-09-12
Access: Selected full-text sections
Status: Deep Learning for Financial Time Series: A Large-Scale Benchmark of Risk-Adjusted Performance, preprint, 2 March 2026. Full-text selected methods/results.
Markets and data dates: Multi-asset daily futures; annual results run 2010–2024 while some aggregate labels say 2010–2025.
Limitations: Teach benchmark scrutiny and comparable model budgets. Reconcile horizon labels and main-result cost treatment against separate breakeven-cost analysis before reproducing rankings. Validation-based seed selection must count as research selection.

The Probability of Backtest Overfitting

Bailey et al., The Probability of Backtest Overfitting, author-hosted manuscript; author publication directory. Search excerpt only for this review; exact manuscript date/publication mapping to verify.

Reviewed: 2026-09-12
Access: Metadata / summary
Status: Bailey et al., The Probability of Backtest Overfitting, author-hosted manuscript; author publication directory. Search excerpt only for this review; exact manuscript date/publication mapping to verify.
Markets and data dates: Selection diagnostics and illustrative experiments; no common market sample specified here.
Limitations: Teach multiple trials, selection bias, and diagnostics alongside chronological evaluation. Do not equate a PBO estimate with prospective strategy validation.

High-frequency trading in a limit order book

Avellaneda & Stoikov, High-frequency trading in a limit order book, Quantitative Finance, 2008, author/university copy. Search excerpt/metadata.

Reviewed: 2026-09-12
Access: Metadata / summary
Status: Avellaneda & Stoikov, High-frequency trading in a limit order book, Quantitative Finance, 2008, author/university copy. Search excerpt/metadata.
Markets and data dates: Stylized order-book market making; simulation assumptions and any empirical sample not inspected.
Limitations: Inventory-dependent quoting baseline before queue and adverse-selection extensions. Full model review and a documented fill model are prerequisites for a teaching implementation.

Deep Hedging

Bühler et al., Deep Hedging, arXiv manuscript first submitted 2018. Abstract/metadata; distinguish manuscript version from later publication.

Reviewed: 2026-09-12
Access: Abstract / metadata
Status: Bühler et al., Deep Hedging, arXiv manuscript first submitted 2018. Abstract/metadata; distinguish manuscript version from later publication.
Markets and data dates: Derivative hedging under costs/constraints; numerical setting, historical sample not verified.
Limitations: Tie ML to risk-sensitive hedging decisions. Use as an advanced comparison after analytic and simple numerical hedges; simulator success is not evidence of live hedging performance.

Time Series Momentum

Moskowitz, Ooi & Pedersen, Time Series Momentum, journal article, 2012, author-affiliated research page. Search/landing summary.

Reviewed: 2026-09-12
Access: Metadata / summary
Status: Moskowitz, Ooi & Pedersen, Time Series Momentum, journal article, 2012, author-affiliated research page. Search/landing summary.
Markets and data dates: Futures across asset classes; original study endpoints not verified here.
Limitations: Separate time-series trend from cross-sectional momentum. Pair with original-paper data in D09 and distinguish the original sample from later extensions and investable implementation.

MIT 18.642: Topics in Mathematics with Applications in Finance, Fall 2024

MIT 18.642: Topics in Mathematics with Applications in Finance, Fall 2024; 2013 predecessor

Reviewed: 2026-09-12
Access: Landing page / documentation
Status: Course / book
Markets and data dates: Resource-specific coverage; verify the selected file, version and availability dates.
Limitations: Lecture, problem-set, and application sequencing; start with newer course, use older material where helpful. Link lectures and author original companion exercises; check reuse terms per asset.

Convex Optimization

Boyd & Vandenberghe, Convex Optimization

Reviewed: 2026-09-12
Access: Landing page / documentation
Status: Course / book
Markets and data dates: Resource-specific coverage; verify the selected file, version and availability dates.
Limitations: Constraints, convexity, duality, portfolio allocation, execution, and arbitrage feasibility. Publicly readable book with examples; public access is not unrestricted republication permission.

An Introduction to Statistical Learning

James et al., An Introduction to Statistical Learning

Reviewed: 2026-09-12
Access: Landing page / documentation
Status: Course / book
Markets and data dates: Resource-specific coverage; verify the selected file, version and availability dates.
Limitations: Python edition, 2023: regression, classification, resampling, regularization and model comparison. Financial chronology requires adaptations beyond generic chapter exercises.

