Quant Letter: September 2023, Week 4

Weekly quantitative finance newsletter

ArXiv

Finance

Hedging Algorithmic Strategies with Ensemble AIS: The study suggests a new method for hedging risky asset portfolios using ensemble algorithmic investment strategies, finding Bitcoin-based AIS to be the best diversifier for S&P 500 index-based AIS. (2023-09-27, shares: 8)

Practical Risk Management and Hedging: The research introduces a simple method for managing and hedging cross-asset portfolio risk using quadratic programming, emphasizing the link between economic ideas and their mathematical expressions. (2023-09-27, shares: 7)

Error Bounds for Deep Calibration of Option Prices: The paper offers quantitative error limits for deep neural networks approximating option prices on a risky asset, demonstrating that DNNs can learn option prices with minimal error without the curse of dimensionality. (2023-09-26, shares: 6)

ATM Skew in ADO-Heston Model: The paper presents a Markovian approximation of the ADO-Heston model, challenging previous beliefs that such approximation cannot replicate the behavior of the vanilla implied skew at small T. (2023-09-26, shares: 4)

Econometric Model for Oil Stocks: The study uses Arbitrage Pricing Theory and Quantile Regression to analyze the risk and return of crude oil stocks, identifying key risk factors and the impact of changes in West Texas Intermediate returns. (2023-09-22, shares: 4)

Performance Evaluation of Equal-Weight Portfolio: The paper compares three portfolio design strategies using historical stock prices from the National Stock Exchange of India, identifying the highest return yielding portfolio for each sector. (2023-09-24, shares: 3)

Miscellaneous

AI and VC Portfolio Simulation for Predicting Startup Success: A new deep learning model can predict the success of startups at Series B and C investment stages, using data like funding metrics and founder characteristics. (2023-09-27, shares: 8)

ML-Assisted Surrogates for Tipping Point Detection in Complex Systems: A new machine learning framework can identify critical changes in complex systems and calculate the likelihood of rare events, using detailed spatiotemporal data. (2023-09-25, shares: 6)

Consistent Explanations in Science-Informed ML Models: An Investigation: The study investigates the reliability of scientific explanations when using explainable machine learning methods on science-based machine learning models, showing varying results based on the type of monotonicity. (2023-09-23, shares: 6)

Gray-box Attack on Deep RL Trading Agent: A study has shown that a gray-box method can significantly reduce the profits of a Deep Reinforcement Learning-based trading agent, highlighting the need for stronger automated trading systems. (2023-09-26, shares: 3)

Recommendation Algorithms in Algorithmic Collusion: Research indicates that recommendation algorithms on e-commerce platforms can affect the dynamics of AI-based pricing algorithms, with profit-based systems promoting collusion and demand-based systems encouraging competition. (2023-09-25, shares: 2)

Detection of Temporal Money Laundering Flows: The newly proposed FaSTMAN framework improves the detection of suspicious money flows in large-scale transaction data by efficiently constructing a temporal graph of sequential transactions, surpassing two leading solutions. (2023-09-24, shares: 2)

Crypto & Blockchain

Uncertainty and Maximal Extractable Value in Decentralized Systems: The study suggests a combined approach to decrease Maximal Extractable Value in decentralized systems, indicating that neither fair ordering nor economic mechanisms alone can mitigate MEV effectively. (2023-09-25, shares: 4)

The Costs of Trading on Uniswap: Efficiency and Slippage Analysis: The research analyzes the costs of trading on a decentralized exchange, showing that costs vary based on trade characteristics, and proposes that DEXs could be a trustworthy alternative to centralized exchanges for trading digital assets. (2023-09-24, shares: 3)

Historical Trending

Valuation of Derivative Contracts: The research introduces a model for assessing a vulnerable derivative with bilateral cash flows, taking into account funding, credit, and wrong-way risks. (2023-08-21, shares: 29)

Optimal Portfolio Liquidation with Mean-Field Control: The research demonstrates a mean-field control problem in portfolio liquidation models, showing that the value function follows a linear-quadratic form. (2022-07-01, shares: 14)

