Algorithmic Trading Shift
A new framework for creating automated trading algorithms, which considers real-world complexity and uses an event-based time concept, has been introduced.
14 sharesSource ↗
Quant LetterNo. 82
January 2025, Week 3
76 items across 8 sections, as sent to readers on 15 January 2025. Paper titles open their ML-Quant page; ↗ goes to the source.
- Items
- 76
- Papers
- 48
- Sections
- 8
- Issue
- No. 82
arXiv
Quantitative-finance and ML-for-finance preprints from arXiv.
15 items
Finance4
HJM and Lifted Heston Model
A multiplicative multi-factor model is suggested to align historical correlations of futures contracts with implied volatility smiles in energy markets.
7 sharesSource ↗
VIX Options Pricing
The same multiplicative multi-factor model is reiterated, highlighting its successful application to German power and TTF gas markets.
7 sharesSource ↗
Liquidity Provisioning Strategies
The pricing of VIX options in the SABR model is examined, revealing infinite VIX futures and call prices, and suggesting a capped volatility process as a solution.
6 shares2 citations todaySource ↗
Economics4
Economics Causal Analysis
Research shows a rise in the use of causal claims in over 44,000 economic papers from 1980-2023, with complex causal narratives more likely to be published in top journals and receive more citations.
16 shares10 citations todaySource ↗
Russian Financial Database
The Russian Financial Statements Database (RFSD) is an open-source collection of annual financial statements from all active Russian firms from 2011-2023, offering data improvements and various economic applications.
11 shares5 citations todaySource ↗
Anonymous Attention Study
A study of the Economics Job Market Rumors (EJMR) online forum highlights its evolving relationship with external information sources like Twitter, and raises concerns about inclusivity and professional ethics in economics.
11 sharesSource ↗
LLMs and SCAs in Econ Experiments
A novel method has been introduced by researchers that uses Large Language Models to simulate economic behavior across different cultures, providing a new tool for economic and behavioral research.
9 shares6 citations todaySource ↗
Crypto & Blockchain4
Causal Relationships on Ethereum
Bitcoin's value increases with Tether minting events, particularly when publicized and amid positive investor sentiment, but this effect lessens after an hour.
163 shares21 citations todaySource ↗
Bidirectional Causal Relationships on Ethereum
The price of Bitcoin rises following Tether minting events, especially when made public and during positive investor sentiment, but this reaction diminishes after 60 minutes.
163 shares21 citations todaySource ↗
Dynamic Causal Relationships on Ethereum
Research shows a dynamic link between Ethereum's transaction fees and economic subsystems, significantly impacting user activity, exchange volumes, and stablecoin transactions.
29 sharesSource ↗
Ethereum Transaction Fees and Economic Activity
A study indicates a dynamic causal connection between Ethereum's transaction fees and economic subsystems, notably affecting user activity, exchange volumes, and stablecoin transactions.
29 sharesSource ↗
Historical Trending3
Congressional Partisanship
The study uses an advanced language model to confirm the rise in partisanship in U.S. congressional speech since the 1990s, also revealing changes in topic content and partisan phrases.
37 shares5 citations todaySource ↗
Economics PhD Influence
The research finds that economics PhD students guided by research-active advisors publish more, but suggests these advisors attract already successful students rather than enhancing their success.
34 shares5 citations todaySource ↗
Reinforcement Learning for Language Models
The paper introduces a new method for training large language models using reinforcement learning feedback, which makes models more beneficial and less harmful by adjusting sensitivity to reward values.
17 shares2 citations todaySource ↗
SSRN
Working papers in finance and economics from SSRN.
14 items
Quantitative5
Credit Risk Modeling
The article discusses the use of normalizing flows and invertible neural networks in credit risk modeling to enhance default time estimation and portfolio risk assessment.
28 sharesSource ↗
Volatility Forecast Combination
The authors introduce a forecast combination scheme with fluctuating weights based on financial decisions, showing better economic performance than existing methods.