Forecasting: Principles and Practice, 3rd edition

Hyndman & Athanasopoulos, Forecasting: Principles and Practice, 3rd edition

Reviewed: 2026-09-12
Access: Landing page / documentation
Status: Course / book
Markets and data dates: Resource-specific coverage; verify the selected file, version and availability dates.
Limitations: Forecast evaluation, intervals, decomposition, benchmark forecasting. Translate educational examples to the site's self-contained Python convention; preserve correct attribution.

QuantEcon: Intermediate Quantitative Economics with Python

Reviewed: 2026-09-12
Access: Landing page / documentation
Status: Course / book
Markets and data dates: Resource-specific coverage; verify the selected file, version and availability dates.
Limitations: Dynamic programming, Markov processes, numerical economics and executable explanations. Choose prerequisites carefully; some lessons exceed beginner scope.

Seeing Theory

Brown University, Seeing Theory

Reviewed: 2026-09-12
Access: Landing page / documentation
Status: Course / book
Markets and data dates: Resource-specific coverage; verify the selected file, version and availability dates.
Limitations: Visual inspiration for probability, distributions, conditioning, and regression. Build original figures with accessible numeric alternatives; visual appeal alone does not demonstrate learning benefit.

CME Institute education

Reviewed: 2026-09-12
Access: Landing page / documentation
Status: Course / book
Markets and data dates: Resource-specific coverage; verify the selected file, version and availability dates.
Limitations: Futures mechanics, contract specifications, basis, margin, commodity and rates examples, videos. Exchange education explains its products; it is not independent evidence of strategy performance.

Cboe Options Institute

Reviewed: 2026-09-12
Access: Landing page / documentation
Status: Course / book
Markets and data dates: Resource-specific coverage; verify the selected file, version and availability dates.
Limitations: Options payoffs, exercise and settlement, index versus equity contracts, short instructional videos. Product conventions must be checked against the exact contract/date.

Kenneth French Data Library

Reviewed: 2026-09-12
Access: Landing page / documentation
Status: Dataset
Markets and data dates: Resource-specific coverage; verify the selected file, version and availability dates.
Limitations: Factors, portfolios and international research returns; dates vary by file. The page documents the January 2025 switch from CRSP FIZ to CIZ and changed dividend reinvestment timing. Pin data vintage and return definition; factor portfolios are not executable constituent-level histories.

Open Source Asset Pricing

Reviewed: 2026-09-12
Access: Landing page / documentation
Status: Dataset
Markets and data dates: Resource-specific coverage; verify the selected file, version and availability dates.
Limitations: Published-signal replication and transparency. Check what is publicly distributed versus requiring underlying licensed equity/fundamental data. Good starting point for a failed-replication exercise.

FRED/ALFRED API documentation

Reviewed: 2026-09-12
Access: Landing page / documentation
Status: Dataset
Markets and data dates: Resource-specific coverage; verify the selected file, version and availability dates.
Limitations: Macro series and real-time vintages; series-specific coverage. Distinguish observation date, release date, revision vintage and download date. Check series-level third-party rights and API access requirements.

SEC EDGAR APIs

Reviewed: 2026-09-12
Access: Landing page / documentation
Status: Dataset
Markets and data dates: Resource-specific coverage; verify the selected file, version and availability dates.
Limitations: Company submissions and XBRL facts. Preserve filing availability times, amendments and units; avoid joining a fiscal period end as if all facts were known then. Respect published access policy.

Binance public data repository

Reviewed: 2026-09-12
Access: Landing page / documentation
Status: Dataset
Markets and data dates: Resource-specific coverage; verify the selected file, version and availability dates.
Limitations: Exchange-maintained spot/futures trade and bar archives. Coverage depends on instrument/file; verify timestamp units, checksum and schema changes. One venue and candles alone cannot establish cross-venue arbitrage execution.

Coin Metrics documentation

Reviewed: 2026-09-12
Access: Landing page / documentation
Status: Dataset
Markets and data dates: Resource-specific coverage; verify the selected file, version and availability dates.
Limitations: Definitions for network and market metrics. Distinguish community coverage from commercial datasets and derived metrics from raw events; definitions and asset coverage need versioning.

