Quant Letter: December 2023, Week-3

Weekly (27th Edition)

SSRN

Recently Published

Quantitative

Tactical ETF Strategy: Volatility Risk Premia & Crisis Alpha Harvesting: The author proposes a systematic method for investing in volatility risk premia via ETFs, utilizing futures contracts for backtesting data. (2023-12-17, shares: 2.0)

Online Portfolio Selection with Deep Sequence Features and Reversal Info: A new algorithm uses machine learning and financial data to improve online portfolio selection and automated trading. (2023-12-19, shares: 2.0)

Fractional Regularization for Sparse Portfolio Optimization: The paper introduces a new L1L2 regularized sparse portfolio optimization model using the ADMM method, and discusses an extension of the model to include a more general L1Lq regularization. (2023-12-17, shares: 3.0)

Financial

ETF Cross-Arbitrage: Due to the unique characteristics of the ETF lending market, ETFs are costlier to borrow than stocks, creating profitable opportunities for cross-ETF arbitrage. (2023-12-18, shares: 2.0)

Cryptocurrency Carry Trade: Risk and Return: Cryptocurrency carry trade provides high returns that are not solely explained by cryptocurrency factors or geopolitical risks, indicating a significant part of returns is a premium for equity market volatility risk. (2023-12-15, shares: 6.0)

Resilience for Stronger Investment Portfolios: The article promotes an adaptive investment strategy that focuses on resilience thinking, active ownership, and moving away from narrow financial models due to fast-paced technological, geopolitical, and environmental changes. (2023-12-18, shares: 2.0)

Outperforming Equal Weighting: The article suggests that an equally-weighted stock portfolio can be improved by avoiding negative exposure to certain factor anomalies, while keeping the portfolio construction process simple. (2023-12-19, shares: 6.0)

Recently Updated

Quantitative

Inflation Forecasting with Economic Narratives: The research finds that economic narratives from Wall Street Journal articles and machine learning algorithms can accurately predict inflation, particularly during economic downturns. (2022-08-04, shares: 2.0)

Volatility Modeling in Asset Markets: The paper investigates the volatilities of nine asset markets from 2013 to 2021, identifying three factors affecting volatility and a strong correlation in the volatility of Iranian stock returns. (2023-10-01, shares: 3.0)

DEX Arbitrage with Deep Reinforcement Learning: The study explores trading performances under arbitrage conditions in decentralized exchanges, using a simulation model and deep reinforcement learning to determine optimal arbitrage strategies for eight cryptocurrency pairs. (2023-04-30, shares: 3.0)

Financial

Market Liquidity Estimation with Machine Learning: Machine learning is used to estimate the average daily bid-ask spread in the US and Chinese stock markets, enhancing performance by capturing more raw data and utilizing learned nonlinear relationships. (2023-03-03, shares: 2.0)

Machine Learning for Portfolio Selection: A new performance ratio is created to address the limitations of the Sharpe ratio under non-Gaussian returns and systemic risk, showing improved portfolio selection performance in terms of profitability and risk reduction. (2023-08-11, shares: 2.0)

Investors' Risk and Return Expectations: The study analyzes risk and return expectations on 19 asset classes from 1987 to 2022, highlighting a strong risk-return tradeoff and the predictive power of expected returns. (2023-05-25, shares: 609.0)

Market Ambiguity and Risk-Return Tradeoff: The risk-return balance in the stock market is affected by the investor's attitude towards ambiguity, with increased market volatility causing a decrease in the equity premium's slope when market optimism is high. (2023-09-21, shares: 2.0)

Automated Market Makers: Revolutionizing Finance: Automated Market Makers (AMMs) can make liquidity provision more accessible and potentially create deeper markets for high-volume, low-volatility assets, resulting in lower trading costs than traditional markets. (2023-05-26, shares: 2.0)

Short-Term Signals: Unlocking Alpha: Investors can achieve substantial net alpha by combining short-term signals with advanced trading rules in a liquid global universe, which helps to reduce transaction costs. (2022-06-01, shares: 3.0)

Unifying Economics and Finance: Solving Equity Premium Puzzle: The issue of a high equity premium and a low risk-free rate, known as the equity risk-premium and volatility puzzle, remains unsolved as current economic models fail to provide a consistent explanation. (2021-03-22, shares: 2.0)

ArXiv

Finance

Return-Diversification Approach for Portfolio Selection: The article suggests a dual-objective model for portfolio selection that optimizes both diversification and expected return, outperforming strategies focused only on diversification or risk-return. (2023-12-15, shares: 8)

Order Size Modelling in Limit Order Book Dynamics: The research introduces a new method using Compound Hawkes Process to model Limit Order Book dynamics, taking into account order size and maintaining a positive spread. (2023-12-14, shares: 5)

Volatility Term Structure in Robust Option Pricing: The research examines the robust option pricing issue, discovering that adding more information does not enhance the robust pricing bounds, contrary to popular belief. (2023-12-14, shares: 4)

