SSRN
Recently Published
Quantitative
Cryptocurrency Replication via Machine Learning: The paper presents a machine learning method to create synthetic cryptocurrency portfolios that mimic the risk-adjusted return profile of cryptocurrencies, providing a safer investment option. (2023-08-26, shares: 3.0)
News Data's Impact on Trading Decisions: The paper suggests a reinforcement learning approach for high-frequency algorithmic trading in futures market using news and price data, tested on the NIFTY 50 index. (2023-08-25, shares: 2.0)
Subscribe now to also get access to www.ml-quant.com in addition to the weekly letter.
Real Returns of Mutual Fund Investors: The paper finds that reported mutual fund returns in China are influenced by subscription and redemption activities, resulting in a lower actual gain coefficient for investors. (2023-08-27, shares: 2.0)
ETF Options Strategies: The research compares various option strategies for Exchange-Traded Funds (ETFs) to the Buy and Hold approach, highlighting the need for strategies to match investor risk profiles and financial goals. (2023-08-28, shares: 3.0)
Barrier Options in Stochastic Volatility Framework: The article introduces a formula for calculating single barrier options in a stochastic volatility framework, not dependent on the model's form or parameters. (2023-08-30, shares: 46.0)
Portfolio Optimization using Machine Learning: The research shows that machine learning models can be used to devise investment strategies and construct optimal portfolios, performing better than traditional strategies on the Mexican Stock Exchange. (2023-08-26, shares: 5.0)
CFO Pay and Hedging: The study explores the impact of a CFO's risk-taking incentives and equity compensation on the hedging strategy of US oil and gas companies. (2023-08-24, shares: 3.0)
Financial
Deep Learning for Derivatives Pricing Study: The research proposes two ways to learn the price of derivatives using neural networks, focusing on price differences and differences between prices of derivatives based on different asset prices. (2023-08-27, shares: 3.0)
Vega-Hedging Costs for Exotic Products: The article suggests a method to calculate future hedging costs of complex financial products using a risk projection algorithm and simulations. (2023-08-24, shares: 16.0)
High Frequency Trading Activity Identification: The article evaluates the reliability of commonly used indicators to identify high frequency traders, showing variations in performance and suggesting that unscaled proxies are more effective at indicating true HFT activity. (2023-08-24, shares: 4.0)
Power Sorting for Factor Performance: The article suggests a new method for creating characteristic-based equity factors called power sorting, showing its superior performance and applicability to multifactor strategies. (2023-08-25, shares: 12.0)
CDS and Options' Transition Risk Information: The research uses CDS and put option prices to gauge the likelihood and impact of climate change policy-related transition risk events, creating a reliable CDS-based index. (2023-08-24, shares: 4.0)
Financial Flexibility and Equity Risk: The research connects limited financial flexibility to levered risk premiums, emphasizing leverage gaps and targets, and argues that leverage alone provides limited insights. (2023-08-27, shares: 3.0)
Leasing and Idiosyncratic Volatility Puzzle: The study explores the negative correlation between past volatility and stock returns, finding that firms with higher volatility tend to lease more capital. (2023-08-29, shares: 2.0)
Sparse Index Tracking Portfolio Asset Selection: The article introduces new methods for asset selection in sparse index tracking portfolios, highlighting its benefits over traditional full replication portfolios. (2023-08-25, shares: 2.0)
Passive S&P 500 Funds' Discretionary Investing: The article challenges the belief that passive index funds strictly mimic their underlying index, showing that S&P 500 index funds do not legally commit to this. (2023-08-27, shares: 4.0)
IPO Outcomes of Disruptive Innovation: The paper presents a new text-based measure of disruptive innovation, developed through machine learning and textual analysis of IPO prospectuses, which accurately predicts IPO results and challenges the hype hypothesis about tech stocks. (2023-08-29, shares: 2.0)
Recently Updated
Quantitative
Machine Learning for Economic Recession Prediction: The paper uses machine learning to predict US economic recessions using market sentiment and economic indicators, using the ARIMA method for backcasting. (2023-06-20, shares: 2.0)
Banks' Risk Hedging for Fixed-Income Securities: The study indicates that banks meet timing requirements for discretionary hedging of fixed-income security and funding risks, but the effectiveness of these strategies is uncertain. (2023-04-19, shares: 2.0)
Generalizing Event Studies: Dollar Tree-Family Dollar Acquisition: The paper suggests a new empirical approach using a synthetic control method and machine learning to compare outcomes in mergers and acquisitions research. (2021-06-07, shares: 134.0)