Rolling Stock Planning Optimization with Maintenance: The paper compares Constraint Programming and Quantum Annealing for optimising rolling stock assignment, with both methods yielding similar results. (2021-09-15, shares: 13)

Enhanced Reliability Estimation for Numeric Predictions: The article suggests a machine learning method to enhance the accuracy of individual predictions in numerical predictive modeling, compatible with any prediction model and not needing distributional assumptions. (2022-07-28, shares: 20.0)

Stochastic Control Policy Gradient Learning: The paper introduces new policy gradient methods for stochastic control with exit time, showing they perform better than traditional methods and can adapt to realistic market conditions. (2022-06-20, shares: 2.0)

Market Consensus Scoring Probability Forecasts: The paper presents a probability scoring rule that evaluates a forecaster's economic performance against market consensus, promoting better information gathering and discouraging groupthink. (2021-07-16, shares: 2.0)

SSRN

Recently Published

Quantitative

Tail-Event Analysis in Imbalanced Time Series Data: The FSDA method combines feature selection and data augmentation to enhance the accuracy of machine learning models in predicting rare events in imbalanced financial data. (2023-09-24, shares: 2.0)

Deep RL vs Traditional Financial Portfolio Methods: Deep reinforcement learning (DRL) surpasses traditional methods in portfolio management, offering better risk-adjusted return and Sharpe ratio. (2023-09-24, shares: 18.0)

Reverse Stress Testing for ALM: The article presents a new toolkit that uses AI and yield curve modelling to detect potential risks in bank balance sheets, illustrated with two hypothetical banks. (2023-09-25, shares: 15.0)

Volatility Spillovers in Asymmetric Covariance: The research introduces a new method for assessing volatility spillovers in asymmetric realized covariance, proving its effectiveness with high-frequency data from major ETFs. (2023-09-25, shares: 4.0)

Multiscale Perspective on Volatility Spillover Network in Chinese Financial Market: The research investigates the progression of systemic risk in China's financial market using a new multiscale decomposition method, uncovering varying network characteristics in high and low volatility spillover networks. (2023-09-25, shares: 4.0)

Short-Selling Cost and IV Spreads in Chinese SSE 50ETF Options Market: The article explores the link between option-implied volatility spreads and option-implied borrow rate in Chinese SSE 50 ETF options, discovering a significant negative correlation and nonlinearity. (2023-09-25, shares: 2.0)

Classifying Trademark Distinctiveness with GPT-3: The article discusses the application of Large Language Models (LLMs) and machine learning in assessing trademarks for registration, showing how an LLM can help identify issues and prepare data for machine learning algorithms. (2023-09-24, shares: 2.0)

Financial

Enhancing Realised Volatility Forecast for Emerging Markets: A study comparing four models for forecasting volatility in emerging markets found the HAR model best for long-term volatility and realised GARCH models for volatility clustering and persistence. (2023-09-26, shares: 3.0)

Limited Impact of Big Three Engagements on Firm Performance: Research indicates that BlackRock, Vanguard, and State Street, the three largest asset managers, have limited capacity for active ownership due to their engagement targets and resource constraints. (2023-09-26, shares: 34.0)

Tax Benefits of Tax-Aware Long-Short Strategies: Tax-aware long-short factor strategies can lead to net capital losses exceeding 100% of the initial investment within three years, due to deferred capital gains, not increased capital losses. (2023-09-26, shares: 4.0)

Credit Line Premium: According to the research, firms with more unused credit lines have higher returns due to greater liquidity needs, but are also more vulnerable to economic shocks. (2023-09-25, shares: 2.0)

Dealer Intermediation and Recovery Post-Default: The research reveals that defaulted U.S. corporate bonds are traded to dealers with prior expertise in the bond, improving recovery rates and emphasizing the importance of dealer expertise. (2023-09-22, shares: 3.0)

Investment and Trading Efficiency Metrics: A study emphasizes the need to measure risk-adjusted returns in investments and trading, highlighting metrics like the Sortino Ratio, Calmar Ratio, and Pareschi Ratio. (2023-09-26, shares: 20.0)