CFTC Commitments of Traders

Reviewed: 2026-09-12
Access: Landing page / documentation
Status: Dataset
Markets and data dates: Resource-specific coverage; verify the selected file, version and availability dates.
Limitations: Positioning reports across futures markets. Model reporting/publication lag, categories and revisions. Aggregated reports do not reveal a particular trader's signal or intraday positioning.

New York Fed data and indicators

Reviewed: 2026-09-12
Access: Landing page / documentation
Status: Dataset
Markets and data dates: Resource-specific coverage; verify the selected file, version and availability dates.
Limitations: Rates, macro-finance indicators and model estimates; choose specific series and vintage later. Estimated indicators should not be presented as directly observed market prices.

AQR data sets

AQR data sets and Time Series Momentum original-paper data

Reviewed: 2026-09-12
Access: Landing page / documentation
Status: Dataset
Markets and data dates: Resource-specific coverage; verify the selected file, version and availability dates.
Limitations: Strategy/factor research series and original-paper versus extended samples. Verify gross/net, construction assumptions and use terms; research series are not fund returns.

Polymarket documentation

Reviewed: 2026-09-12
Access: Landing page / documentation
Status: Market documentation
Markets and data dates: Resource-specific coverage; verify the selected file, version and availability dates.
Limitations: Event/market distinctions, outcome tokens, order books, negative-risk structures, resolution and fees. The landing page distinguishes US documentation; do not assume the same venue/entity/rules across products.

Kalshi documentation

Reviewed: 2026-09-12
Access: Landing page / documentation
Status: Market documentation
Markets and data dates: Resource-specific coverage; verify the selected file, version and availability dates.
Limitations: Event-contract data and market metadata; documented APIs and demo environment. Exact contract resolution terms and economic equivalence must be checked separately for cross-venue exercises.

Uniswap documentation

Reviewed: 2026-09-12
Access: Landing page / documentation
Status: Market documentation
Markets and data dates: Resource-specific coverage; verify the selected file, version and availability dates.
Limitations: Separate v2, v3 and v4 mechanics; liquidity ranges, swaps, ticks and hooks. The algebra of a simple constant-product example does not describe every pool version.

Deribit API documentation

Reviewed: 2026-09-12
Access: Landing page / documentation
Status: Market documentation
Markets and data dates: Resource-specific coverage; verify the selected file, version and availability dates.
Limitations: Instruments, options and futures market data. Record payoff currency, multiplier, expiry and quote units for each selected instrument; APIs alone are not a backtest dataset.

Ethereum: maximal extractable value

Reviewed: 2026-09-12
Access: Landing page / documentation
Status: Market documentation
Markets and data dates: Resource-specific coverage; verify the selected file, version and availability dates.
Limitations: Ordering, arbitrage, liquidations and transaction-level frictions. Teach mechanism and user execution consequences; theoretical atomic payoff does not remove failed-transaction costs or other economic risks.

Flashbots documentation

Reviewed: 2026-09-12
Access: Landing page / documentation
Status: Market documentation
Markets and data dates: Resource-specific coverage; verify the selected file, version and availability dates.
Limitations: Transaction supply chain and MEV infrastructure. Use for architectural diagrams and research context; infrastructure provider material has its own perspective and is not a profitability benchmark.

Active learning increases student performance in science, engineering, and mathematics

Freeman et al., Active learning increases student performance in science, engineering, and mathematics, PNAS, 2014; research synthesis, abstract/method excerpt reviewed.

Reviewed: 2026-09-12
Access: Abstract / metadata
Status: Learning design
Markets and data dates: Resource-specific coverage; verify the selected file, version and availability dates.
Limitations: Supports prioritizing active problems and feedback. Undergraduate STEM evidence does not by itself establish an effect size for this self-paced finance site; evaluate the proposed modalities locally.

Making Audio and Video Media Accessible

W3C WAI, Making Audio and Video Media Accessible, living guidance; landing/guidance reviewed.