Financialization-Assetization in Energy Industry: The paper explores the growing influence of financial markets and actors in the energy industry, a trend known as financialization and assetization. (2023-06-08, shares: 3.0)
Nonbanks' Leverage and Liquidity: Despite high-risk assets and short-term leverage, Nonbank mortgage companies maintain low bankruptcy rates by quickly reducing operating costs and financing after negative shocks. (2023-03-20, shares: 2.0)
Financial
Statistical Arbitrage with Deep Learning: High Returns and Outperformance: The research introduces a new data-driven method for statistical arbitrage that uses machine learning to maximize risk-adjusted returns, outperforming all benchmark approaches in a study on daily US equities. (2021-06-08, shares: 2.0)
ML Beats Benchmark Models in Stock Beta Estimation: Machine learning models, especially random forests, are more effective than traditional models in predicting market trends and reducing errors, improving market-neutral strategies and minimum variance portfolios. (2021-10-01, shares: 2.0)
Corporate Bond Pricing: Implications of Institutional Demand: The research indicates that corporate bond pricing is significantly affected by institutional demand, with different institutions having different preferences for maturity credit risk and liquidity. (2021-01-15, shares: 1744.0)
Index-Linked Trading: Differences in Returns: The study reveals that index-linked trading leads to differences in returns and volatilities, with firm fundamentals influencing risk premiums in expected returns. (2021-04-26, shares: 241.0)
Alumni Ties and Private Equity Investments: Impact on Endowment Choices: The research finds that university endowments tend to invest in private equity funds managed by their alumni, but this doesn't necessarily lead to better performance. (2023-03-13, shares: 152.0)
Machine Learning Predictability: Factors Affecting Stock Return Forecasts: The study shows that while machine learning strategies can predict short-term returns for small firms and early historical data, they have not provided significant economic gains for most of the U.S. market in the past 20 years. (2023-08-18, shares: 7.0)
Hedge Funds' Demand Linked to Future Stock Returns: The correlation between short-term institutional demand and future stock returns is only significant in hedge funds, not in other institutions with short investment horizons. (2022-04-06, shares: 72.0)
Pricing Errors Impact Options Prices & Greeks: Pricing errors in the base asset can inflate options prices and affect option Greeks, leading to inefficient risk management and hedging if not considered. (2021-06-15, shares: 2.0)
Tradeoff Between Tracking Error & Turnover in Portfolio Rebalancing: There's a balance between tracking error and turnover in rebalancing choices, with calendar-based methods being less efficient than tolerance bands, and no proof that rebalancing choices can consistently boost expected returns. (2021-06-03, shares: 2.0)
ArXiv
Finance
Insights from Level 3 Order Book Data: Empirical Analysis on Financial Market: The study presents a new physical model, inspired by statistical physics, that uses Level 3 order book data to predict price volatility and expected returns, surpassing traditional and machine learning methods by incorporating principles of statistical physics. (2023-08-28, shares: 4)
GPU-Accelerated LOB Simulator: JAX-LOB: The paper introduces JAX-LOB, the first GPU-powered limit order book simulator capable of processing multiple books simultaneously, designed for efficient large-scale simulations of LOB dynamics for research, calibration, and reinforcement learning training. (2023-08-25, shares: 5)
Quantum Stock Price Prediction: The article investigates the use of Quantum Algorithms in predicting stock prices of companies like Apple and Visa, comparing the accuracy of these quantum models with traditional models. (2023-08-25, shares: 4)
Joint Calibration of Volatility Models: The authors present a non-parametric method for joint calibration of a volatility model and a correlated stochastic short rate model, demonstrating its effectiveness on market data and comparing it with sequential calibration. (2023-08-28, shares: 3)
Miscellaneous
Few-Shot Text Classification for Finance: Conversational GPT models are suggested for efficient text classification in finance, providing a practical solution for tasks with limited labels and achieving top-tier results. (2023-08-28, shares: 2)
Deep Algorithm for Nonlinear Equations: A new deep learning algorithm has been developed to solve complex mathematical equations, offering improved accuracy and less complexity than similar models. (2023-08-28, shares: 5)
TimeTrail: Financial Fraud Patterns: TimeTrail, a new technique for detecting financial fraud, uses advanced analysis to explain fraud patterns, outperforming traditional methods in accuracy and interpretability. (2023-08-27, shares: 2)
Crypto & Blockchain
