SSRN
Recently Published
Quantitative
Deep Reinforcement Learning for US Equities Trading: The study shows that Deep Reinforcement Learning can effectively interpret synthetic alpha signals in financial trading, outperforming the market benchmark. (2023-11-27, shares: 3.0)
Machine Learning for Portfolio Performance: The study introduces a method to determine the impact of individual factors on portfolio performance, providing insights into the economic value of return predictability in machine learning models. (2023-11-29, shares: 2.0)
Machine Learning for Path-Dependent Contracts: The study introduces a new method for pricing financial products with early-termination features using machine learning algorithms and Chebyshev interpolation techniques. (2023-11-28, shares: 5.0)
Historical Calibration of SVJD Models with Deep Learning: The paper suggests using deep neural networks to calibrate parameters of Stochastic Volatility Jump Diffusion models, proving to be more accurate, robust, and faster than other methods. (2023-12-01, shares: 2.0)
Financial
Competition between ETFs and Mutual Funds: The research indicates that less transparent active ETFs do not affect mutual fund investor flows, instead, the reputation of the cloned mutual funds helps the new ETFs attract more flows. (2023-12-05, shares: 2.0)
Disagreement Proxies and Price Impact: The study presents a new framework to better understand investor disagreement, introducing a more accurate measure that can predict returns. (2023-11-28, shares: 39.0)
Pricing VIX Derivatives in Stochastic Volatility Model: The paper introduces a new stock price model based on continuous-state branching processes, providing a formula for VIX put option price. (2023-11-30, shares: 4.0)
Causal Reductionism in Finance Limits: The research suggests that traditional methods of studying finance and econometrics may be flawed, proposing a new approach of considering multiple causal factors. (2023-11-28, shares: 31.0)
Recently Updated
Quantitative
RL and Deep Stochastic Optimal Control for Quadratic Hedging: The study compares Reinforcement Learning and Deep Trajectory-based Stochastic Optimal Control for hedging a European call option, finding both methods perform similarly under various market conditions. (2023-11-20, shares: 3.0)
ML and IRB Capital Requirements: Advantages, Risks, and Recommendations: The article explores the potential of machine learning in improving bank capital requirements and enhancing financial inclusion through better credit risk measurement. (2023-06-25, shares: 2.0)
Machine Learning Framework for Portfolio Choice: The paper presents a computational framework for solving dynamic portfolio choice problems with multiple risky assets and transaction costs, suggesting that having more assets can mitigate some illiquidity caused by transaction costs. (2023-08-18, shares: 2.0)
Volatility and Mispricing with Sentiment and Institutional Investors: The research suggests that high investor sentiment and increased institutionalization can decrease excess volatility and mispricing in stock returns. (2022-12-11, shares: 2.0)
Sharpening Sharpe Analysis with Machine Learning: The research shows that over 95% of mutual funds have multidimensional investment styles, and those that change their styles often outperform their new style benchmarks. (2023-10-27, shares: 2.0)
ArXiv
Finance
Stock return distribution: A new model suggests that financial markets often underreact to small events and overreact to major ones, with a stronger reaction to positive events. (2023-12-05, shares: 5)
Monotonic risk measures: Monotonic mean-deviation measures have been characterized from a general model, providing new examples of consistent risk measures and establishing the consistency and normality of the natural estimators of the measures. (2023-12-02, shares: 4)
Rough Volatility: Range Volatility Estimators: The study further analyzes volatility dynamics using range-based proxies, confirming that log-volatility behaves like fractional Brownian motion and the rough fractional stochastic volatility model predicts better. (2023-12-03, shares: 7)
FinMem: LLM Trading Agent: The article presents FinMem, a new Large Language Model-based system designed to improve financial decision-making by retaining crucial information beyond human capabilities. (2023-11-23, shares: 27)
ESG Raw Scores vs Aggregated Scores: The paper compares the predictive power of raw and aggregated Environmental, Social, and Governance (ESG) scores on company stock returns and volatility, with raw ESG data proving most predictive. (2023-11-30, shares: 5)