Reviewed: 2026-09-12
Access: Landing page / documentation
Status: Learning design
Markets and data dates: Resource-specific coverage; verify the selected file, version and availability dates.
Limitations: Captions, transcripts, descriptions of essential visual information, and accessible playback are production requirements for media. This is accessibility guidance, not a study of learning effectiveness.

Binance USDⓈ-M: funding history, funding intervals and market metadata

Official funding-history and funding-info sections specify timestamps, pagination, associated mark prices and interval/cap adjustments.

Reviewed: 2026-09-12
Access: Selected documentation sections
Status: Living official documentation; not academic performance evidence
Markets and data dates: Crypto perpetual markets supported by the named venue/tool; no historical dataset downloaded or tested.
Limitations: Endpoint windows and product conventions differ. Current interval settings do not establish historical settings; sample dates and missing intervals must be audited. Documentation examples are not a dataset.

Bybit V5: order creation, position modes and asynchronous acknowledgement

Official order interface covers linear/inverse contracts, client order IDs, position mode, reduce-only behavior and asynchronous request acceptance.

Reviewed: 2026-09-12
Access: Selected documentation sections
Status: Living official documentation; not academic performance evidence
Markets and data dates: Crypto perpetual markets supported by the named venue/tool; no historical dataset downloaded or tested.
Limitations: An accepted request is not a confirmed fill. Market-order handling and account modes are venue-specific; testnet behavior cannot establish live liquidity or execution quality.

Bybit V5: private order stream and cancel/fill races

Order-stream documentation explicitly describes two Filled status messages that can arise when execution and cancellation race.

Reviewed: 2026-09-12
Access: Selected documentation sections
Status: Living official documentation; not academic performance evidence
Markets and data dates: Crypto perpetual markets supported by the named venue/tool; no historical dataset downloaded or tested.
Limitations: An order-status stream alone is not a complete cash ledger. Reconcile executions, fees, funding and balances after reconnects; this documentation review did not test delivery guarantees.

Hyperliquid funding: payment timing, price reference and contract distinctions

The technical section distinguishes the rate formula horizon from payment frequency and identifies the oracle-price reference used to calculate payment notional.

Reviewed: 2026-09-12
Access: Selected documentation sections
Status: Living official documentation; not academic performance evidence
Markets and data dates: Crypto perpetual markets supported by the named venue/tool; no historical dataset downloaded or tested.
Limitations: Do not copy an eight-hour rate into an hourly ledger. HIP-3 conventions can differ; resolve the technical formula, contract configuration and worked-example horizon before implementation. No current rate or return is predicted.

Hyperliquid: API wallets, signing identities and nonce coordination

Official guidance separates the signing API wallet from the account queried for state and explains nonce sharing across processes and subaccounts.

Reviewed: 2026-09-12
Access: Selected documentation sections
Status: Living official documentation; not academic performance evidence
Markets and data dates: Crypto perpetual markets supported by the named venue/tool; no historical dataset downloaded or tested.
Limitations: Nonce pruning and wallet reuse have operational consequences. Documentation is not a credential-management implementation; concurrent processes and recovery need explicit testing.

Hyperliquid exchange API: client IDs, reduce-only orders and scheduled cancel

Selected order, cancel-by-client-ID and scheduleCancel sections describe order controls and a scheduled cancel-all mechanism.

Reviewed: 2026-09-12
Access: Selected documentation sections
Status: Living official documentation; not academic performance evidence
Markets and data dates: Crypto perpetual markets supported by the named venue/tool; no historical dataset downloaded or tested.
Limitations: Scheduled cancellation removes open orders; it does not close existing positions or guarantee a flat portfolio. Trigger limits and timing constraints must be checked against the current API.

Hyperliquid API: request weights, connection limits and cancellation capacity

Official limits distinguish IP-weighted requests, WebSocket resources and address-based action budgets.

Reviewed: 2026-09-12
Access: Selected documentation sections
Status: Living official documentation; not academic performance evidence
Markets and data dates: Crypto perpetual markets supported by the named venue/tool; no historical dataset downloaded or tested.
Limitations: Batching can count differently under different limits. Do not bake this review date’s numerical limits into a permanent strategy assumption; review them for the actual account and traffic pattern.

CCXT documentation: unified exchange access and adapter references

Official documentation index links market definitions, pagination, orders, error handling and WebSocket guidance for a unified exchange interface.