Fee Mechanism for Proof-of-Stake Protocol: The research expands the transaction fee system in blockchain's proof-of-stake protocol, adding a long-term utility model for miners and a new parameter for user-miner incentives and contract validity. (2023-08-26, shares: 9)
Grover Search for Portfolio Selection: The study presents explicit oracles for Grover's algorithm to align with investor preferences, possibly choosing portfolios with optimal Sharpe ratios, tested using quantum simulators. (2023-08-24, shares: 4)
Historical Trending
Neural Networks for Market Risk Scenario Generation: The study extends the use of generative adversarial networks (GANs) to a full internal market risk model, suggesting that GAN-based models can be a data-driven alternative for market risk modeling. (2021-09-21, shares: 31)
Interpretability of LSTM Models for Stock Prediction: The research examines the effect of correlated features on the interpretability of LSTM models for predicting oil company stocks, concluding that adding a correlated feature does not enhance the interpretability of these models. (2022-01-02, shares: 22)
Solvability of Differential Riccati Equations in Algorithmic Trading: The research investigates a differential Riccati equation with indefinite matrix coefficients, using it to address two algorithmic trading issues with a constant absolute risk-aversion utility function. (2022-02-15, shares: 14)
Automated Market Makers: Mathematical Framework: The research introduces a new framework for Automated Market Makers, suggesting a unique fee structure and a novel AMM that reduces divergence loss. (2022-10-03, shares: 19)
Trading with Stochastic Price Impact: The research uses singular perturbation methods to study optimal trading in a market with stochastic price impact, showing how stochastic trading frictions affect optimal trading through numerical experiments. (2021-01-25, shares: 48)
Stochastic Gradient Descent for SDE Optimization: A novel continuous-time stochastic gradient descent method is developed for optimizing stochastic differential equation models, with potential applications in mathematical finance, including training stochastic point process models. (2022-02-14, shares: 33)
RePec
Finance
NPS Internal vs. External Management: Market Impact of Trading Strategies: The research investigates the trading strategies and market impact of the National Pension Service of Korea, identifying differences between internal and external management methods, with the latter serving as a market stabilizer and the former increasing volatility. (2023-08-30, shares: 14.0)
Exchange Rates Forecasting with Interpretable Machine Learning: The Light Gradient Boosting Machine model has been found to be the most effective at predicting 12 exchange rates due to its ability to extract short-term information and robustness on small datasets. (2023-08-30, shares: 14.0)
Volatility & Expected Returns: Past & Present: The research confirms previous findings that stock returns are influenced by aggregate-volatility risk and idiosyncratic volatility, and suggests that recent asset-pricing models fail to consistently account for this, except for the models by Stambaugh and Yuan, and Barillas and Shanken. (2023-08-30, shares: 22.0)
Sectoral & Regional Volatility: CDS Spreads & Equities: The research examines the volatility connection between the CDS and equity markets in the US, UK, EU, and Japan, finding that this connection is stronger during crisis periods and that equity is the main transmitter of volatility. (2023-08-30, shares: 17.0)
Econometrics and Analytics for Movie Success Forecasts: Social media data and hybrid strategies combining econometrics and machine learning can enhance forecast accuracy in commercial sectors like the film industry. (2022-02-06, shares: 22.0)
Housing Price Trends with Machine Learning: Machine learning can effectively predict housing price trends and variations, considering land use and transportation interactions, as evidenced in a Toronto and Hamilton Area study. (2022-07-04, shares: 21.0)
Machine Learning vs. Dictionary Methods for Disclosure Sentiment: Machine-learning methods, particularly the random-forest-regression-tree method, provide a more accurate measure of disclosure sentiment than dictionary-based methods. (2022-11-08, shares: 13.0)
Financial Data Modeling with Heterogeneous Tail Factors: The suggested Factor-HGH model for the combined distribution of financial factors and asset returns shows benefits in capturing data stylized facts and enhancing portfolio performance, particularly with highly tail heterogeneous cryptocurrencies. (2023-08-30, shares: 14.0)
Enhanced VaR Estimation: Machine learning is enhancing the accuracy and reliability of Value at Risk (VaR), a tool used in risk management for estimating potential portfolio losses. (2023-08-30, shares: 25.0)
GitHub
Finance
Advanced Trading App: Explores a web app that integrates technical, fundamental research and sentiment analysis for investment trading. (2022-11-24, shares: 94.0)