12 sharesSource ↗
LeadLag Relationships in Stock Futures
The study investigates the relationships between different maturity stock index futures contracts in China, finding that price discovery is led by near-month contracts.
6 sharesSource ↗
Enhancing Inflation Forecasts
A thick modeling approach combining time and frequency-domain models enhances inflation forecasts, particularly during periods of high inflation volatility.
10 sharesSource ↗
Predicting College Closures
Predictive models of financial distress in colleges can be developed using a comprehensive dataset on college characteristics, with missing data identified as a key issue.
4 sharesSource ↗
Financial9
Advancing Towards Net-Zero Economy
Climate-related bonds issued by governments and international organizations are vital for a net-zero economy, offering investors protection against long-term climate risks and demonstrating government dedication to climate action.
8 shares3 citations todaySource ↗
Inflation's Impact on Risky Asset Prices
A study using inflation swap prices shows that inflation sensitivity changes over time, with good inflation decreasing corporate credit spreads and increasing equity values, while bad inflation can have the reverse impact.
7 sharesSource ↗
Estimating Location Value with SHAP
The study uses AI to calculate the location value of Swiss apartments, distinguishing it from the building's value and comparing across different areas.
8 sharesSource ↗
Impact of Geopolitical Risk on Mutual Funds
The research shows that active funds in the U.S. lessen the impact of geopolitical shocks by selling stocks of firms exporting to China and buying lottery stocks.
3 sharesSource ↗
Equity Risk Premium
The article suggests a new method for predicting the equity risk premium using a deep learning combination forecast, which aggregates firm-level return predictions.
15 sharesSource ↗
Limited Risk Transfer
The article defines and estimates risk transfer using data on U.S. investors' portfolio holdings, flows, and returns, and develops a model that explains the observed risk transfer.
5 sharesSource ↗
Index Rebalancing
The article reveals that index funds suffer adverse selection costs from changes in the stock market composition, and that an annual rebalancing strategy improves fund returns.
3 sharesSource ↗
Climate Betas
The article shows the sensitivity of asset returns to climate risk proxies, and suggests that adding new asset classes to a basic equity-bond portfolio improves diversification during climate stress but increases tracking error.
2 sharesSource ↗
Insider Trading Strategy for Market Outperformance
The research suggests a portfolio construction strategy using Monte Carlo simulations and insider trading transactions, showing consistent outperformance of the S&P 500 across most performance metrics.
3 sharesSource ↗
RePEc
Economics working papers from RePEc's NEP field reports.
10 items
Finance3
Insider Trading Model
The proposed insider trading model suggests that a higher correlation coefficient leads to more informative equilibrium prices, reducing trading intensity and the insider's expected payoff.
15 sharesSource ↗
Group Mean-Differences
The research proposes a compensation-factor to estimate the additional observed variables needed to recover the determinacy coefficient size after eliminating a group mean-difference, impacting the validity of the factor score predictor.
4 sharesSource ↗
Robust Haberman Linking
The comparison between robust Haberman linking and invariance alignment for factor models shows that Haberman linking performs better when item intercepts are used, with varying results depending on the loss function chosen.
2 sharesSource ↗
Statistical3
Impurity Functions in Trees
The research disproves the similarity between impurity and concave functions in decision trees, suggesting a combination of Gini Index and Entropy may be more effective.
5 sharesSource ↗
ECB Press Conference Sentiment
The research uses FinBERT to analyze sentiment in ECB president's introductory statements, finding that the sentiment about monetary policy significantly affects subsequent press conference content.
3 sharesSource ↗
State Equations in High-Dimensional Regression
The article explores the interchangeability of various state equations in high-dimensional statistics through parameter transformations.
3 sharesSource ↗
Machine Learning1
Polarized Online Discourse
The study uses social network analysis to examine the traits of 'deniers' and 'believers' in controversial topics on social media. It found that 'deniers' show more consistency, which can aid in controlling information spread and identifying false information.