Reviewed: 2026-09-12
Access: Landing page / documentation index
Status: Living official documentation; not academic performance evidence
Markets and data dates: Crypto perpetual markets supported by the named venue/tool; no historical dataset downloaded or tested.
Limitations: Documentation landing/index reviewed; the manual URL could not be retrieved in this pass. A unified method name does not verify identical venue behavior. Pin a package release and inspect the specific connector before implementation.

Hummingbot Strategy V2: funding-rate arbitrage sample script

Maintainer-hosted Python example compares normalized funding rates, estimates entry costs, creates paired position executors and records funding payments.

Reviewed: 2026-09-12
Access: Selected source-code sections
Status: Maintainer example implementation; not a research paper
Markets and data dates: Crypto perpetual markets supported by the named venue/tool; no historical dataset downloaded or tested.
Limitations: Selected source sections inspected and matched byte-for-byte to the pinned commit; not installed, backtested or audited. Quoted rates are extrapolated and connector intervals need validation. A sample script is not evidence of profitable automation.

Hummingbot Strategy V2: executor lifecycle and orchestration

Executors manage discrete order workflows; the orchestrator creates, stops and reports on those workflows.

Reviewed: 2026-09-12
Access: Selected documentation sections
Status: Living official documentation; not academic performance evidence
Markets and data dates: Crypto perpetual markets supported by the named venue/tool; no historical dataset downloaded or tested.
Limitations: Framework abstractions do not remove partial fills, stale data, connector errors or residual positions. Specific strategy/connector coverage and software release were not tested.

Freqtrade futures: missing funding, margin assumptions and liquidation accounting

Official futures documentation identifies account-isolation assumptions, missing funding histories and limitations in liquidation-fee and cross-position simulation.

Reviewed: 2026-09-12
Access: Selected documentation sections
Status: Living official documentation; not academic performance evidence
Markets and data dates: Crypto perpetual markets supported by the named venue/tool; no historical dataset downloaded or tested.
Limitations: Substituting zero for missing funding can misstate results; zero is not an observed rate. The documented liquidation and shared-account limitations must be incorporated into the experiment design.

Freqtrade lookahead analysis: detect future-dependent signals

The diagnostic compares baseline and sliced backtest behavior to detect indicator and signal dependence on future information.

Reviewed: 2026-09-12
Access: Selected documentation sections
Status: Living official documentation; not academic performance evidence
Markets and data dates: Crypto perpetual markets supported by the named venue/tool; no historical dataset downloaded or tested.
Limitations: Untriggered signal paths are not verified and can produce false negatives; limit-order/custom-price behavior can introduce false positives. Passing the diagnostic is not proof of a causal or profitable strategy.

HftBacktest: partial fills, queue models and replay limitations

Official simulator documentation explains partial/no-partial-fill exchange models and queue-position assumptions in market-data replay.

Reviewed: 2026-09-12
Access: Selected documentation sections
Status: Living official documentation; not academic performance evidence
Markets and data dates: Crypto perpetual markets supported by the named venue/tool; no historical dataset downloaded or tested.
Limitations: Replay cannot change historical order books, so market impact is absent and liquidity-taking fills can be unrealistic. Funding, margin and liquidation accounting require separate validation.

NautilusTrader: event-driven backtesting and execution-model reading guide

The official overview maps historical event processing to shared strategy components and links data, matching, fills, account and margin models.

Reviewed: 2026-09-12
Access: Overview / reading guide
Status: Living official documentation; not academic performance evidence
Markets and data dates: Crypto perpetual markets supported by the named venue/tool; no historical dataset downloaded or tested.
Limitations: Overview and reading guide reviewed; linked subchapters were not comprehensively audited. Sharing components with live trading does not establish equivalent fills or exchange behavior.

Point-in-Time Audit Before Alpha: Public-Archive Availability and a Negative Matched-Budget Study on BTC Perpetual Futures

Zeng, Yang, Han and He audit whether public archives were usable at each decision. Their matched-budget adaptive search did not establish superiority over random search, and evaluated historical-holdout runs had positive information coefficients but negative net Sharpe under primary costs.