Efficient Data Manipulation: Introduces NVTabular, a tool for processing large-scale tabular data in deep learning-based recommendation systems. (2020-04-03, shares: 946.0)
Curated Resources for Hedging: Provides a detailed list of resources for understanding and implementing Deep Hedging. (2021-12-27, shares: 37.0)
Evaluate Option Trading Strategies: Discusses a Python library specifically designed for evaluating option trading strategies. (2023-07-04, shares: 7.0)
Quantitative Analysis Tools: Presents a collection of tools and examples for Quantitative Analysis, including MonteCarlo Simulations, Linear Regression, and TimeSeries Analysis. (2022-04-08, shares: 10.0)
Trending
Python in Excel: Python programming language can be integrated and utilized in Microsoft Excel. (2023-08-16, shares: 271.0)
Python Stochastic Processes Simulation and Visualization Library: Python library that allows for the simulation and visualization of random processes. (2022-09-07, shares: 59.0)
CodeLlama Model Inference Code: Details the inference code used in CodeLlama models. (2023-08-24, shares: 3286.0)
Automatic Generation of Visualizations and Infographics with Language Models: Investigates the use of large language models in automatically generating visualizations and infographics. (2023-03-09, shares: 820.0)
Graph of Thoughts Implementation for Problem Solving with Language Models: Official implementation of the Graph of Thoughts method, which uses large language models to solve complex problems. (2023-08-18, shares: 556.0)
LinkedIn
Trending
Losses of Liquidity Provision in Constant Function Markets: A recent working paper examines the predictable losses of liquidity providers in constant function markets and concentrated liquidity markets, using data from Uniswap v3. (2023-08-30, shares: 1.0)
Continuous Statistical Jump Models for Financial Regimes: A new statistical jump modeling approach in finance is introduced, allowing for estimation of probabilities in different financial regimes for better risk management. (2023-08-30, shares: 1.0)
Improved Results with Fusion Algorithm: Keyword and Vector Search Combination: Weaviate has launched a fusion algorithm that enhances search efficiency and accuracy by merging traditional keyword-based search with vector search. (2023-08-30, shares: 1.0)
GRASFI Conference: Sustainable Finance and Biodiversity Impact on Corporate Bonds: The GRASFI conference at Yale School of Management focused on sustainable finance, particularly the influence of biodiversity events on the pricing of corporate bonds. (2023-08-30, shares: 1.0)
Successful Investing through Diversification: The article highlights the significance of diversification in both passive and quant investing, advocating for the expansion of the investment universe. (2023-08-28, shares: 1.0)
Informative
Leveraging ESG Data: Standardizing Scores and Predicting Future Returns: ESG Data: Standardizing Scores and Predicting Returns: The article advises on how to effectively use ESG data from various providers, suggesting that investors should consider ESG rating dispersion in their investment strategy. (2023-08-30, shares: 1.0)
Enhancing Trading Analytics with VisualHFT Features: Trading Analytics Enhanced with VisualHFT Features: The article reveals new features in the VisualHFT project to improve trading market analytics, including real-time studies, market data integration, and easy data analysis. (2023-08-30, shares: 1.0)
Global Financial Compensation: The eFinancialCareers 2023 Financial Services Compensation Survey offers a global overview of financial services industry salaries, aiding in salary negotiation and career planning. (2023-08-26, shares: 1.0)
AI in Financial Services Webinar: The World Alliance of International Financial Centers and NVIDIA are conducting a webinar on the influence of Generative AI in finance, featuring expert discussions and a live Q&A. (2023-08-26, shares: 1.0)
Inflation and Investment Strategies: Elisabetta B.'s blog explores the economic factors influencing inflation and how this knowledge can assist investors in strategy development. (2023-08-30, shares: 2.0
Podcasts
Quantitative
MultiFactor Investing with Asim Turk: Professor Zoro and Asim Turk discuss the MultiFactor Investing Model, covering topics such as data manipulation for stocks, multiple linear regression for returns, and result visualization. (2023-08-30, shares: 1)
Decelerating Resilience: Aahan Menon from Prometheus Investment Research talks about their macro framework, liquidity, and the potential for further growth in the bond market. (2023-08-25, shares: 2)
Erosion of Software Architecture Quality in AI Code Generation: The rise of AI-powered code generation is causing a decline in software architecture quality, leading to questions about design responsibility. (2023-08-30, shares: 0)
Twitter
Quantitative
Machine Learning Predicts Alphas: Machine learning is used to accurately forecast mutual fund alphas, with fund momentum and investor sentiment as main indicators. (2023-08-25, shares: 5)