Residual U-net for Multi-Agent Trade Execution: The paper discusses the application of a deep residual U-net with self-attention to solve the continuous time-consistent mean variance optimal trade execution problem for multiple agents and assets, surpassing the constraints of finite difference methods. (2023-12-14, shares: 4)

Risk Budgeting and Diversification: A new framework for Risk Budgeting in portfolio optimization is introduced, which balances risk from assets and factors using various risk measures. (2023-12-18, shares: 4)

Convergence of Hawkes Processes in Financial Markets: The research identifies the weak convergence of a nearly-unstable Hawkes process with a heavy-tailed kernel, useful for creating a scaling limit for a financial market model. (2023-12-14, shares: 3)

Managing ESG Ratings in Sustainable Portfolios: A nonlinear optimization model for portfolio selection considering risk, return, and ESG criteria is proposed, resolving discrepancies between different agencies' ESG ratings. (2023-12-17, shares: 3)

Learning Merton's Strategies in Incomplete Market: The study applies reinforcement learning to determine optimal portfolio policies in an incomplete market, showing its efficiency and robustness compared to the traditional plug-in method. (2023-12-19, shares: 3)

Crypto & Blockchain

Implications of Artificial Latency in PBS: The study examines the effects of artificial latency in the Ethereum network's Proposer-Builder Separation framework, highlighting increased profits for node operators but also potential network inefficiencies and centralization risks. (2023-12-15, shares: 6)

Blockchain Risk Parity: Efficient Investing: Blockchain technology is being utilized to create risk-managed portfolios with three different funds, each inversely related to the asset's risk, giving investors the ability to choose their preferred risk or return level. (2023-12-12, shares: 3.0)

Historical Trending

Futures Price Prediction with Graph Neural Networks: A new model for predicting futures prices in high-frequency trading, using graph neural networks, outperforms existing models in China's futures market. (2023-03-29, shares: 23)

Linking Investor Expectations and Market Price Movement: A study introduces a model for predicting market asset prices based on the correlation between investors' expectations and market price movement. (2019-12-24, shares: 22)

Portfolio Evaluation with Rewards: The study examines how periodic reward structures affect long-term portfolio strategies, especially when short-selling is not allowed, by transforming the issue into a single-period optimization problem. (2023-11-21, shares: 13)

Agent-Based Modeling with Language Models: The research introduces Smart Agent-Based Modeling (SABM), a new framework that combines Large Language Models with Agent-Based Modeling to simulate real-world situations more accurately, as demonstrated in three case studies. (2023-11-10, shares: 13)

ArXiv ML

Recently Published

Time-Warp-Attend: Learning Topological Invariants in Dynamical Systems: The study proposes a deep-learning framework for classifying dynamical regimes and identifying bifurcation boundaries in different systems, offering insights into large-scale physical and biological systems. (2023-12-14, shares: 8)

Maximizing Non-differentiable Objectives: The guide introduces reinforcement learning as an extension of supervised learning, offering an easy-to-understand method for learning advanced deep reinforcement learning algorithms such as proximal policy optimization. (2023-12-13, shares: 29)

RePec

Machine Learning

Impact of Evaluation Metrics on ML Models for Stock Market Indices: The study reveals that the choice of machine learning algorithm significantly affects the financial performance of trading systems, with the random forest algorithm proving most effective. (2023-12-20, shares: 32.0)

ML Methods for Selecting Mutual Funds with Positive Alpha: Machine-learning methods can help select profitable mutual fund portfolios, with the study indicating that past performance predicts future performance for active funds, benefiting investors with access to advanced prediction methods. (2023-12-20, shares: 22.0)

Predicting Corporate Credit Ratings with ML: The research recommends restricted CART models for predicting corporate credit ratings using machine learning techniques, emphasizing the role of company size in credit rating prediction. (2023-12-20, shares: 20.0)

Finance

Asset Bubbles and Trading Strategies: The chapter discusses trading strategies in a single risky asset market with a price bubble, showing that wealth preserving strategies can outperform simply holding the asset. (2023-12-20, shares: 37.0)

Short Selling and Arbitrage: The study reveals that arbitrage opportunities in financial markets can only be exploited through short selling, emphasizing the bankruptcy risk involved. (2023-12-20, shares: 29.0)

Advances in Mathematical Finance: The book celebrates Peter Carr's contributions to Quantitative Finance, featuring new research results and tributes from family and friends. (2023-12-20, shares: 26.0)

Volatility Estimators for Cryptocurrencies: The paper studies the realized volatility of cryptocurrencies, showing that the best predictors for Bitcoin and Ethereum come from 30-day implied volatility. (2023-12-20, shares: 24.0)

EMA Trading Strategies with Partial Information: The study investigates optimal trading strategies for a partially informed trader under Gaussian price dynamics, proving that optimal strategies depend on current price and an exponentially weighted moving average price. (2023-12-20, shares: 24.0)

