Quant Letter: April 2024, Week-4
Quant Letter: April 2024, Week-4
Weekly (43rd Edition)
SSRN
Recently Published
Quantitative
Trading Volume Alpha: The article emphasizes the importance of predicting trading volume in portfolio optimization, noting that the benefits can be as significant as those from return prediction. (2024-04-21, shares: 9.0)
AI Analytics Post-IPO: Companies that adopt AI analytics after their IPOs experience a smaller drop in innovation quality, as AI analytics helps alleviate the pressure to meet short-term financial targets and disclosure obligations. (2024-04-22, shares: 2.0)
Automated Market Makers Liquidity Pool Design: The study suggests a model for optimal liquidity in automated market makers, showing that exchange rate volatility increases the optimal transaction fee. (2024-04-20, shares: 5.0)
Local Volatility Calibration: A study introduces a new calibration criterion for local volatility models that minimizes the gap between theoretical and market implied volatilities, balancing calibration error reduction and overfitting prevention. (2024-04-20, shares: 3.0)
Byzantine-Robust Federated Learning with Clustering Model Updates: The paper presents FedCmp, a method to protect federated learning from Byzantine attacks by detecting malicious updates using a multiround voting system. (2024-04-20, shares: 3.0)
Financial
SEC Regulation of Investments: The SEC is adjusting its regulatory authority to accommodate the rise of crypto assets, in line with its mission to safeguard investors and facilitate capital formation. (2024-04-22, shares: 3.0)
Mutual Funds and Social Responsibility: Chinese mutual funds investing in socially responsible investments (SRI) have shown improved performance, indicating that SRI can be financially beneficial for investors and funds. (2024-04-20, shares: 2.0)
Gaussian Processes for Volatility: The study enhances the modeling of the implied volatility surface in option pricing by incorporating temporal dynamics into a Gaussian Process, which performs better than traditional models. (2024-04-22, shares: 4.0)
Machine Learning vs Trend Following in Chinese Futures: Machine learning models, especially Multilayer Perceptron (MLP), excel in predicting returns in the Chinese commodity futures market, due to their ability to identify complex patterns and use both volume and price data. (2024-04-22, shares: 2.0)
Neural Networks in Merger Arbitrage: The use of feed forward neural networks (FFNNs) in making merger arbitrage investment decisions proves effective, outperforming other models and increasing risk-standardized deal returns on average. (2024-04-22, shares: 3.0)
Bitcoin Risk in Equity Portfolios: The study indicates that Bitcoin's risk impact on equity portfolios has grown, particularly after COVID-19, highlighting a need for investment professionals to manage Bitcoin-related risks. (2024-04-18, shares: 2.0)
Credit Factor Spillovers: The research identifies 21 bond factors that generate significant positive alpha in bond and CDS markets, with similar factor performance across equity and credit markets and noticeable momentum in bond factors. (2024-04-23, shares: 3.0)
Regime Asset Allocation: The article suggests new portfolio construction methods that use macroeconomic regime information, offering a strategic and analytical alternative to the usual tactical asset allocation approach. (2024-04-19, shares: 2.0)
Option Distributions: The study uses an initial density forecast and monthly index options' bid-ask prices to predict one-month equity index returns, finding that the implied physical significantly improves the initial and implied risk neutral. (2024-04-20, shares: 3.0)
Recently Updated
Quantitative
Machine Learning for Crop Yield Prediction: The study uses machine learning to improve crop yield predictions in eFarming, enhancing forecast accuracy and reliability. (2023-06-10, shares: 2.0)
Crisis Prediction for U.S. Regional Banks: Techniques such as clustering, ridge regression, and sequential feature selection can predict U.S. regional banking crises, helping to improve risk-adjusted returns. (2024-01-18, shares: 2.0)
Financial
Volatility Disagreement: The study uses a model to demonstrate how differing investor opinions on future market volatility can influence the trading of volatility derivatives and impact the stock market, particularly during periods of market instability. (2023-11-29, shares: 2.0)
ETFs Impact: The research shows that ETF ownership and trading have different effects on stock volatility, and that these two mechanisms work together, resolving the debate on the role of ETFs in spreading shocks or providing liquidity to stocks. (2024-03-15, shares: 2.0)
Double-Exponential Jumps: The paper introduces a volatility model that combines double-exponential jumps and GARCH volatility diffusion, effectively capturing major market changes, particularly during the COVID-19 crisis, and suggests its potential use in improving option market fitness and hedging. (2024-04-01, shares: 3.0)