Risk Management through Predictive Simulation: A paper suggests using a simulation approach with mortality-linked securities and stochastic mortality rates to manage capital risk in the insurance industry and meet regulatory capital requirements. (2023-09-23, shares: 2.0)

Superhuman Speed in Futures Trading Requires Stricter Regulation: A study suggests that high-frequency traders' use of low-latency trading algorithms for arbitrage opportunities increases execution costs for other market participants, proposing batch auctions as a solution. (2023-09-22, shares: 4.0)

Recently Updated

Quantitative

Financial Statement Fraud: News Analysis: A new system using news coverage and machine learning can accurately detect financial fraud in Chinese companies from 2001 to 2022. (2023-01-27, shares: 236.0)

Modeling Nested Data in Operation Research: A paper suggests the contextual effects model as a better method for handling nested data in research, over fixed effects and multilevel models. (2023-08-28, shares: 2.0)

CSRM Assurance's Impact on Nonprofessional Investors' Judgments: Egyptian investors' decisions are significantly influenced by the reliability of cybersecurity risk management, according to research. (2023-08-06, shares: 2.0)

Public Asset Purchases and Private Risk Sharing: Government interventions that increase interest rates enhance risk sharing efficiency, while those that lower rates promote inefficient risk-taking, a study suggests. (2023-02-10, shares: 2.0)

Idiosyncratic Volatility and Stock Returns in Industries: A higher level of idiosyncratic volatility predicts lower stock returns, especially in central industries more exposed to such shocks. (2023-09-01, shares: 3.0)

The Future of Tax Law: Legal Singularity Ahead: The idea of legal singularity, where law becomes entirely comprehensive and predictable, could be achieved through AI and new technologies in tax law. (2023-08-28, shares: 2.0)

Financial

Corporate Bond Funds and Underpriced New Issues: From 2002-2019, active investment-grade corporate bond mutual funds saw significant positive returns, largely due to underpriced new bond offerings to funds with strong underwriting relationships. (2022-10-29, shares: 121.0)

Data Breach Impact on Bank Operations: Data breaches at banks lead to a loss of insured and brokered deposits and negatively impact stock returns, but do not affect long-term operations; banks often increase lending after a breach, likely due to CEO compensation incentives. (2023-03-11, shares: 74.0)

Enhanced Model for Valuing S&P 500 and VIX Options: Current pricing models for SP 500 and VIX options are insufficient, but a new model controlling higher-order moments of risk-neutral return distribution outperforms alternatives and reconciles the two markets. (2021-01-09, shares: 155.0)

Dynamic Price Impact Model: A dynamic model reveals that the predictability of noise trading flows influences return predictability at both individual asset and factor levels, potentially causing asset price bubbles when flows show excessive momentum. (2022-11-23, shares: 234.0)

Diversification Paradox and Uncertainty Allocation: Firms diversify investments into various industries to mitigate risks, but this can result in decreased growth, equity value, and financial constraint, leading to inefficient resource allocation and a drop in overall productivity. (2023-09-20, shares: 8.0)

Pension Plan Systems and Risk Sharing: Including defined benefit pension funds in an asset pricing model enhances its performance in matching historical equity premium and riskless rate, and offers significant risk sharing benefits. (2023-04-14, shares: 124.0)

Stock Market Index Construction Challenges: The process of creating an index, particularly in less active stock markets, greatly influences the index and its statistical characteristics, as shown by a historical Finnish stock market database. (2020-11-16, shares: 73.0)

VIX Modeling for Insiders: The Barndorff-Nielsen Shephard volatility model is expanded to include a jump Ornstein-Uhlenbeck equation with a non-zero stochastic mean-reversion level, offering insights into variance swap pricing. (2022-05-31, shares: 86.0)

Volatility and Regime Switching in Sustainable Indices: The performance and volatility of sustainable indices are influenced by regime-switching, with positive shocks impacting volatility differently than negative ones, as demonstrated by a study of major sustainable indices from 2009 to 2017. (2021-07-29, shares: 2.0)

