Optimal Hedging
The study investigates the best way to hedge risk using derivatives in incomplete markets, focusing on an investor exposed to two assets and using vanilla options as hedging tools.
8 shares2 citations todaySource ↗
Quant LetterNo. 39
March 2024, Week 1
98 items across 7 sections, as sent to readers on 6 March 2024. Paper titles open their ML-Quant page; ↗ goes to the source.
- Items
- 98
- Papers
- 67
- Sections
- 7
- Issue
- No. 39
arXiv
Quantitative-finance and ML-for-finance preprints from arXiv.
18 items
Finance8
Insider Trading Detection
The article introduces an unsupervised machine learning technique for detecting potential insider trading by analyzing large datasets, using principal component analysis and autoencoders.
7 shares2 citations todaySource ↗
Volatility Strategy
The research looks at the growth of derivative markets in China, focusing on a short-volatility strategy using ETF options data, and suggests model improvements based on volatility forecasts.
7 sharesSource ↗
Deep Learning for Pricing
The study introduces a new deep learning method for pricing European options in diffusion models, transforming the option pricing equation into an energy minimization problem and using deep artificial neural networks.
6 shares4 citations todaySource ↗
Digitwashing
The discrepancy between a company's digital transformation promises and actual results can lead to a stock price crash, worsened by economic policy uncertainty and unprofitable firms.
4 shares2 citations todaySource ↗
Fourier Pricing of Multi-Asset Options
The RQMC quadrature enhances the scalability of Fourier methods in pricing multi-asset options, surpassing traditional methods and offering practical error estimates.
4 shares3 citations todaySource ↗
Properties of EVaR
The Lambert function has been used to successfully calculate the Entropic Value-at-Risk (EVaR) measure for various distributions like Poisson, Gamma, and Laplace.
3 shares5 citations todaySource ↗
Fill Probabilities in Order Book
A new stochastic model accurately calculates fill probabilities for limit orders at different price levels in the order book, effectively capturing its dynamics.
2 shares3 citations todaySource ↗
Miscellaneous5
Robust Utility Valuation
The article presents a new method for maximizing utility with semistatic strategies for exotic options, introducing a robust form of convex integral functionals and establishing key results, which provide a solution for the robust utility maximization problem and a representation of associated indifference prices.
9 sharesSource ↗
MambaStock Prediction
The paper presents MambaStock, a new Mamba-based model for predicting stock prices using historical market data, which outperforms previous methods in accuracy, aiding investors in making informed decisions.
7 shares31 citations todaySource ↗
Bandit Profit-maximization
The study explores a sequential profit-maximization problem, optimizing price and marketing expenditures across multiple markets with different demand curves, and introduces near-optimal algorithms for this problem in an adversarial bandit setting, proving an upper and lower regret bound for monotonic demand curves.
6 sharesSource ↗
Transformer for Time Series
The research investigates the use of transformer models in financial time series prediction, showing promising results with synthetic data and insightful findings on S&P500 data volatility prediction.
3 shares4 citations todaySource ↗
Historical Trending5
Two-Way-Fixed-Effects Regression for Panel Data
The article introduces a new estimator for calculating average causal effects of binary treatment with panel data, offering better performance and robustness than traditional two-way estimators.
361 sharesSource ↗
Market Microstructure and Pricing
The paper presents a discrete binary tree for pricing contingent claims, which is arbitrage-free, market-complete, and maintains all parameters controlling the historical price dynamics.
57 sharesSource ↗
Local Volatility in Rate Models
The article explains the implementation of Local Volatility in market modeling to replicate most swaption prices within a single model, but short-term swaption volatility cannot be accurately generated due to the use of a normal distribution.
37 sharesSource ↗
Gamma Hedging
The research uses rough path theory to show that a specific hedging strategy can replicate other European options, even without a specific pricing model.
32 shares2 citations todaySource ↗
Talent Hoarding
The study reveals that talent hoarding by managers in companies discourages employees from seeking new roles, affecting career growth and talent distribution within the organization.
30 shares12 citations todaySource ↗
SSRN
Working papers in finance and economics from SSRN.
27 items
Quantitative9
Inflation Model
The article presents a new model for predicting inflation volatility, claiming superior performance over traditional methods.
20 sharesSource ↗
Intraday Volatility Prediction
The paper proposes a new method for predicting intraday volatility in financial data using Ito semimartingale models and a Two-side Projected-PCA procedure.
2 sharesSource ↗
Banking Stability Prediction
The research uses the CAMELS framework and machine learning to assess the performance of major banks in top GDP countries, with the aim of predicting future performance.