Arbitrage-Free Model: The research presents a four-factor arbitrage-free Nelson-Siegel-Svensson model that is similar to the DNSS model, providing a virtually arbitrage-free DNSS model and broadening the scope for more detailed structural analysis. (2024-04-15, shares: 16.0)
Smart Rebalancing: The article discusses how smart rebalancing can enhance investment strategies by minimizing trading costs, especially in smart beta and factor strategies. (2023-03-11, shares: 2.0)
ETF Indexing: The study reveals that the indexing strategy of ETFs greatly affects the value of its assets, with larger bid-ask spreads for equally weighted ETF index assets. (2024-04-16, shares: 2.0)
Equity Lender Base: The research indicates that short sellers mainly borrow from a few repeated lenders, implying that lending-side issues contribute to market inefficiency. (2022-12-09, shares: 2.0)
Investment Banks in Equity Financing: The study shows that confidentially marketed public offerings (CMPOs) generally raise more capital and attract more investors than registered direct offerings (RDOs), but high-risk firms favor RDOs, and both offerings yield negative average abnormal returns. (2024-01-17, shares: 3.0)
Currency Portfolios: The research suggests that high foreign exchange (FX) ambiguity results in high currency carry returns, indicating that FX ambiguity encompasses aspects of uncertainty not covered by FX volatility. (2024-03-26, shares: 2.0)
ArXiv
Finance
Statistical Edge: The article shows that using past trading data to predict the likelihood of new order execution can give traders a statistical advantage, backed by simulations and real-world trading evidence. (2024-04-22, shares: 6)
Tail Risk: The article proves the joint identifiability and elicitability of tail risk measures and the corresponding quantile, making it easier to fit, compare, and validate models for tail risk measures. (2024-04-22, shares: 5)
Buyback Contracts: The article proposes a new method for pricing and managing share buyback contracts using optimized heuristic strategies, which overcomes the limitations of traditional optimal control methods. (2024-04-21, shares: 5)
Asymmetric CAPM: The article presents a new capital asset pricing model that takes into account the different risks of falling and rising prices, offering a more accurate measure of market risk. (2024-04-22, shares: 4)
Network Contagion Centrality: The article presents a new method for assessing the risk of financial contagion across networks, including a statistical validation technique for practical application. (2024-04-22, shares: 3)
Farmers' Climate Change Adaptation: The study introduces the Environmental, Social, Economics (ESE) score, a comprehensive sustainable credit rating system for farmers, incorporating agricultural sustainability factors into personal credit assessments. (2024-04-22, shares: 2)
Checkerboard Copula Dependence: The research suggests using the checkerboard copula for selecting the copula when not all marginal distributions of a random vector are continuous, as it contains the least information among all possible copulas. (2024-04-23, shares: 2)
Distortion Risk Measures: The paper offers a comprehensive framework for extreme distortion risk measures, establishing precise lower and upper limits based on the first two moments and some shape information of the underlying distributions. (2024-04-21, shares: 2)
Economics
Efficient Stochastic Volatility Analysis: The article discusses a Bayesian analysis of stochastic volatility models, using a new approximation method and applying it to study excess holding yields. (2024-04-22, shares: 7)
Competition in Crowdsourcing: The study examines the balance between competition and collaboration in crowdsourcing communities, noting that increased skill levels lead to more competition but also leniency towards non-threatening members. (2024-04-22, shares: 7)
US Higher Education Skills Dataset: The research uses natural language processing to analyze over three million U.S. course syllabi, creating detailed skill profiles for institutions and academic majors to aid in workforce development research. (2024-04-19, shares: 5)
Firm Growth Models Revisited: The paper explores firm growth models, presenting new theoretical and empirical findings on firm size and growth rates, and suggesting a lack of understanding about the mechanisms driving firm growth. (2024-04-23, shares: 4)
Statewide Travel Forecasting with GPS: A study uses smartphone GPS data to predict travel demand in Indiana, showing a 5-15% difference from traditional models, with growth expected in suburban areas and urban corridors. (2024-04-20, shares: 3)
Natural Capital and Climate Change: The article proposes treating Earth's natural capital as a stock option to maintain and restore ecosystems, in response to the climate crisis caused by excessive fossil fuel use. (2024-04-22, shares: 2)