Reviewed: 2026-09-12
Access: Selected full-text sections: abstract, data admission, evaluation and limitations
Status: Preprint; no journal publication verified
Markets and data dates: Binance BTCUSDT USD-M perpetuals; archive interval [2024-08-01, 2026-08-01), five-minute decisions; revised admission retained 727 complete UTC days.
Limitations: The holdout is retrospective: the protocol was locked after the historical interval, not prospectively. Funding availability uses an assumed delay, open interest is optional because publication time was unverified, and known-template audit success does not detect every possible leakage mechanism.

AutoQuant: An Auditable Expert-System Framework for Execution-Constrained Auto-Tuning in Cryptocurrency Perpetual Futures

Deng presents auditable configuration selection within a fixed signal family, combining delayed execution, funding alignment, cost scenarios and staged screening. The protocol exposes configuration fragility; it does not guarantee higher returns.

Reviewed: 2026-09-12
Access: Selected full-text sections: abstract, sample definitions, execution and evaluation
Status: arXiv manuscript with a listed Expert Systems with Applications article reference, 133924; publisher edition not directly retrieved
Markets and data dates: BTC/USDT 4-hour core [2019-09-08, 2021-12-31), extended through 2025-10-14; ETH through 2025-11-24 and SOL/AVAX through 2025-12-16. Training, screening and holdout windows differ and are separately specified.
Limitations: Small-account simulations use linear costs and exclude market impact/institutional capacity. Screening reuses validation information. Funding has a disclosed fallback, venue snapshots require pinning, and results are conditional on the chosen signal family and cost conventions.

Optimal Liquidation of Perpetual Contracts

Donnelly, Lin and Lorig study planned inventory unwinding with funding, market impact, inventory risk and terminal penalties. Linear payoff exposure admits an analytic strategy; nonlinear exposure uses stated approximations.

Reviewed: 2026-09-12
Access: Selected full-text sections: abstract, setup and execution interpretation
Status: Preprint; manuscript journal field says TBA
Markets and data dates: Stochastic-control model and numerical illustrations of perpetual inventory liquidation; no historical trading sample verified.
Limitations: Liquidation here means planned position unwinding, not an exchange forced-liquidation engine. The funding/impact model and approximation regimes are restrictive; neither closed-form optimality nor numerical illustrations establish live execution quality.

Exploring risk and return profiles of funding rate arbitrage on CEX and DEX

The publisher excerpts describe a funding-arbitrage backtest across centralized and decentralized venues, comparing leverage and holding benchmarks.

Reviewed: 2026-09-12
Access: Search-indexed primary publisher abstract and section excerpts; direct article retrieval failed
Status: Journal research article in Blockchain: Research and Applications, article 100354, according to the primary publisher index; exact publication date unverified
Markets and data dates: Binance, BitMEX, ApolloX and Drift; BTC, ETH, XRP, BNB and SOL; six-month analysis is described, but exact observation endpoints were not verified.
Limitations: Access is limited to indexed publisher material. Do not adopt the reported best-performing return or drawdown as an attainable expectation. Verify sample endpoints, liquidation handling, capital denominator, risk-free benchmark and cost/market-impact assumptions in the full paper first.

A Shared Template Without Shared Feedback: Funding Rates in Cryptocurrency Perpetual Futures

Zhang studies why the same nominal funding template can produce different effective feedback when arbitrage capital, leverage limits, averaging, flat regions and caps interact.

Reviewed: 2026-09-12
Access: Search-indexed primary SSRN abstract/metadata; direct page and PDF retrieval blocked
Status: Working-paper record on SSRN; journal publication not verified
Markets and data dates: Model with an empirical illustration covering 200 Binance USDT-margined perpetuals according to the indexed abstract; exact sample dates unverified.
Limitations: This is an access-limited candidate, not a reviewed empirical result. Retrieve and pin the exact manuscript, sample and identification design before implementing its model or treating associations with directional liquidation as causal evidence.