Belief Dispersion Predicts Returns: A machine learning model shows that the variation in investors' beliefs significantly forecasts stock returns. (2023-08-25, shares: 5)
Empirical Asset Pricing Notes: Lecture notes discuss empirical asset pricing, including return predictability, volatility, interest rates, and other topics. (2023-08-29, shares: 4)
Retail vs. Institutional Sentiment & Stock Returns: Micaletti uses Sentix survey sentiment measures and machine learning to predict stock returns based on investor sentiment. (2023-08-27, shares: 4)
Interpreting ML Models with Shapley Values: ManGroup's blog post discusses the use of Shapley values in interpreting machine learning models. (2023-08-24, shares: 4)
Scaling Equity Strategies with VIX: The article suggests that using VIX for volatility scaling can enhance equity strategies performance, especially post-transaction costs. (2023-08-25, shares: 3)
CNNs in Trading: Discusses the application of convolutional neural networks in trading and suggests further reading on the topic. (2023-08-28, shares: 3)
Miscellaneous
Improved interpretability of XGenTimeSeries: The article introduces XGenTimeSeries, a tool that improves understanding of data augmentation in energy time series applications. (2023-08-30, shares: 2)
Synthetic market data for trading: Jonathan Kinlay's paper examines the pros and cons of creating synthetic market data for trading strategies. (2023-08-25, shares: 2)
Treasury market sentiment predicts bond returns: The article reports on a study that identifies the Sentix Survey's Treasury market sentiment index as a key forecaster of US bond returns. (2023-08-24, shares: 2)
Factors of investor overconfidence: The paper studies the factors influencing investor overconfidence, using data from the UBS-Gallup Investor Optimism Survey. (2023-08-24, shares: 2)
Quant investing in emerging markets: The article presents David Blitz from Robeco's insights on quantitative investing in emerging markets. (2023-08-25, shares: 1)
Business Confidence Predicts Commodity Futures Returns: The article analyzes the correlation between business confidence fluctuations and commodity futures returns. (2023-08-26, shares: 1)
Earnings Acceleration Predicts Returns: The article investigates the predictability of returns using four different definitions of earnings acceleration. (2023-08-25, shares: 0)
Statistical Jump Models for Market Classification: The article introduces continuous statistical jump models as a method for identifying various market conditions. (2023-08-30, shares: 0)
Market Timing Analysis: The article offers a detailed analysis of market timing strategies. (2023-08-25, shares: 0)
Market Timing Linked to Returns: The article lacks sufficient information for a summary. (2023-08-24, shares: 0)
Reddit
Quantitative
RustQuant Contributors Wanted: The article discusses the skills required to be an effective quant in a large hedge fund. (2023-08-24, shares: 0.0)
Climate Finance Machine Learning: The piece is a request for guidance on data and metrics for a climate finance machine learning project. (2023-08-24, shares: 0.0)
Paper with Code
Trending
Understanding Academic Documents: The article highlights the storage of scientific knowledge predominantly in books, scientific journals, and PDFs. (2023-08-30, shares: 614.0)
OmniQuant: Calibrated Quantization for LLMs: The article presents the OmniQuant technique for LLMs, which delivers strong performance in various quantization settings while preserving computational efficiency. (2023-08-29, shares: 64.0)
Generating Deployable Models: The paper presents Prompt-2-Model, a technique that employs a natural language task description for training deployable models. (2023-08-25, shares: 158.0)
Large Language Model Autonomous Agents Survey: The article explores the difficulties and potential future paths in a specific field, drawing on past research. (2023-08-26, shares: 129.0)
Expanding Context Lengths: The article delves into the application of context length extrapolation methods for managing longer sequences in models. (2023-08-24, shares: 262.0)
WizardMath: Enhancing Math Reasoning with Language Models: The presented model excels in mathematical reasoning, performing exceptionally well on GSM8k and MATH benchmarks according to comprehensive tests. (2023-08-29, shares: 6294.0)
Rising
Code Llama: Foundation Models for Code: Code Llama, a series of advanced language models for coding, provides superior performance, supports extensive input contexts, and can execute programming tasks without prior training. (2023-08-26, shares: 4628.0)
Plan-and-Solve: Enhancing Zero-Shot Chain-ofT-hought Reasoning with Language Models: The quality of reasoning steps is enhanced and calculation mistakes are corrected by providing more detailed instructions in PS prompting. (2023-08-25, shares: 4019.0)
BEVBert: Map Pretraining for Navigation: The article details the creation of a local metric map to compile incomplete observations, eliminate duplicates, and model navigation dependency. (2023-08-28, shares: 94.0)