4 sharesSource ↗
Historical Trending3
Foreign Portfolio Investment and Index Crash Risk
Research shows that exchange rate fluctuations and investor sentiment significantly impact country index crash risk, while net foreign portfolio investment has minimal effect.
24 sharesSource ↗
Machine Learning for Stock Market Prediction in Africa
The XGBoost machine learning algorithm is the most effective at predicting crises in the African stock market, with stock prices and exchange rates being key factors, according to a study.
22 sharesSource ↗
Finance Research Trends from Machine Learning
The paper employs machine learning models to identify trends in finance research topics from 1976 to 2015, revealing growth and shrinkage in topics and a consistent pattern in topic coverage among researchers.
14 sharesSource ↗
Machine learning
The general machine-learning papers the letter carried in 2023-25.
9 items
Recently Published6
Empirical Study of Autoregressive Pre-training
The study presents Toto, a series of video models trained on over 1 trillion visual tokens, showing strong performance in tasks like image recognition and object tracking.
131 shares22 citations todaySource ↗
Decentralized Diffusion
The paper suggests Decentralized Diffusion Models, a framework for distributing AI model training across separate clusters, reducing costs and increasing resilience to GPU failures.
80 shares14 citations todaySource ↗
The GAN is dead; long live the GAN!
The study introduces R3GAN, a simplified GAN baseline that outperforms StyleGAN2 on various datasets and competes well against other state-of-the-art GANs and diffusion models.
67 shares110 citations todaySource ↗
GenMol: Drug Discovery Generalist
Drug Discovery Generalist: The paper presents GenMol, a molecular generative model that surpasses previous models in new generation and fragment-constrained generation, offering a unified approach for drug discovery tasks.
47 shares58 citations todaySource ↗
Neuro-Symbolic AI Review 2024
Neuro-Symbolic AI has grown since 2020, focusing on learning and inference, but still lacks in areas like explainability, trustworthiness, and Meta-Cognition.
40 shares91 citations todaySource ↗
RoboPanoptes: The All-seeing Robot
The All-seeing Robot: RoboPanoptes, a robot system, learns complex manipulation skills from human demonstrations using a visuomotor policy, enabling it to perform tasks like unboxing in narrow spaces and sweeping oversized objects.
16 shares22 citations todaySource ↗
Historical Trending3
Adjoint Matching
The study presents Adjoint Matching, a new algorithm that enhances dynamical generative models by refining reward fine-tuning, leading to improved consistency, realism, and adaptability to unseen human preference reward models.
222 shares227 citations todaySource ↗
Grokking Softmax Collapse
The study investigates 'grokking' in deep learning, introduces Softmax Collapse and naïve loss minimization concepts, and suggests a new activation function and training algorithm for grokking without regularization.
50 shares41 citations todaySource ↗
Unity in VAEs
A new mixture-of-experts prior for Variational Autoencoders for multimodal data has been proposed, replacing hard constraints with a soft one, leading to better latent representation and improved imputation of missing data modalities.
41 shares20 citations todaySource ↗
GitHub
Repositories the letter featured.
9 items
Finance5
Skimpy: Data Frame Summary Tool
Data Frame Summary Tool: Skimpy is a tool that provides summary statistics for data frame variables in the console.
422 shares
Dendrite SDK: AI Agent Builder
AI Agent Builder: The article explores tools for developing web AI agents capable of interacting with and extracting data from websites.
219 shares
StockStats: DataFrame Wrapper
DataFrame Wrapper: The article introduces a StockDataFrame wrapper, based on pandas.DataFrame, that supports stock statistics indicators.
1,324 shares
Monaco: Model Uncertainty Library
Model Uncertainty Library: The piece presents a Monte Carlo library for quantifying uncertainties and sensitivities in computer models.
108 shares
Cosmograph: Visual Analytics Binder
Visual Analytics Binder: The article discusses a binder for cosmograph visual analytics for large graphs.
71 shares
Trending4
Arch Agent Gateway
Arch is a smart gateway developed by Envoy proxy's core contributors, designed for agents to securely and seamlessly integrate user prompts with external APIs.