2 sharesSource ↗
Seeking Alpha: Investment Advice
Investment Advice: Investment advice from Seeking Alpha offers timely and relevant information for savvy investors, impacting immediate market returns and 90-day drift returns.
3 shares1 citation todaySource ↗
AI Risk Package
The authors introduce a Python package and metrics for managing risks in Artificial Intelligence applications, emphasizing their interpretability and reproducibility.
14 sharesSource ↗
Asset Pricing Frictions
The notes detail a Big Data Asset Pricing course, covering asset pricing basics, transaction costs, market liquidity risk, and machine learning.
2 sharesSource ↗
Barycentric ML Optimization
The study suggests a method to decrease machine learning algorithms' execution time in high-dimensional spaces using the barycentric correction procedure.
3 sharesSource ↗
Robust Inference for Financial Portfolios
The article highlights the importance of accurately modeling asset dependence in financial portfolios, emphasizing the significance of correlation-concordance matrices during market stress.
2 sharesSource ↗
Corporate Refinancing and Bond Returns
The paper reveals that the urgency of refinancing maturing debt can increase future corporate bond returns, particularly during periods of high default and liquidity risk.
2 sharesSource ↗
Financial18
Local Edgeworth
The article introduces a mathematical model to estimate changes in level-volatility in a Brownian semimartingale, incorporating skewness and kurtosis through fluctuating correlations and volatility changes.
3 shares1 citation todaySource ↗
Greenium Search
The study introduces a robust green score and expected returns to calculate the greenium, the expected return of green securities compared to brown, which is found to be more negative in greener countries and over time.
25 shares20 citations todaySource ↗
Commodity Pricing
The article shows that a latent-factor model using the Instrumented Principal Component Analysis methodology surpasses existing models in explaining variations in commodity futures returns, with momentum, expected shortfall, and idiosyncratic volatility as key factors.
3 shares1 citation todaySource ↗
Asset Ratio
The article proposes a strategy to improve weak medium-term returns in retirement portfolios by adjusting the stock percentage based on the earnings yield of stock and the current yield of bonds, with caution needed when stock prices exceed sustainable levels.
3 sharesSource ↗
Toxicity Trade-off
The paper investigates the liquidity provision game in decentralized exchanges, revealing a tradeoff between toxicity and competitiveness in liquidity provision and offering a new guideline for liquidity provision in the decentralized financial market.
6 sharesSource ↗
Insider Trading Detection
The article proposes a machine learning method for detecting potential insider trading by analyzing large datasets of trading positions.
2 sharesSource ↗
Limit Order Book Simulations Review
The piece reviews models of Limit Order Books simulations, emphasizing the role of AI in improving these models and the significance of price impacts in algorithmic trading.
2 shares19 citations todaySource ↗
Overcoming Markowitz's Instability
The paper shows the hierarchical risk parity (HRP) approach is superior to the traditional Markowitz portfolio allocation method in terms of noise reduction and robustness.
4 shares3 citations todaySource ↗
Dispelling Myths in Optimization
The article discusses the widespread use of mean-variance optimization in quantitative finance, dispels associated myths, and introduces the concept of mean-variance-equivalent distributions.
7 shares6 citations todaySource ↗
Global FX Ambiguity in Portfolios
The study examines the effect of global foreign exchange ambiguity on currency portfolios, finding that high ambiguity leads to high currency carry returns and uncovers uncertainty not captured by FX volatility.
3 shares2 citations todaySource ↗
Volatility Risk Premiums in Swaption Market
The study investigates how unconventional monetary policies and pandemics affect volatility risk premiums in the USD interest rate swaption market from 2007 to 2022.
118 sharesSource ↗
Mutual Fund Efficiency & Internationalization
The research finds that UK equity mutual funds with a domestic focus perform better than those with an international focus.
86 sharesSource ↗
Robust Stochastic Volatility Models
The paper suggests four principles to evaluate the suitability of a Stochastic Volatility model for valuing derivative securities across various asset classes.
2 sharesSource ↗
Procyclicality of Risk-based Initial Margin Models
The study presents a method to evaluate the responsiveness of initial margin calculation models during periods of high market volatility.
64 sharesSource ↗
Mutual Fund Outflows and First-Mover Advantage
The study suggests that mutual fund outflows after poor performance are due to a firstmover advantage in the asset market, not investor behavior, affecting mutual fund industry regulation and understanding.
4 sharesSource ↗
Cryptocurrency Factor Diversification
Research indicates that adding size and momentum-based cryptocurrency factors to a stock-bond portfolio can significantly diversify it, with machine-learning asset allocation strategies enhancing these benefits.
3 sharesSource ↗
Global Mutual Fund Flows Study
The research shows that investors' decisions are influenced by the performance of mutual funds, with variations based on the fund's size and market position.