Crane, Karra & Soto: Total Recall? Evaluating the Macroeconomic Knowledge of Large Language Models

Crane, Karra & Soto: Total Recall? Evaluating the Macroeconomic Knowledge of Large Language Models · Federal Reserve Finance and Economics Discussion Series working paper 2025-044 · June 2025; later versions not established

Reviewed: 2026-09-12
Access: Abstract on Federal Reserve 2025 research index only
Status: Federal Reserve Finance and Economics Discussion Series working paper 2025-044
Markets and data dates: Macroeconomic data and release dates; Exact series, date ranges and model snapshots not inspected
Limitations: Motivates explicit vintage and availability checks. The lesson does not quantify model error rates or reproduce the study.

Federal Reserve Bank of St. Louis: FRED API real-time periods

Federal Reserve Bank of St. Louis: FRED API real-time periods · Living primary provider documentation · Reviewed 12 September 2026

Reviewed: 2026-09-12
Access: Documentation page
Status: Living primary provider documentation
Markets and data dates: Series-specific macroeconomic data; Series-dependent; no dataset downloaded
Limitations: API semantics do not establish the exact intraday availability of every vintage; source release timestamps and time zones must be verified separately.

Bongiorno, Manolakis & Mantegna: End-to-End Large Portfolio Optimization for Variance Minimization with Neural Networks through Covariance Cleaning

Bongiorno, Manolakis & Mantegna: End-to-End Large Portfolio Optimization for Variance Minimization with Neural Networks through Covariance Cleaning · Journal of Finance and Data Science 12 (2026), 100179; journal reference verified on author manuscript record · v3, 21 April 2026

Reviewed: 2026-09-12
Access: Abstract and version/publication metadata only
Status: Journal of Finance and Data Science 12 (2026), 100179; journal reference verified on author manuscript record
Markets and data dates: US equities; Abstract reports out-of-sample January 2000–December 2024; training and point-in-time data details not inspected
Limitations: Motivates comparing covariance estimators under common constraints. Architecture, execution assumptions and claimed rankings were not reproduced; this module teaches transparent baseline calculations.

Banerjee, Boneva, Pinter & Sushko: Monetary policy transmission to exchange rates: the role of currency carry trades

Banerjee, Boneva, Pinter & Sushko: Monetary policy transmission to exchange rates: the role of currency carry trades · BIS Bulletin 124; institutional primary research · 6 May 2026

Reviewed: 2026-09-12
Access: Publication summary only; PDF methods not inspected
Status: BIS Bulletin 124; institutional primary research
Markets and data dates: FX and leveraged currency positions; Exact estimation endpoints not available in inspected summary
Limitations: Motivates stress testing exchange rates jointly with leveraged positioning. It is not a recommendation to hold carry or evidence of executable arbitrage.

Du, Nesmith & Heppe: Does Financial Stress Affect Commodity Futures Traders’ Positions?

Du, Nesmith & Heppe: Does Financial Stress Affect Commodity Futures Traders’ Positions? · Federal Reserve working paper 2025-082 · September 2025, revised November 2025

Reviewed: 2026-09-12
Access: Abstract on Federal Reserve research index only
Status: Federal Reserve working paper 2025-082
Markets and data dates: US commodity futures; 2008 global financial crisis and COVID-19 crisis; exact window endpoints not inspected
Limitations: Different crisis mechanisms motivate multiple stress paths; aggregate position findings are not causal forecasts of individual futures returns.

Aldasoro, Beltrán & Grinberg — Stablecoin flows and spillovers to FX markets

Aldasoro, Beltrán & Grinberg — Stablecoin flows and spillovers to FX markets · BIS Working Paper 1340, 27 March 2026

Reviewed: 2026-09-12
Access: Primary publisher summary and abstract checked 2026-09-12
Status: BIS Working Paper 1340, 27 March 2026
Markets and data dates: Four USD stablecoins, 27 fiat currencies, 64 exchanges, 2021–2025
Limitations: Motivates studying stablecoin conversion as a segmented FX route with balance-sheet constraints. The identification and quantitative estimates require full-paper review; a parity gap is not a frictionless trading profit.

Polymarket — How positions work

Polymarket — How positions work · Living official product documentation

Reviewed: 2026-09-12
Access: Documentation inspected 2026-09-12
Status: Living official product documentation
Markets and data dates: Conditional outcome-token mechanics; no empirical sample
Limitations: Operational starting point for split, merge and redemption workflows. Record the exact market rules and collateral/version before applying a generic binary model; platform behavior may change.