Total Positivity and Convexity in Options: The chapter explores total positivity and relative convexity properties in option pricing models, demonstrating that these properties generally hold in time-homogeneous local volatility models. (2023-12-20, shares: 23.0)

Derivatives' Risks in a Network Model: The paper introduces a one-period XVA model for bilateral and centrally cleared trading, illustrating its potential for stress testing a financial network or optimizing a defaulted clearing member's portfolio. (2023-12-20, shares: 23.0)

Modified Stochastic Volatility Model for Derivative Pricing: The article suggests an improved 4/2 stochastic volatility model with a new formula for derivative prices, enhancing calibration speed and capturing market volatility. (2023-12-20, shares: 22.0)

Volatility Spillovers between Oil and Financial Markets: The article uses a GARCH-VAR-Spillover Index method to study the two-way volatility relationship between oil and stock markets during financial crises. (2023-12-20, shares: 20.0)

Backtestability and the Ridge Backtest: The article proposes a formal definition of backtestability for a statistical functional of a distribution and explores its connection with elicitability and identifiability. (2023-12-20, shares: 19.0)

Effectiveness of Short-Term Market Swings in Predicting Realized Volatility: The article assesses the new VIX1D volatility index's effectiveness in predicting short-term market fluctuations and realized volatility. (2023-12-20, shares: 18.0)

Alternative Data and Trade Credit Financing: The research reveals that the use of alternative data, specifically online sales data, boosts trade credit financing for firms in China. (2023-12-20, shares: 18.0)

Jumps and Gold Futures Volatility Prediction: The article studies the efficiency of the jump component in predicting Chinese gold futures volatility using high-frequency data. (2023-12-20, shares: 17.0)

ChatGPT as a Quant Asset Manager: The research suggests a quantitative investment approach that includes recommendations from ChatGPT, demonstrating its potential to enhance portfolio efficiency. (2023-12-20, shares: 16.0)

GitHub

Finance

Scalable Timeseries ML with Polars: The article explores the application of Polars in large-scale timeseries machine learning, particularly in parallel feature extraction and panel data forecasts. (2023-06-05, shares: 677.0)

Minimal ML Algorithm Implementations: The article provides practical examples of how to implement different machine learning algorithms. (2016-10-05, shares: 9779.0)

Deep RL for Portfolio Optimization: The article investigates the use of Deep Reinforcement Learning for optimizing investment portfolios. (2020-03-04, shares: 71.0)

pyflux: Time Series Library for Python: The article presents an open-source library designed for handling time series data in Python. (2016-02-16, shares: 2069.0)

Safe RL Baselines Repository: The article introduces a repository focused on safe reinforcement learning baselines. (2022-02-22, shares: 329.0)

Trending

Interactive ABIDES Simulation: ABIDES is a simulation system that enables interactive, agent-based, discrete event modeling. (2019-03-06, shares: 332.0)

AI Custom Builds: The repo discusses the need for precise instructions for AI systems to produce accurate results. (2023-04-29, shares: 46808.0)

Opensource Language Model: KnowLM: A new open-source framework has been created for handling large language models with extensive knowledge. (2023-04-01, shares: 821.0)

Genetic Feature Selection: sklearngenetic: A new scikitlearn module has been introduced that utilizes genetic algorithms for feature selection. (2016-06-09, shares: 289.0)

News & Social

Quant Trading Firms Respond to NeurIPS AMDs Chip Allegations: Quant Trading Firms' use of NeurIPS AMDs chips has sparked a dispute with Nvidia, as per The Information. (2023-12-18, shares: 4)

Market Making Model Analysis in HFT: A paper presents a straightforward market making model for high frequency trading in the North American stock market, without including performance analysis. (2023-12-19, shares: 1)

Quant Investing with AQR: An article delves into quantitative investing and AQR, offering valuable insights into the sector. (2023-12-15, shares: 1)

Christmas Quant Gift Ideas: The article suggests Christmas gift ideas related to quantitative analysis, including books on various financial and mathematical topics and mechanical pencils. (2023-12-17, shares: 17.0)

Computation Challenges with Matrix Multiplication: The article discusses the importance of matrix multiplication in statistics and AI, highlighting the challenges posed by large matrices and high computational costs. (2023-12-14, shares: 6)

Paper with Code

Rethinking UNet Encoder in Diffusion Models: A novel technique has been found that skips the encoder at certain adjacent timesteps and cyclically reuses the encoder features from previous timesteps for the decoder. (2023-12-20, shares: 140.0)

Pointcept: Simpler, Faster Point Transformer V3: The article does not seek to introduce any new concepts or improvements to the attention mechanism in machine learning. (2023-12-20, shares: 119.0)

Agent Attention: Softmax and Linear Attention Integration: Agent Attention enhances the traditional attention module by incorporating an extra set of agent tokens. (2023-12-19, shares: 73.0)