Quant Letter: January 2025, Week-1
Weekly (75th Edition)
Abstract
In our SSRN’s Recently Published categories, top Quantitative studies include AI Investment Decisions—revealing how AI might sharpen portfolio outcomes yet exacerbate expertise imbalances—and Decoding China's Industrial Policies, employing NLP on 3 million documents. The Factor Investing with Delays paper highlights the ramifications of transaction-cost modeling in illiquid bond markets, while an Inflation Volatility Model surpasses classical benchmarks in capturing price-level asymmetry. Topping the SSRN Financial list is Optimal Execution Strategies Incorporating Internal Liquidity, which merges limit and market orders to reduce market impact, and Congress and Market Declines, connecting legislative sessions to equity dips. Notable “Recently Updated” SSRN entries analyze ML Takeover Verification, RL Option Hedging, and Learning Strategies in Broker-Mediated Markets—all reinforcing machine learning’s growing importance in finance.
Recent arXiv papers offer the most significant developments this week, with those at the top of each section holding particular importance. In Large Language Models for Sentiment Trading (46 shares), researchers showcase how the OPT model outperforms traditional sentiment-analysis techniques in U.S. equity prediction. Next, a Robust Portfolio Selection Model ties risk-aversion directly to portfolio diversification, while BSDEs with Singular Terminal Values tackles optimal liquidation in dark pool markets.
Additional highlights include Market Interactions, which gives brokers an information edge over price-only traders, and Volatility Modeling, introducing a path-lifting approach for rough volatility. Rounding out the top-tier arXiv Finance lineup are Mid-Price Forecasting, Ambiguous Quantiles, Travelling Waves in Finance, Multi-Agent LLM Financial Trading, and Hidformer in Forecasting. Meanwhile, Crypto & Blockchain work at the top underlines Spillover Effects in Crypto Market, an NLP-based approach spotlighting data-quality gaps.
Finally, RePec contributions—like Commodity Futures Selection, which finds simpler covariance methods can outperform complex estimators—offer strategic investment insights. GitHub repositories, such as Reinforcement Learning Investment, highlight open-source approaches for alpha creation, and DeepSeekV3 Report proposes a massive 671B-parameter language model.
SSRN
Recently Published
Quantitative
AI Investment Decisions: AI enhances investment performance, but it benefits those with greater financial expertise more, potentially increasing existing performance disparities. (2024-12-29, shares: 3.0)
Decoding China's Industrial Policies: Large Language Models are utilized to interpret China's industrial policies from 2000 to 2022, extracting structured data from 3 million government documents. (2024-12-31, shares: 31.0)
Factor Investing with Delays: The study presents a new method for calculating transaction costs in the infrequently traded corporate bond market, emphasizing the significant impact of delay costs in illiquid securities. (2024-12-28, shares: 2.0)
Inflation Volatility Model: A new semiparametric volatility model is introduced for estimating inflation volatility, outperforming standard models and offering a new measure to explain the fluctuating asymmetric inflation level-volatility relationship. (2024-12-29, shares: 2.0)
Volatility Prediction in Chinese Futures: A new deep learning method is introduced for predicting Chinese futures market movements, demonstrating superior predictability compared to existing benchmarks. (2024-12-30, shares: 4.0)
Financial
Optimal Execution Strategies Incorporating Internal Liquidity: The paper introduces a new algorithmic execution model that combines interbank limit and market orders with internal liquidity, aiming to reduce market impact and enhance execution efficiency. (2024-12-28, shares: 3.0)
Congress and Market Declines: Stock market performance tends to drop when Congress is in session, likely due to high disapproval ratings and media-induced behavioral biases. (2024-12-27, shares: 2.0)
Limited Risk Transfer Between Investors: The paper reveals that risk transfer among U.S. investors is minimal, challenging macrofinance models that forecast a larger risk transfer due to varying equity premiums. (2024-12-28, shares: 5.0)
Monetary Transmission in Equity Markets: The study suggests that contractionary monetary policy can cause equity outflows from mutual funds, which are then absorbed by banks charging a premium for market making. (2025-01-01, shares: 3.0)
Altered Monthly Asset Returns: The termination of the CRSP tape in 2025 will alter 9.62% of monthly returns, but these changes will not significantly impact time-series average premia or their significance. (2024-12-28, shares: 3.0)
Recently Updated
Quantitative
ML Takeover Verification: The study uses machine learning models to predict the accuracy of corporate takeover rumors, emphasizing the effectiveness of TabNet and the need for data imbalance and dimensionality reduction. (2024-12-21, shares: 8.0)
RL Option Hedging: The study introduces a Reinforcement Learning-based algorithm for hedging European call option payoffs, proving its effectiveness against the standard delta hedging strategy. (2024-12-17, shares: 9.0)