Valuing Post-Revenue Biopharmaceutical Assets: The research introduces a new model for predicting future sales of post-revenue biopharmaceutical assets, aiding more strategic investment decisions in the biotech and pharmaceutical sectors. (2023-12-04, shares: 5)
Crypto & Blockchain
DeFi: Protocols, Risks, Governance: The article discusses the benefits of decentralized finance (DeFi) over traditional finance, the function of smart contracts, and the associated risks, highlighting the need for more research on scalability and auditing. (2023-12-02, shares: 7)
DeFi Market Misconduct Analysis: The paper investigates the rise of blockchain and DeFi, potential market misconduct, and the challenges of creating a DeFi regulatory framework, suggesting possible regulation strategies. (2023-11-29, shares: 7)
Just-in-Time Liquidity Paradox in Decentralized Exchanges: The research analyzes the paradox of just-in-time (JIT) liquidity provision in decentralized exchanges, which can reduce liquidity, and suggests a two-tiered fee structure to counteract this. (2023-11-30, shares: 6)
Uniswap Daily Transaction Indices: The study explores the effect of Layer 2 solutions on DeFi by analyzing millions of transactions from Uniswap, offering insights into adoption, scalability, and decentralization in the DeFi sector. (2023-12-05, shares: 5)
Cryptocurrency Tail Risk and Systemic Risk Estimation: The paper introduces an expectile-based approach to assess the tail risk of cryptocurrencies, presenting the Marginal Expected Shortfall as a tool to measure the impact of a single cryptocurrency on the market's systemic risk. (2023-11-28, shares: 5)
Historical Trending
Physics-Informed Convolutional Transformer for Volatility Prediction: The paper presents a new architecture using physics-informed neural networks and convolutional transformers for better predicting financial market volatility. (2022-09-22, shares: 23)
Theory and Stock Return Predictions: The research indicates that the predictability of cross-sectional return predictors decreases by half in post-sample scenarios, implying that theoretical models don't improve predictions and peer-review often misinterprets mispricing as risk. (2022-12-20, shares: 48)
Optimal Stopping with Neural Networks: The article highlights the benefits of using randomized neural networks to approximate solutions for optimal stopping problems, proving they are more efficient and faster than other machine learning methods. (2021-04-28, shares: 38)
Quantitative
RL for Limit Order Book Trading: The article explores the application of reinforcement learning in model-based limit order book trading. (2023-12-01, shares: 3)
Extracting SEC Company Filings with Python: The article reviews the SEC Company Filings, a Python library that extracts financial data from over 120 million data points. (2023-12-02, shares: 3)
Simple Factor Model for Forex: The article introduces a straightforward long-term factor model for foreign exchange. (2023-12-01, shares: 1)
Impact of Generative AI on Finance Markets and Services: The article examines the significant influence of Generative AI on financial markets and services, highlighting the importance of regulatory dynamics in its implementation. (2023-11-27, shares: 1)
Paper on empirical Bayes and out-of-sample returns: The article introduces a new paper that applies empirical Bayes to discover out-of-sample returns from 70,000 long-short trading strategies. (2023-11-28, shares: 1)
Analyst underreaction decline and momentum strategy profitability: The article explores a paper suggesting that the profitability of a 12-month momentum strategy has decreased due to less analyst underreaction to news. (2023-11-29, shares: 1)
Morgan Stanley research on GenAI impact on work and skills automation challenges: The article discusses a Morgan Stanley research on the effects of GenAI on work, advocating for side-hustles and discussing skills that are difficult to automate with AI. (2023-12-01, shares: 0)
Miscellaneous
Obfuscation: More Sinister?: The article explores the idea of effective obfuscation, questioning if it's a real accelerationism or a disguise for something darker. (2023-12-04, shares: 0)
GenAI in Financial Services: The article shares a report by Oliver Wyman about the impact of GenAI in the financial services sector. (2023-12-03, shares: 0)
Analyst Disagreement and Future Returns: The article reviews studies showing a negative link between analyst disagreement and future returns, emphasizing the importance of proxy choice in empirical results. (2023-12-03, shares: 0)
Generative AI in Virtual Meetings: The article cannot be summarized due to lack of information. (2023-12-03, shares: 0)