Deep Learning

Bayesian ANN for Efficiency Analysis: The study presents a novel method for frontier estimation in econometrics, merging Data Envelopment Analysis and Stochastic Frontier Analysis via Bayesian artificial neural networks, tested on a dataset of large US banks. (2023-09-28, shares: 16.0)

GitHub

Finance

Temporian: Temporal Data Engineering and Augmentation: Temporian is a Python tool that improves and expands time-based data for machine learning uses. (2023-01-17, shares: 140.0)

Obanalytics: Limit Order Book Data Visualization and Analysis: R package is a tool for examining, visualizing, and rebuilding limit order book data. (2015-10-06, shares: 140.0)

Vnpy: Quantitative Trading Platform Development: An open-source structure is available for creating quantitative trading platforms using Python. (2023-08-14, shares: 43.0)

Betterquant2: Programmatic Quantitative Trading: QUANT is a software designed for quantitative and algorithmic trading. (2023-09-11, shares: 31.0)

LongLoRA: Efficient Longcontext Finetuning for LongQA: Efficient longcontext finetuning is a technique for guided finetuning of the LongQA dataset. (2023-09-21, shares: 361.0)

Trending

AutoGPT: GPT4 Autonomy: The article explores an open-source project that aims to make GPT4 fully self-sufficient. (2023-03-16, shares: 149153.0)

mtwebsockets: Metatrader4MetaTrader5 as Web Socket Client: The piece provides a guide on transforming MetaTrader4/5 into a web socket client. (2020-01-19, shares: 47.0)

kani Hackable Microframework for Chatbased Language Models: The article presents kani カニ, a microframework for chat-based language models with customizable tool usage. (2023-07-14, shares: 359.0)

sqlc Typesafe Code Generation from SQL: The article offers a tutorial on creating type-safe code from SQL. (2019-06-21, shares: 9104.0)

catala: Programming Language for Literate Programming Law Specification: The piece introduces a new programming language designed for literate programming law specification. (2020-04-17, shares: 1754.0)

LinkedIn

Trending

AI Forecasting: Theory and Applications: A new book, Forecasting with Artificial Intelligence: Theory and Applications, provides in-depth case studies on data mining techniques and predictive algorithms. (2023-09-28, shares: 1.0)

Big Data Forecasting with Global Models: The author has written two chapters on global forecasting models in a book about artificial intelligence applications in forecasting. (2023-09-28, shares: 1.0)

Top Position in M6 Financial Duathlon: The author's team secured first place in the M6 Financial Duathlon Competition by using historical data and risk mitigation strategies for successful forecasting and investment. (2023-09-28, shares: 1.0)

Informative

Calibrating Market Impact Model: Nicholas Westray and Kevin Webster published a paper on adjusting a market impact model in the face of biased trading data. (2023-09-28, shares: 1.0)

Training LLMs for Quant Finance: The 19th Quantitative Finance Conference will include a presentation by Ioana Boier from NVIDIA on managing LLMs for quantitative finance. (2023-09-28, shares: 2.0)

Guide to Natural Language Processing: The book Natural Language Processing with Transformers is a useful guide for quantitative researchers, even those not primarily focused on NLP. (2023-09-28, shares: 1.0)

Higher Returns of High R&D Stocks: Larry Swedroe's blog suggests that high R&D stocks yield higher returns due to increased systematic risk not considered in standard asset pricing models. (2023-09-28, shares: 1.0)

Short-Term Futures Quant Traders in NYC or Remote (US): A leading proprietary trading group is hiring Quant Portfolio Managers for short-term futures trading, with competitive pay and remote work options in the US. (2023-09-28, shares: 1.0)

Verifiable Sequencing Rules for Decentralized Exchange Design: A Fordham University seminar will focus on creating reliable decentralized exchanges using verifiable sequencing rules to prevent price manipulation. (2023-09-28, shares: 1.0)

VIX Flattening and Volatility Positions: Tipping Point?: The article suggests a possible shift in volatility positions due to the flattening of VIX at the curve's front end. (2023-09-28, shares: 1.0)