Learning Strategies in Broker-Mediated Markets: The article reveals that brokers in a broker-mediated market have a strategic advantage and can profit from information leakage in client trading. (2024-12-24, shares: 15.0)
Mutual Fund Alpha and FOMC Meetings: The article studies the ability of mutual fund managers to generate positive alpha around FOMC meetings, and its impact on investor flows and fund performance. (2024-12-21, shares: 6.0)
Financial
Multivariate Volatility: The article introduces a multivariate version of the Rough Fractional Stochastic Volatility model for analyzing logvolatilities, providing an estimator and confirming its theory through simulation. (2024-12-20, shares: 12.0)
ETFs for Discipline: The study shows that active exchange-traded funds (AETFs) help investors remove underperforming managers, leading to improved sector efficiency and price informativeness. (2024-11-21, shares: 9.0)
Foreign EPUF and US Equity: The article reveals that foreign economic policy uncertainty significantly predicts excess U.S. stock returns, primarily affecting equity prices through cash flow news. (2024-12-19, shares: 10.0)
DeepFake and Trading Signals: Advanced AI-powered deepfake technology threatens financial markets by manipulating market sentiment, disrupting trading systems, and exploiting retail investors. (2024-08-08, shares: 2.0)
Twin Stars: Rates and Currency Risk: The currency risk premium links the neutral interest rates of two countries, with global interest rates showing strong correlations across foreign countries. (2024-12-20, shares: 8.0)
ArXiv
Finance
Large Language Models for Sentiment Trading: The OPT model, a large language model, has proven superior in predicting stock market returns using sentiment analysis of U.S. financial news, outdoing traditional methods like the Loughran-McDonald dictionary model. (2024-12-26, shares: 46)
Robust Portfolio Selection Model: A new portfolio selection model is proposed to minimize estimation errors and over-diversification, with an algorithm developed to establish a direct link between the risk-aversion coefficient and robustness level. (2024-12-27, shares: 6)
BSDEs with Singular Terminal Values: A novel model is presented to address a stochastic control issue in optimal liquidation with dark pools, using a system of backward stochastic differential equations with jumps and singular terminal values. (2024-12-26, shares: 5)
Market Interactions: The research shows brokers in a mediated market have a strategic edge over traders who only use prices for information, and that brokers can gain economic value from leaked information. (2024-12-30, shares: 5)
Volatility Modeling: The paper presents a method for jointly lifting a Brownian motion and a low-regularity adapted stochastic rough path, useful for modeling rough volatility. (2024-12-30, shares: 5)
Mid-Price Forecasting: The article introduces the Adaptive Learning Policy Engine, a new approach to mid-price forecasting using NASDAQ data, which outperforms other machine learning models. (2024-12-26, shares: 4)
Ambiguous Quantiles: The study introduces Choquet Expected Shortfall, a new class of risk measures, and provides optimization algorithms and examples using financial data. (2024-12-27, shares: 2)
Travelling Waves in Finance: The paper examines the use of a second-order differential operator in the Black-Scholes framework, leading to a variant of the Harry Dym equation for potential use in studying financial-market volatility. (2024-12-26, shares: 6)
Multi-Agent LLM Financial Trading: TradingAgents, a new stock trading framework, employs large language models to simulate real-world trading dynamics, enhancing trading performance. (2024-12-28, shares: 5)
Hidformer in Forecasting: The research explores the application of the Hidformer model, a Transformer-based neural network, in stock price forecasting, combining technical analysis and advanced machine learning for improved accuracy. (2024-12-27, shares: 3)
Crypto & Blockchain
Spillover Effects in Crypto Market: The paper discusses the application of text mining and natural language processing in finance, emphasizing the need to improve data quality and model understanding for better financial predictions. (2024-12-29, shares: 11)
Oracle Limitations in Blockchain: The study reveals a lack of research on the challenges oracles present in incorporating blockchain technology into accounting systems, despite numerous articles on the topic. (2024-12-29, shares: 8)
Text Mining in Finance System: The research highlights a gap in studies addressing the constraints of oracles in the implementation of blockchain technology in accounting systems, despite abundant literature on the subject. (2024-12-29, shares: 8)
Crypto Risk Propagation: The article proposes a strategy to manage cryptocurrency risks during crises like the new crown outbreak, after analyzing how risks spread among cryptocurrencies during such extreme events. (2024-12-28, shares: 2)
ArXiv ML
Recently Published
nm FD-SOI CNN Accelerator: The paper introduces IMAGINE, a compute-in-memory SRAM for processing convolutional neural networks, which offers high energy efficiency and competitive accuracies on MNIST and CIFAR-10. (2024-12-27, shares: 8)