Global Minimum Tax for Corporations: The Global Minimum Tax on large multinational firms boosts tax revenues in both tax haven and non-haven countries, but gradual rate increases may lead to a split tax rate and lower revenues in non-haven countries. (2024-04-22, shares: 2)
Tracking Economy with Public Ledgers: The article suggests a tool to track every cent in the economy using a serial number and a public ledger, to enhance public spending efficiency and effectiveness, while preserving privacy and enabling statistical analysis. (2024-04-19, shares: 2)
Miscellaneous
Continuous-time Risk-sensitive RL: The article discusses continuous-time risk-sensitive reinforcement learning. It shows its similarity to maintaining the martingale property of a process involving the value function and the q-function. The paper also suggests an algorithm that includes risk sensitivity and proves its effectiveness for Merton's investment problem and its enhanced performance in the linear-quadratic control problem. (2024-04-19, shares: 3)
Crypto & Blockchain
Optimal Liquidity in AMMs: The study presents a model for ideal liquidity provision in automated market makers, indicating that exchange rate volatility increases the optimal transaction fee and the pricing formula is tied to the performance of underlying assets. (2024-04-20, shares: 7)
Blockchain Governance: Liquid Democracy: The research investigates how DAOs can enhance human cooperation, emphasizing the role of the Network Nervous System's staking mechanism in aligning personal interests with the long-term success of the DAO. (2024-04-21, shares: 4)
Historical Trending
Neural Networks for Swaptions Pricing: The article introduces a new method for pricing financial derivatives using advanced deep learning techniques, enhancing efficiency and precision in computational finance. (2024-04-17, shares: 6)
AI in Financial Trading: Recommender Systems: The text explores the use of Artificial Intelligence, particularly Recommender Systems, to mimic traditional asset selection and portfolio construction, integrating AI data analytics with AI-based portfolio construction methods. (2024-04-17, shares: 4)
ArXiv ML
Recently Published
Language Model as Q-Function: The research explores Direct Preference Optimization (DPO) in Reinforcement Learning From Human Feedback (RLHF), showing its ability to assign credit and its similarity to search-based algorithms in language generation. (2024-04-18, shares: 107)
SelfEvolution of Large Language Models: The article discusses self-evolution methods in large language models, providing a conceptual framework and suggesting future improvements. (2024-04-22, shares: 30)
SubRiemannian Bridge Sampling: The study demonstrates a method for bridge simulation on sub-Riemannian manifolds, showing how machine learning can be adapted for training on these manifolds. (2024-04-23, shares: 21)
Historical Trending
DreamerV3: Algorithm Mastery: DreamerV3, a universal algorithm, excels in over 150 varied tasks, including diamond collection in Minecraft without human input, expanding the scope of reinforcement learning. (2023-01-10, shares: 3254)
StructLM: Knowledge Grounding: Despite the limitations of large language models in handling structured data, the new StructLM series, trained on a comprehensive dataset, outperforms task-specific models on 16 out of 18 datasets and sets new benchmarks on 8 Structured Knowledge Grounding tasks. (2024-02-26, shares: 55)
Unified Model Editing Framework: The article presents EMMET, a new algorithm that combines the ROME and MEMIT model editing techniques under the preservation-memorization objective. (2024-03-21, shares: 49)
RePec
Finance
AI in Finance: The article highlights the importance of combining financial expertise with data analytics skills in the era of big data and AI to better manage the financial system. (2024-04-24, shares: 26.0)
Efficient Frontier Kinks: The paper explores the characteristics of efficient frontiers in portfolio optimization, demonstrating the absence of tangency and the universal presence of kinks in portfolio choices. (2024-04-24, shares: 16.0)
Sectoral Contagion Risk: The study uncovers the structure of sectoral risk contagion, emphasizing the need for accurate identification of risk contagion for effective regulation due to strong inter-sector effects. (2024-04-24, shares: 14.0)
Memory-Enhanced Commodity Momentum: The research suggests a memory-enhanced momentum strategy for commodity futures markets, which outperforms traditional momentum and is independent of the overall commodity market movement. (2024-04-24, shares: 12.0)
Statistical
Predicting Systemic Financial Risk: A research suggests that machine learning models and financial stress index can effectively predict systemic financial risk, with stock and money markets being the most influential. (2024-04-24, shares: 26.0)