MIT UBS' Value in Generative AI in Financial Services: MIT and UBS report explores the application and value of Generative AI in financial services, citing examples like RCBC Kasisto, Cowbell Insurance, and Goldman Westpac. (2023-12-01, shares: 0)
RePec
Finance
Constrained index tracking optimization models: The research investigates two methods of including liquidity constraints in portfolio optimization, finding that these constraints increase liquidity and tracking errors. (2023-12-06, shares: 34.0)
Bond Selection: The chapter discusses the challenges of bond selection and the use of traditional optimization techniques, highlighting the need for thorough analysis in portfolio construction. (2023-12-06, shares: 20.0)
Dynamic limit order placement's impact on stock market quality: The study investigates the impact of two system upgrades by the Australian Securities Exchange on dynamic limit order placement activities and market quality, revealing both positive and negative effects. (2023-12-06, shares: 19.0)
Forecasting Parameters in SABR Model: Two methods for predicting parameters in the SABR model, the vector autoregressive moving-average model and epsilon-support vector regression, both provide accurate fits, with the SABR model yielding superior pricing results. (2022-02-26, shares: 15.0)
Statistical
Real Estate Appraisals: ML vs Traditional Methods: Research indicates that machine learning, particularly XGBoost, offers the most precise predictions in Automated Valuation Models for residential properties, suggesting a need for regulators to consider various methods. (2023-12-06, shares: 31.0)
Bitcoin Futures Forecasting with ML: Machine learning algorithms have proven to be more effective than traditional models in predicting Bitcoin futures prices, maintaining an average classification accuracy consistently over 50%. (2023-12-06, shares: 18.0)
Survival Models for Startup Failures: The research finds that advanced machine learning models like MTLR and Random Forest are more accurate in predicting startup failures than standard models. (2023-12-06, shares: 16.0)
GitHub
Finance
Statistical ML Discovery: The article explores a machine learning package designed for accurate scientific discovery through statistical analysis. (2023-08-24, shares: 113.0)
Advanced Sentiment Trading App: The piece presents a new web app for trading and investment research, featuring real-time sentiment analysis. (2022-11-24, shares: 146.0)
Scalable Realtime Datastore: The piece examines a scalable datastore specifically created for metrics events and real-time analytics. (2013-09-26, shares: 26787.0)
Koopa Learning for Time Dynamics: The article announces the launch of a code for learning nonstationary time series dynamics using Koopman Predictors, set for NeurIPS 2023. (2023-08-22, shares: 83.0)
UnbiasedGBM: Repository for Gradient Boosting Decision Tree: The article discusses a repository for Unbiased Gradient Boosting Decision Tree that offers unbiased feature importance. (2023-05-14, shares: 19.0)
Trending
Large Language Models for Quant Finance: The Citigroup Centre Auditorium is holding an event called Frontiers in Quantitative Finance: Large Language Models for Quantitative Finance. (2023-12-05, shares: 1.0)
Hierarchical PCA: Statistical Approach in Factor Investing: Hierarchical Principal Component Analysis (HPCA) is presented as a new statistical method in factor investing, providing dynamic market adaptability and superior performance than traditional methods. (2023-12-04, shares: 1.0)
Measuring Information Flows in Option Markets: A research article in the Journal of Derivatives introduces a new method to measure information flows in option markets. (2023-12-05, shares: 1.0)
Knowledge Graphs Enhance GenAI and LLMs Accuracy: A benchmark study shows that knowledge graphs enhance the accuracy of data Q&A in GenAI, emphasizing their role in democratizing data for businesses. (2023-12-05, shares: 1.0)
Winning Team's Project Addresses Feature Engineering and Modeling: A project team successfully tackled feature engineering and modeling, but faced challenges optimizing machine learning models due to limited computing power. (2023-12-05, shares: 1.0)
AI Replicability and Finance Implications: The ADIA Lab Market Prediction Competition Awards Ceremony included a discussion on the impact of AI and p-hacking in quantitative finance. (2023-12-05, shares: 1.0)
Podcasts
Quantitative
Market States and Recession Prediction: The podcast explores the influence of AI in finance, potential recession indicators, and the effect of market volatility, featuring insights from industry expert Michael Khouw. (2023-12-05, shares: 14)