Tensor Network EDAs in Evo Optimization: The study explores the use of tensor networks in evolutionary optimization algorithms, concluding that better generative models don't always improve optimization performance and suggests adding a mutation operator for better results. (2024-12-27, shares: 8)
LASER: Locally Adaptive Nonparametric Regression: The paper presents LASER, a new nonparametric regression method that adapts to the local Hölder exponent of the regression function, outperforming other locally adaptive methods in various experiments. (2024-12-27, shares: 7)
Training SWE Agents with SWE-Gym: The authors introduce SWE-Gym, the first training environment for software engineering agents, featuring real-world Python tasks and showing significant improvements in task resolution rates. (2024-12-30, shares: 6)
AI Impact on Human Skills Demand: The research studies the effect of AI on skill demand and compensation in the U.S., finding increased demand for AI-complementary skills and decreased demand for substitute skills, indicating AI's transformative impact on workforce skills. (2024-12-27, shares: 5)
Historical Trending
Mitigating Optimistic Bias: A novel bootstrapping method is suggested to reduce bias in the empirical entropic risk estimator, a tool used in high-stakes decision making, and is applied to insurance contract design. (2024-09-30, shares: 18)
Decentralized Intelligence in GameFi: A proposed GameFi ecosystem integrates advanced AI agents into gaming platforms, improving player engagement and economic interaction within gaming ecosystems. (2024-12-24, shares: 13)
Structure Learning in Gaussian Graphical Models: A new algorithm for Gaussian graphical model selection is presented, offering theoretical guarantees on computational and statistical complexity when data is sampled according to the Glauber dynamics. (2024-12-24, shares: 12)
PDE Flow Map Learning from Noisy Data: A new computational technique has been introduced for modeling the evolution of dynamical systems, specifically targeting the complex problem of modeling partially-observed partial differential equations on high-dimensional non-uniform grids. (2024-07-15, shares: 11)
RePec
Adaptive Online Portfolio Selection: The article introduces a new online portfolio selection strategy that considers transaction costs and uses an adaptive scheme for sequential parameter decision, yielding higher cumulative returns and competitive Sharpe ratios than existing strategies. (2025-01-01, shares: 10.0)
Commodity Futures Selection: The article finds that traditional sample covariance matrix performs better in portfolio selection than both naive allocation and advanced covariance estimators, challenging previous equity-focused studies. (2025-01-01, shares: 21.0)
Stochastic Non-Dominance Measures: The research introduces measures of stochastic non-dominance to analyze scenarios where stochastic dominance rules are not applicable, using the Wasserstein distance as the measure. (2025-01-01, shares: 8.0)
Selecting Factors in Time Series: The paper suggests a new eigenvalue ratio criterion to determine the number of factors in static approximate factor models, validating its effectiveness through a Monte Carlo study. (2025-01-01, shares: 6.0)
GitHub
Finance
Reinforcement Learning Investment: The piece explores a deep reinforcement learning framework for developing alpha factors for quantitative investment using GFlowNet and PythonPyTorch. (2024-05-08, shares: 27.0)
DeepSeekV3 Report: DeepSeekV3 is a robust language model with 671B parameters, 37B of which are used for each token. (2024-12-31, shares: 10160.0) Editor note: seems like a true cheap openai alternative.
AutoJobsApplier Agent: The article talks about AutoJobsApplierAIAgent, an AI tool developed to streamline job searching by automating the application process. (2024-08-04, shares: 23896.0)
Charting Library Integrations: The article demonstrates how the Charting Library can be combined with other libraries, frameworks, and data transports. (2018-02-14, shares: 1476.0)
AI Agent Toolkit: The article presents a detailed toolkit for managing an AI agent service, built using LangGraph FastAPI and Streamlit. (2024-08-04, shares: 641.0)
News
Multistrategy Hedge Fund Boom Persists: Despite Citadel founder Ken Griffin's assertion, a Bloomberg report indicates that the multistrategy hedge fund boom is still ongoing. (2024-12-31, shares: 2)
Banks Expand Tech and Quant Teams: New groups are being established as the new year begins. (2024-12-31, shares: 1)
Odey Sues Financial Times for Libel: Hedge fund manager Crispin Odey is filing a libel lawsuit against the Financial Times, demanding at least £79m in compensation. (2024-12-31, shares: 1)
Goldman Sachs MD Joins Jane Street: Another Managing Director leaves the company just before the bonus season starts. (2024-12-31, shares: 0)
Twitter
Research Papers Compilation: The article compiles popular research papers from 2024, discussing topics like alternative data commodities, equities, machine learning, and trading. (2024-12-31, shares: 8)
Momentum Strategies Review by Tobias Wiest: The article discusses Tobias Wiest's comprehensive review on different momentum strategies such as cross-sectional timeseries, residual volatility-scaled, industry factor momentum, and their theoretical justifications. (2024-12-26, shares: 0)