Cryptocurrency Investor Protection: A study finds that UK cryptocurrency investors tend to favor high-risk investments and lack diversification, influenced by demographic traits, risk tolerance, tech literacy, and emotional attitudes. (2024-04-24, shares: 12.0)
Option-Implied Kurtosis: Including risk-neutral volatility skewness and kurtosis in volatility forecasting models is more accurate than extrapolation, which may lead to less accurate forecasts, according to a research. (2024-04-24, shares: 11.0)
Machine Learning
Predicting Output Trends in China: Machine learning study on Chinese data from 1993-2016 reveals credit as a better output predictor than money, but its predictive power has lessened post-2007 due to financial development. (2024-04-24, shares: 28.0)
Challenges in Reusing ML Applications: The article categorizes machine learning applications into four types based on reuse strategies and offers insights for their development and deployment. (2024-04-24, shares: 23.0)
CostSensitive ML for Startup Investments: Machine learning models used to predict the success of Israeli startups can reduce investment risk, but may also limit potential profits by predicting fewer successful startups. (2024-04-24, shares: 22.0)
Addressing sample bias in ML: The research suggests two control function methods to improve machine learning accuracy when training and prediction samples differ, reducing prediction error and selection bias. (2024-04-24, shares: 17.0)
ML in hierarchical time series forecasting: A multi-output regression model using variables from different hierarchical levels can provide reliable forecasts for supply chain decisions, especially during deep promotional discounts. (2024-04-24, shares: 13.0)
GitHub
Finance
Machine Learning for Algorithmic Trading: HandsOn Machine Learning for Algorithmic Trading is a new book by Packt that focuses on the application of machine learning in trading algorithms. (2019-05-07, shares: 1293.0)
Data Quality and Machine Learning: The article discusses an AI package designed to enhance the quality of unorganized real-world data used in machine learning. (2018-05-11, shares: 8605.0)
Bitcoin Volatility Forecasting: The piece investigates the use of GARCH and Multivariate LSTM models for predicting Bitcoin volatility, useful in crypto trading and risk management. (2021-07-18, shares: 188.0)
Supplemental Material for Algorithmic Trading: The article offers extra resources for understanding algorithmic trading and quantitative strategies. (2020-01-20, shares: 238.0)
Trending
Bridging LLM and Recommender System: The article explores the combination of Language Model and Recommender System to improve user interaction. (2023-09-07, shares: 302.0)
Productionise Jupyter Notebooks: The piece offers a tutorial on efficiently scheduling and operationalizing Jupyter Notebooks. (2020-02-28, shares: 842.0)
Scikitlearn compatible neural network: The article evaluates a scikit-learn compatible neural network library that uses PyTorch for sophisticated machine learning. (2017-07-18, shares: 5618.0)
LinkedIn
Trending
FTC Bans Non-Competes in Quant Industry: The FTC has prohibited noncompete clauses, marking a significant change in employment practices in the quant industry. (2024-04-24, shares: 1.0)
ML Asset Allocation: Recent research suggests a novel asset allocation method that uses optimal portfolio weights in a supervised learning algorithm, leading to improved Sharpe ratios and lower trading expenses. (2024-04-23, shares: 1.0)
Decomposing Options P&L: Olivier Daviaud from JP Morgan provides insights on dissecting the profit and loss of an options portfolio from a buy-side firm's viewpoint. (2024-04-24, shares: 1.0)
Diversification with Markku Kurtti: Electrical engineer and investor, Markku Kurtti, shares his unique perspective on diversification and its associated opportunity costs. (2024-04-23, shares: 1.0)
Harvesting Tail Risk Diversification: Julien Royer emphasizes the significance of diversification in portfolio building, especially with regards to alternative risk premia strategies. (2024-04-24, shares: 1.0)
Informative
MultiStrats Secrets Revealed: Financial experts will discuss the rapid growth of MultiStrats in asset management at the Battle of the Quants - Worldwide event on May 9th. (2024-04-24, shares: 1.0)
Prof John Mulvey's Recap: QWAFAxNEW members attended a talk by Prof John Mulvey on his paper about multi-step approach for asset allocation in regime-aware portfolio models. (2024-04-24, shares: 1.0)
Fathmat Bakayoko AI Conference: Fathmat Samira Bakayoko is co-leading a project to create an algorithmic trading strategy based on real-time analysis of SEC 8-K reports and additional disclosures. (2024-04-23, shares: 1.0)