2 sharesSource ↗
Volatility, Leverage, and Skewness in Stock Returns
The study finds that variance shocks strongly influence the conditional skewness of index returns, impacting asset pricing, portfolio selection, and risk management applications.
2 sharesSource ↗
RePEc
Economics working papers from RePEc's NEP field reports.
18 items
Finance4
Investment Strategies Development
A new scale for assessing short and long-term investment strategies was developed and proven reliable for understanding investment decision-making processes.
16 sharesSource ↗
Estr OIS Market Efficiency
The study finds that only investors skilled in navigating the bid-ask spread can profit from mispricing in Euro Short Term Rate Overnight Index Swaps.
16 sharesSource ↗
Mutual Funds Active Investors
The study reveals that individual investors in mutual funds act as momentum buyers and contrarian sellers, with older and larger transaction investors more likely to be momentum buyers.
14 sharesSource ↗
Credit Rating Announcements Trading Responses
The research finds that investors react differently to changes in credit ratings from issuer-paid and investor-paid agencies, and can earn significant abnormal returns by using information from both.
11 sharesSource ↗
Statistical6
AI Assets Connectedness with Traditional Classes
Research suggests AI tokens can diversify traditional assets under normal market conditions, but fail to do so during extreme market shocks.
16 sharesSource ↗
Spatial Autocorrelation in Hedonic Models
The study proposes a spatial cross-validation strategy to correct bias in error estimates of tree-based algorithms in real estate data due to spatial autocorrelation.
23 sharesSource ↗
Financial Machine Learning with R
The paper explores the limitations of machine learning in finance, offering advice on method selection and referencing R libraries for computation.
20 sharesSource ↗
AI for human learning
The paper presents a framework for integrating AI into education and proposes a learning design model for AI-based learning support systems.
12 sharesSource ↗
Machine learning in supply chain
The research reviews the use of machine learning in supply chain management, providing insights for future studies in this area.
12 sharesSource ↗
Stock Return Forecasting with Machine Learning
The article uses machine learning to predict stock returns, challenging the efficient market hypothesis due to its strong predictive power. It also shows that machine learning models are effective in out-of-sample performance.
17 sharesSource ↗
Historical Trending8
Adaptive Portfolio Selection with Transaction Costs
The paper introduces an adaptive moving average method with peer impact for online portfolio selection, which considers the influence of other risky assets for accurate return predictions, and an adaptive mean-variance model for risk measurement.
23 sharesSource ↗
Volatility, Growth Options, and Returns
The study reveals that growth firms and high idiosyncratic volatility firms outperform the CAPM during periods of high aggregate volatility, thus lowering their risk.
21 sharesSource ↗
Robust Testing of Risk Premia
The article introduces new tests for risk premia in linear factor models that are robust to small sample sizes and weak identification of risk premia, and revisits two empirical applications to show differences from traditional tests.
11 sharesSource ↗
Independent Directors and Financial Fraud
The research finds that companies with dissenting independent directors, identified through machine learning predictions and Chinese board voting data, have a lower future risk of financial fraud.
9 sharesSource ↗
Adaptive Portfolio Selection
The paper discusses the use of polynomial series, specifically Taylor and Bernstein series, to solve dynamic portfolio optimization problems.
9 sharesSource ↗
Volatility, Growth, and Returns
The research suggests using the characteristic function to estimate linear models with errors in financial econometrics, with applications to the capital asset pricing model.
8 sharesSource ↗
Robust Testing
A new test is introduced for identifying changes in risk exposures of large financial asset portfolios, revealing portfolio weight dynamics across different regimes.
8 sharesSource ↗
Independent Directors
The research uses micro-scale job-household data and machine learning to analyze spatiotemporal patterns in Tokyo, highlighting urbanization and suburbanization trends.
7 sharesSource ↗
Machine learning
The general machine-learning papers the letter carried in 2023-25.
4 items
Recently Published4
Active Inference for Stat Analysis
A new method called Active inference, which uses machine learning to collect data and focuses on areas where the model is uncertain, achieves the same accuracy with fewer samples than previous methods.
32 shares53 citations todaySource ↗
Behavior Generation with VQ-BeT
The Vector-Quantized Behavior Transformer (VQ-BeT), a new model for behavior generation, improves multimodal action prediction, conditional generation, and partial observations, and speeds up inference.
32 shares221 citations todaySource ↗
Preventing Reward Hacking
A novel method to prevent reward hacking in AI systems uses state occupancy measure instead of action distribution, effectively avoiding significant drops in true reward.