1,240 shares
OpenAI: Structured Output Apps
Structured Output Apps: Developers are provided with sample applications to help them understand Structured Outputs.
369 shares
Resume Matcher: Resume Improvement
Resume Improvement: Resume Matcher is an open-source tool that uses AI to improve resumes by comparing and ranking them against job descriptions.
7,577 shares
LightningAI: Multi-GPU AI Pretraining
Multi-GPU AI Pretraining: AI models of any size can be pretrained and finetuned on multiple GPUs TPUs without any code modifications.
28,792 shares
News
Industry news: funds, hiring, markets and regulation.
6 items
Quantitative3
Data Quality Challenge for Quants
A Bloomberg Research Survey reveals that the primary challenge for Quants Research Analysts and Data Scientists is the timeliness and quality of data coverage.
6 shares
Prop Trading Firms Avoid PhD Quants
The article debates whether a PhD is necessary for a job in quantitative finance.
4 shares
Point72 Quant Platform Head Leaves
The piece explores the consequences of leaving a successful hedge fund for a less prosperous one.
3 shares
Miscellaneous3
Quant Gains by Renaissance Tech and Two Sigma
In 2024, quant hedge funds and multistrategy firms experienced significant double-digit growth, as reported by Business Insider.
3 shares
Citadel Securities Hiring from Two Sigma and Millennium
The electronic trading firm is actively hiring in 2025.
2 shares
Two Sigma Founders Arbitration
Two Sigma Investments' co-founders, John Overdeck and David Siegel, are in arbitration over a disagreement about the hedge fund's future.
2 shares
Podcasts
Episodes on markets, quant methods and economics.
6 items
Quantitative3
Machine Learning in Investing
Phil Wool of Reliant Global Advisors examines the influence of machine learning on strategic investing, US equities dynamics, and the tech sector's potential in emerging markets.
18 shares
Backtesting Insights
Matt Amberson from ORATS discusses the importance of backtesting in refining options trading strategies and forecasting future performance.
17 shares
Implied Volatility and Gamma
Mat Cashman from the Options Clearing House explains the inverse correlation between Gamma and Implied Volatility in options trading.
13 shares
Related3
Global Data Insights
In a 2025 podcast, Bruce Kasman and Joseph Lupton analyze the stability of growth, inflation, and central bank policies amid US policy uncertainties.
7 shares
Elite Network Dynamics
Vuk Vukovic explores the impact of political connections on income inequality and introduces his unique market research and trading approach using crowd wisdom and social media analysis.
7 shares
Political Investment Strategies
Hal Lambert discusses how political beliefs shape investment strategies and the potential effects of deregulation on various sectors, especially under a Trump administration.
7 shares
X / Twitter
Posts from quant researchers on X.
7 items
Quantitative3
Portfolio Protection Study
The article proposes a varied portfolio protection strategy that includes SPX rolling puts, Trend, Long Rates Vol, and Quality to control equity drawdowns.
4 shares
Investing Research Recap
The recent investment research discusses Bitcoin trading strategies, market betas estimation, LLMs biases, macro risks, and portfolio protection strategies.
2 shares
Short-Term Reversion Indicators Comparison
The most effective short-term mean-reversion indicators for global equities are those that include current price measures in the intraday or multi-day trading range.
1 shares
Miscellaneous4
ChatGPTs Cognitive Decline
The article debates the possible cognitive decline from using ChatGPTs, likening it to how calculators affected mental math abilities.
0 shares
Entropy Visual History
The article offers a detailed, non-mathematical visual explanation and history of the concept of entropy.
0 shares
Stock Signals from Insider Trades
The paper studies a vast global dataset of insider trades, creating a collective signal for each stock and discovering significant alphas in roughly 23% of countries.
0 shares
Industry Momentum Strategies Enhancement
The article investigates the improvement of industry momentum strategies with news sentiment and dispersion, noting substantial performance improvements particularly at more detailed industry levels.
0 shares