Energy Landscape Theory: The HTSR theory uses the Energy Landscape theory from protein folding to create a complex energy landscape in AI, based on the activations at each layer of a model. (2024-04-23, shares: 1.0)
Future of Asset Allocation: The AIM Summit London will host a panel discussion on future trends and shifts in asset allocation, featuring experts from Goldman Sachs, University Pension Plan Ontario, and Citi. (2024-04-23, shares: 1.0)
Predict Electricity Prices: The 2024 Challenge Data, organized by ENS and Institut Louis Bachelier, calls on data scientists to predict electricity prices to aid green electricity provider Elmy's purchasing strategies. (2024-04-24, shares: 1.0)
Podcasts
Quantitative
Navigating ETFs with James Sayffart: James Sayffart predicts that ETFs, particularly Bitcoin ETFs, will dominate over mutual funds due to the complexities of the mutual fund industry. (2024-04-18, shares: 14)
Reading the Market with AI: Barry Ritholtz interviews Ashish Shah about his career and roles in various financial institutions, including Goldman Sachs Asset Management LP. (2024-04-19, shares: 12)
Insights on Trading Strategies with Nick Baltas: Nick Baltas explains the differences between alpha, beta, smart beta, and factors, and discusses why momentum is a profitable strategy in trading. (2024-04-19, shares: 10)
Related
Gods Among Men: In a podcast, Professor Ken French discusses his differing views with Professor Eugene Fama, the idea of long-term investing, and common misunderstandings about stock buybacks. (2024-04-19, shares: 8)
Geopolitics and Real Yields: The LGIM Real Assets Research Team predicts four major trends that will influence private investment performance and capital distribution for the upcoming decade. (2024-04-22, shares: 7)
News
Quantitative
Jane Street's Restraining Order Failure: Jane Street Capital's attempt to secure a restraining order against Millennium Management over alleged trading strategy theft was unsuccessful. (2024-04-22, shares: 4)
IMF Warns of Hedge Fund Risk: The IMF has raised concerns about potential financial instability due to a few hedge funds dominating short positions in the US Treasury futures market. (2024-04-19, shares: 3)
Twitter
Quantitative
Predicting Trading Volume: Article 1: A study uses machine learning to accurately predict trading volume based on various factors including technical signals and firm characteristics. (2024-04-23, shares: 5)
Top Trading and Investing Books: Article 2: The article suggests six informative books on trading, investing, and portfolio management. (2024-04-20, shares: 2)
GBRT for Order Book Modeling: Article 3: The article explores the application of Gradient Boost Regression Tree in Limit Order Book modeling. (2024-04-18, shares: 2)
Equities Premium Challenges: Article 1: The article highlights the challenges in predicting the risk of rare disasters and their impact on equity risk premiums due to limited data on macroeconomic disasters. (2024-04-22, shares: 1)
Calibrating Models: Article 2: The article discusses the finance sector's efforts to improve technologies for calibrating models, which have previously led to substantial financial losses due to poor calibration. (2024-04-22, shares: 1)
Whisper Training Data: Article 3: The article reports on accusations against OpenAI for allegedly using illegally obtained data to train its Whisper system. (2024-04-21, shares: 1)
Miscellaneous
Forecasting Tools Underperform: Forecasting tools such as Prophet, TIDE, and XGBoost often have difficulty accurately predicting financial data sets that cover several business cycles and new turning points. (2024-04-19, shares: 1)
Dynamic Factor Modeling with Python: Metran, a Python library, utilizes Dynamic Factor Modeling and auto regressive techniques for a variety of applications, not just hydro timeseries. (2024-04-18, shares: 1)
Analyst Forecast Dispersion Predicts: A recent study shows that the dispersion of analyst forecasts usually predicts stock returns negatively. (2024-04-22, shares: 0)
Optimizing Output Quality: Article 1: The article offers tips on enhancing the quality of interactions with Language Model Machines. (2024-04-19, shares: 0)
PseudoMathematics in Finance: Article 2: The article reviews a 2014 paper about pseudomathematics and financial fraud, particularly the impact of backtest overfitting. (2024-04-19, shares: 0)
Stock vs. Bond Trends: Article 3: The article posits that Democrats boost the stock market and Republicans favor the bond market, with both showing stronger momentum under the GOP. (2024-04-18, shares: 0)
Paper with Code
Trending
State Space Model: The article offers a detailed review of SSM, including experimental comparisons and analysis to underline its features and benefits. (2024-04-18, shares: 196.0)
Beomi: The article presents a method for efficiently scaling Transformer-based Large Language Models (LLMs) to manage infinitely long inputs within limited memory and computation. (2024-04-18, shares: 87.0)