Improving Execution Quality in FX through Technology and Automation: The piece discusses the role of technology and automation in improving execution quality in forex trading. (2023-09-28, shares: 2.0)

Leveraging GenAI Models for Competitive Edge in Financial Services: OReilly Media's Generative AI for Finance event will highlight how financial firms use generative AI models to boost productivity and competitiveness. (2023-09-28, shares: 1.0)

Podcasts

Quantitative

Machine Learning in Quant Finance Conference Part 02: ML in Quant Finance Conf Pt 02: The second part of the Machine Learning in Quant Finance Conference is available to watch, featuring discussions on the application of machine learning in quantitative finance. (2023-09-22, shares: 12)

Machine Learning in Quant Finance Conference Part 01: ML in Quant Finance Conf Pt 01: The first part of the Machine Learning in Quant Finance Conference is available for viewing, focusing on the integration of machine learning in the financial sector. (2023-09-22, shares: 12)

Twitter

Quantitative

Returns in Bear vs. Bull Markets: The article suggests that equity factors and raw returns usually perform better in bear markets due to the slow correction of mispricing. (2023-09-25, shares: 4)

Data Sources Across Disciplines: The new paper offers a detailed list of potential data sources across various fields including finance, medicine, retail, etc. (2023-09-24, shares: 2)

DeFi Exchanges Liquidity Modeling: The article explores the process of modeling liquidity in decentralized finance (DeFi) exchanges. (2023-09-23, shares: 2)

Quant Signals: The article delves into the subject of rough quant signals. (2023-09-28, shares: 1)

Miscellaneous

Low-risk anomaly in bonds: The article analyzes the existence of a low-risk anomaly in the corporate bond market. (2023-09-27, shares: 1)

Technology enables new creation modes: The article investigates the role of technology in fostering new creative processes. (2023-09-28, shares: 0)

Long-term pricing effect on short-term dynamics: The article explores the use of long-term pricing effects in understanding short-term price fluctuations. (2023-09-26, shares: 0)

AI's impact on knowledge workers: The article shares HBR's research on how AI is affecting knowledge-based professions. (2023-09-26, shares: 0)

Capacity constraints in corporate bond strategies: The article delves into the increasing interest in systematic corporate bond strategies, their limitations, and the effects of high-turnover strategies. (2023-09-25, shares: 0)

Blogs

Related

Free Strategies for Traders: Quantifiedstrategies.com is providing 200 complimentary trading strategies for active traders. (2023-09-25, shares: 2)

Seasonal Trends in India VIX: Seasonal patterns are observed in the India VIX index. (2023-09-24, shares: 1)

Reddit

Quantitative

Junior Quant Researcher Switching Jobs: The author shares their diverse role experiences in a systematic trading startup due to the absence of specialization. (2023-09-23, shares: 14.0)

Translating Methods: Non-Quant to Quant: explores the analogy between alpha decay in physics and its use in investment strategies. (2023-09-25, shares: 1.0)

Weekly Megathread: Education, Career, and Hiring Tips: The post invites new and aspiring quants to join weekly megathreads to consolidate information and minimize redundancy. (2023-09-25, shares: 5.0)

Paper with Code

Trending

Diffusion UNet: Generation Quality Unleashed: The paper explores the untapped potential of diffusion UNet in enhancing generation quality. (2023-09-27, shares: 633.0)

Efficient Finetuning of Large Language Models: LongLoRA uses LLaMA2 7B or LLaMA2 70B to broaden context on a single 8x A100 machine. (2023-09-23, shares: 353.0)

Logical Deduction Failure in LLMs: The article explores a hypothesized failure in logical deduction caused by the Reversal Curse. (2023-09-24, shares: 86.0)

Rising

RunpeiDong DreamLLM: Multimodal Comprehension and Creation: The article presents DreamLLM, a learning framework that improves Multimodal Large Language Models by combining multimodal comprehension and creation. (2023-09-22, shares: 65.0)

Multimodal Comprehension and Creation: The article explains how COCO DEViT outperforms the current state-of-the-art open-vocabulary by 6.9 AP50 and reaches 50 AP50 in new categories. (2023-09-27, shares: 48.0)