10 shares55 citations todaySource ↗
Variance in Fair Classification
Variance in predictions across different models is a major source of error in fair binary classification, and a new metric, self-consistency, is proposed to measure and reduce randomness, challenging the effectiveness of common algorithmic fairness methods.
65 shares51 citations todaySource ↗
GitHub
Repositories the letter featured.
7 items
Finance7
HSBC Order Book Project
University students and HSBC's AI team have developed a profitable trading strategy using machine learning algorithms on level2 limit order book data.
9 shares
Math Finance Analysis
The article explores the application of financial modeling and quantitative analysis in finance.
3 shares
ML USU Repository
The piece details a Github repository for a machine learning course managed by a professor.
28 shares
Real-time Data Pipelines
The article provides a guide on creating real-time feature pipelines using Python.
80 shares
DataDreamer Data Generation
The article explores the creation of synthetic data and training of align models using DataDreamer Prompt.
524 shares
Python Made Easy
The article offers tips for a smoother experience with Python programming.
10,260 shares
Smart Contract Empowerment
The piece provides guidance on creating reliable and efficient smart contracts.
57,633 shares
Podcasts
Episodes on markets, quant methods and economics.
3 items
Quantitative3
Mastering Quant Finance Interviews
Katherina DuongBernet provides tips for successful interviews in the quant finance sector, covering employer expectations, common questions, and preparation strategies.
19 shares
Specialty Finance and Data Investing
Barry Ritholtz of Bloomberg Radio interviews David Snyderman from Magnetar Capital LLC, discussing his career and role in the company.
13 shares
Spreadbites High Yield Conference Takeaways
Stephen Dulake and Samantha Azzarello discuss global credit market trends following J.P. Morgan’s 2024 High Yield & Leveraged Finance Conference.
8 shares
X / Twitter
Posts from quant researchers on X.
21 items
Quantitative8
Machine Learning for Stock Prediction
Machine learning algorithms have proven to be more effective than traditional models in predicting weekly stock performance.
8 shares
Weekly Research Topic Recap
A weekly summary of research on topics like ESG investing, Macro Machine Learning, Volatility, etc. has been released.
6 shares
Empirical Asset Pricing Review
A recent paper offers an extensive review of empirical asset pricing and machine learning studies.
5 shares
Multipletesting for FX Trading Signals
A new framework for testing predictive signals surpasses current frameworks when used on over 20,000 FX trading rules across 30 currencies.
3 shares
Quant Trading Books
The author suggests three beneficial books on quant trading, two of which were instrumental in their early hedge fund career.
3 shares
Private Equity
The article offers an in-depth analysis and discussion on the subject of private equity.
2 shares
International Finance Notes
The article provides detailed lecture notes covering a range of topics within international finance.
1 shares
Equity Risk Factor Paper
The article introduces a research paper focused on Equity Risk Factor Regimes.
1 shares
Miscellaneous8
Stock Forecasting Comparison
The article explores the prediction of a stock's performance over the market, noting that simple models can be as effective as complex ones.
0 shares
Weirdness Prompting AIs
The article delves into the complexities and unpredictability associated with prompting artificial intelligence.
0 shares
PyRIT AI Red Team Tool
Microsoft employs PyRIT, a Python-based tool, for risk identification in generative AI.
0 shares
Highyield spreads and momentum factor performance
Momentum factor performance is strongly predicted by high-yield spreads and often crashes after significant drawdowns in market reversals.
0 shares
Current regime compared to late 90s boom
The current economic climate closely resembles the late 90s boom, particularly June 1997, suggesting potential for further growth.
0 shares
Microsoft's AI Red Team and PyRIT tool
Microsoft's AI Red Team utilizes the PyRIT Python Risk Identification Tool to manage AI risks and ensure compliance.
0 shares
Modality Aware Transformers for TimeSeries on FRED Data
Modality Aware Transformers (MAT) are employed for TimeSeries on FRED Data, with corresponding R code used to measure sentiment from texts.
0 shares
Related5
High Frequency Arbitrage
The article explains how high-frequency statistical arbitrage uses advanced tech and models to take advantage of brief market inefficiencies.
14 shares
Stop Loss Strategy Pros And Cons
The piece highlights how hedge funds have led the way in using high-frequency statistical arbitrage to profit from tiny, fleeting price differences in various assets.
14 shares
Responsible Investing Path
The piece discusses how changes in investors' expectations of central bank actions have recently affected financial markets.
4 shares
High Frequency Arbitrage
The article explores the inclusion of a frequently ignored asset class into a sustainable multi-asset portfolio.
3 shares
Benchmark for RL
The article criticizes some research tools for being either too slow without significant computational resources or not challenging enough.
60 shares