The Quant Finance LIE: The podcast debunks the idea of a full stack quant in finance, suggesting individuals to focus on one primary area instead. (2023-11-29, shares: 13)
Embracing Reality: Debunking AGI Hype: Filip Piekniewski, an AI expert, debunks hype about artificial general intelligence and the singularity, focusing on real AI advancements. (2023-12-04, shares: 8)
24 Interest Rate Derivatives Forecast: Srini Ramaswamy and Ipek Ozil predict the state of interest rate derivatives markets in 2024 in a podcast recorded in December 2023. (2023-12-05, shares: 6)
Videos
Quantitative
Language to SQL Generator with LLM: Rami Krispin explains how LLM models can be used to convert language into code, specifically developing a language to SQL translator via the OpenAI API. (2023-12-05, shares: 0.0)
The Biggest LIE in Quant Finance: Krispin delves into the use of LLM models for translating language into code, focusing on the creation of a language to SQL translator through the OpenAI API. (2023-12-06, shares: 9.0)
Yield Farming: Costs, Returns, and Risks: The article debates the concept of a 'full stack quant' in quantitative finance, arguing that while such professionals exist, they typically specialize in a particular area rather than mastering all aspects. (2023-11-29, shares: 68.0)
Covariance Matrix and Shrinkage: The article explores the problems of unstable covariance matrix in contemporary statistics and suggests a practical solution through statistical shrinkage. (2023-11-29, shares: 8)
Quantitative
Microstructure Book Recs: The author recounts their recent experience with microstructure signals. (2023-11-24, shares: 54.0)
Unveiling OMMs: The article debunks the notion of secret strategies in trading, emphasizing the importance of swift decision-making. (2023-12-02, shares: 114.0)
Quant Research: Manipulating LOB: The paper explores a model for testing potential manipulation by trading algorithms and discusses the role of machine learning in asset management. (2023-11-29, shares: 64.0)
Transitioning from Prop Shop to Academia: The author considers transitioning from a proprietary trading firm to a school research lab, evaluating the trade-off between autonomy, collaboration, and financial gain. (2023-11-29, shares: 63.0)
Analyzing the Absence of Algo Traders in Forex: The article explores the underutilization of forex trading by algorithmic traders, despite its benefits like liquidity, affordability, and abundant price data. (2023-11-25, shares: 50.0)
Rising
PNL & Sharpe for Fund Hiring: The article explores what daily profit expectations should be for a trading strategy used by funds like Milleniumcitadeletc. (2023-11-26, shares: 47.0)
Work Culture: Hedge Funds in NY vs London: The article provides insights on what to consider when choosing a permanent workplace. (2023-11-23, shares: 95.0)
Books: Good Plot Only: The article examines the daily tasks of quant High-Frequency Trading (HFT) traders, considering that models are developed by researchers and implemented by developers. (2023-11-28, shares: 154.0)
HFT Traders' Daily Activities: The article offers guidance to a statistics student debating whether to take advanced courses in Bayesian statistics, Machine learning, or Partial Differential Equations (PDEs). (2023-11-26, shares: 135.0)
ArXiv ML
Recently Published
Training Data Extraction from Language Models: The study shows that large amounts of training data can be extracted from different machine learning models, highlighting that current techniques do not prevent data memorization. (2023-11-28, shares: 115)
Benefits of Overparameterization in ML: The paper supports the theory that larger model size, more data, and more computation enhance performance in random feature regression, similar to shallow networks with only the last layer trained. (2023-11-24, shares: 66)
Enhanced Sample Quality with Self-Attention Guidance: Denoising diffusion models are becoming increasingly popular due to their high-quality and diverse generation capabilities. (2023-11-23, shares: 16676.0)
Historical Trending
Universalizing Weak Supervision for Any Label Type: The article introduces a universal technique for weak supervision frameworks that can be applied to any label type, demonstrating improvements in various settings including learning-to-rank and regression problems. (2021-12-07, shares: 71)
Edge Directionality in Heterophilic Graphs: The study presents Directed Graph Neural Network (Dir-GNN), a new deep learning framework for directed graphs that surpasses traditional models in heterophilic benchmarks. (2023-05-17, shares: 186)
No comments:
Post a Comment