Milionis et al. — Automated Market Making and Loss-Versus-Rebalancing

Milionis et al. — Automated Market Making and Loss-Versus-Rebalancing · arXiv manuscript v6, revised 20 August 2026; publication status beyond this record not checked

Reviewed: 2026-09-12
Access: Abstract-only review 2026-09-12
Status: arXiv manuscript v6, revised 20 August 2026; publication status beyond this record not checked
Markets and data dates: Continuous-time model and Uniswap v2 ETH–USDC illustration; empirical sample dates not verified
Limitations: Separates inventory market exposure from fees and adverse selection against a rebalancing benchmark. Abstract-only access; no empirical magnitude is used here. Discrete block trading, fees and oracle changes alter a continuous model.

Uniswap — V2 pricing

Uniswap — V2 pricing · Official protocol documentation, V2

Reviewed: 2026-09-12
Access: Documentation inspected 2026-09-12
Status: Official protocol documentation, V2
Markets and data dates: Constant-product pools; no historical sample
Limitations: Operational reference for reserve-based swap arithmetic and external price checks. The lessons label simplified fee and execution assumptions; other pool versions require different formulas.

Aave — V3 overview

Aave — V3 overview · Official protocol documentation, V3

Reviewed: 2026-09-12
Access: Overview inspected 2026-09-12
Status: Official protocol documentation, V3
Markets and data dates: Lending protocol mechanics; no empirical sample
Limitations: Starting point for collateral, debt and health-factor research. Parameters differ by network, asset and governance state. Examples are generic lending models, not copied live parameters.

Daian et al. — Flash Boys 2.0

Daian et al. — Flash Boys 2.0 · Foundational arXiv research record, 2019; subsequent venue status not checked in this pass

Reviewed: 2026-09-12
Access: Abstract-only review 2026-09-12
Status: Foundational arXiv research record, 2019; subsequent venue status not checked in this pass
Markets and data dates: Ethereum decentralized exchange ordering; sample endpoints not checked
Limitations: Explains why transaction ordering belongs in execution analysis. Historical mechanisms cannot be assumed to match current proposer/builder markets; use alongside the recent LVR checkpoint.

François, Gauthier, Godin & Pérez-Mendoza: Deep Hedging with Options Using the Implied Volatility Surface

François, Gauthier, Godin & Pérez-Mendoza: Deep Hedging with Options Using the Implied Volatility Surface · arXiv research preprint; no journal status established by this review · v3, 12 August 2025

Reviewed: 2026-09-12
Access: Abstract and version metadata only
Status: arXiv research preprint; no journal status established by this review
Markets and data dates: S&P 500 index options and simulated markets; Current abstract reports historical out-of-sample straddles 2020–2023; training dates not inspected
Limitations: Version 1 search excerpts describe a different data window. This record follows v3. Claimed hedge rankings are not taught as established results; simulator, trading costs and data construction require full-text review and replication.

Gatheral & Jacquier: Arbitrage-free SVI volatility surfaces

Gatheral & Jacquier: Arbitrage-free SVI volatility surfaces · Foundational research manuscript; publication mapping not verified here · v4, 21 March 2013

Reviewed: 2026-09-12
Access: Abstract and version metadata only
Status: Foundational research manuscript; publication mapping not verified here
Markets and data dates: SPX options illustration; Exact quote date not inspected
Limitations: Supports the importance of static arbitrage constraints; the elementary convexity check below is not a full SVI calibration or sufficient global surface validation.

Banerjee, Cordova, De Pooter & Grishchenko: Gauging the Sentiment of FOMC Communications through the Eyes of the Financial Press

Banerjee, Cordova, De Pooter & Grishchenko: Gauging the Sentiment of FOMC Communications through the Eyes of the Financial Press · Federal Reserve Finance and Economics Discussion Series working paper 2025-048 · July 2025

Reviewed: 2026-09-12
Access: Abstract on Federal Reserve research index only
Status: Federal Reserve Finance and Economics Discussion Series working paper 2025-048
Markets and data dates: FOMC-related financial press and asset prices; Abstract gives May 1999–November 2022
Limitations: The study motivates separating text measurement from event timing. Association with prices does not establish a tradable signal; article availability, revisions and dictionary construction were not inspected.