Quant LetterNo. 40
95 items across 9 sections, as sent to readers on 13 March 2024. Paper titles open their ML-Quant page; ↗ goes to the source.
Quantitative-finance and ML-for-finance preprints from arXiv.
5 items
A new model for large equity markets over long periods of time has been introduced, using rank volatility stabilized models, which aligns well with empirical data and allows for relative arbitrage.
7 shares4 citations todaySource ↗
A new framework for pricing debt securities under varying short-rate diffusion processes has been proposed, including an efficient algorithm for pricing callable/putable debts and convertible bonds, and aligns perfectly with the current market interest rates.
4 sharesSource ↗
A new concept of GG-convex conjugate, a geometrically convex function suitable for financial risk measures, has been introduced, with a general dual representation for GG-convex risk measures derived, using Orlicz risk measures as a key example.
2 shares1 citation todaySource ↗
The first article investigates how companies' algorithm-based pricing strategies can result in collusion and price discrimination, negatively impacting consumer surplus and societal welfare.
2 shares3 citations todaySource ↗
Credit Expansion and Business Cycles: The study reveals that the 1999-2010 U.S. business cycle was primarily driven by credit expansion in private-label mortgages, leading to a significant boom and bust in housing-related industries. This contradicts several previous theories about the cycle's cause.
2 sharesSource ↗
Working papers in finance and economics from SSRN.
31 items
The paper introduces a new method for deducing covariance functions from sparse data using Convolutional Neural Networks, offering high accuracy and low computational time.
3 sharesSource ↗
The paper finds that Graph Attention Networks yield better investment outcomes compared to traditional complex system-based optimization models.
5 sharesSource ↗
The review examines 69 studies on machine learning's role in detecting code smell (error detection), emphasizing the importance of data preprocessing techniques.
2 sharesSource ↗
The paper explores numerical solutions for the Black-Scholes Partial Differential Equation in options pricing, using specific approximation methods.
2 sharesSource ↗
The study presents FinText, a tool that enhances volatility forecasts using business news archives and identifies key phrases impacting volatility.
2 sharesSource ↗
The paper uncovers a V curve pattern in the Shanghai stock market, showing a stronger dependency of volatility on negative return.
2 sharesSource ↗
The study finds a correlation between cryptocurrency market volatility and trading spread, suggesting portfolios based on the Johansen procedure yield the highest returns.
4 shares1 citation todaySource ↗
The article discusses the challenge of rapidly analyzing and using vast social media data during natural disasters for effective crisis management.
2 sharesSource ↗
The study reveals that high-volume closing auctions distort the closing price and reduce intraday liquidity in major European markets.
2 sharesSource ↗
Swing pricing can mitigate liquidity risk in open-ended funds during market stress, but its effectiveness may be limited by factors like increased volatility, leverage, and non-disclosure practices.
2 sharesSource ↗
Global credit supply shocks significantly impact country-level financial outcomes, emphasizing the need for sector-specific credit market monitoring.
2 sharesSource ↗
Mutual funds, especially ESG funds, recall loaned shares before voting record dates, with recalled shares more likely to vote against management proposals, attracting higher fund flows without poor performance due to lost lending revenues.
2 sharesSource ↗
The study reveals that firms are negatively impacted by court decisions that undermine trademark redeployability, leading to a decrease in book leverage.
2 shares1 citation todaySource ↗
The article disputes common models of flight-to-liquidity shocks, arguing that the post-2010 fall in the natural rate was due to a slowdown in productivity growth, not liquidity shocks.
2 sharesSource ↗
Opportunities: Big data and AI are transforming finance and accounting sectors by altering data processing and decision-making, with increased use of machine learning and AI analytics for empirical evidence analysis.
2 sharesSource ↗
The paper identifies a factor linked to climate news indices and commodity markets that contributes to the excess volatility of climate-sensitive assets.
3 shares1 citation todaySource ↗
The study suggests that financial institutions can reduce climate risk through securitization, particularly with diversified Mortgage-Backed Securities that have lower spatial concentration and correlation in wildfire events.
31 sharesSource ↗
The research finds the Welch 2022 age-decayed slope-winsorized beta estimator to be the most effective among 11 market beta estimators in forecasting and hedging performance across international stock markets.
7 shares1 citation todaySource ↗
The study introduces a method to simulate asset price and variance under the Hull and White stochastic volatility model, offering a way to generate unbiased estimates for the price of derivatives instruments.
3 sharesSource ↗
A new stock market strategy that uses a financial stress indicator and sentiment analysis from ChatGPT reading Bloomberg daily market summaries has improved performance in major equity markets.
2 sharesSource ↗
A new predictor, MCOR, uses a moderate cross-section of currency returns and covariates correlation to robustly predict carry trade returns, introducing a new currency investment strategy called MaC strategy.
3 sharesSource ↗
A unified framework reveals that most macroeconomic factors have increasing unconditional term structures with large risk premia at business cycle horizons.
8 shares3 citations todaySource ↗
Machine learning, particularly Random Forest, shows strong evidence of return predictability and investment gains for individual corporate bonds, especially in private bonds.
3 sharesSource ↗
In the Markowitz market, market clearing is the same as the market portfolio's clearing, and the CAPM equation is equivalent to the semi-clearing condition, which determines the composition value percentage of the market portfolio.
3 sharesSource ↗
Corporate bond credit spreads are affected by default risk and convenience services, with the European Central Bank's corporate quantitative easing programs significantly impacting corporate bonds' convenience yields.
2 sharesSource ↗
A new method for approximating the beta coefficient in investment projects is proposed, using a simulation model of expected returns to determine the project's market beta when the risk-return relationship is unobservable.
2 sharesSource ↗
The 'overnight-intraday return gap' in stock markets is largely due to retail investors' trading intensity, who tend to trade more actively in low per-share price and high volatility stocks.
7 shares1 citation todaySource ↗
The paper suggests three linked solutions to the absence of a comprehensive asset pricing theory: state-space representation, explanatory gain decomposition approach, and shifting research focus from asset returns to asset prices.
2 sharesSource ↗
In Pursuit of the Perfect Portfolio discusses the work of finance scholars in portfolio management and the potential for a perfect risk-reward balance.
2 sharesSource ↗
Research shows that laws requiring firms to disclose data breaches increase future stock price crash risk, especially for firms with weak governance and high financial constraints.
2 sharesSource ↗
A paper improves Option-based portfolio insurance (OBPI) to provide better portfolio protection over a rolling window.
2 sharesSource ↗
Economics working papers from RePEc's NEP field reports.
14 items
A study of 21 equity indices in 17 developed economies found that global variables are the main determinants of local volatility persistence.
24 sharesSource ↗
A non-Gaussian dynamic currency hedging strategy for globally diversified investors is stable, robust, and highly risk reductive, outperforming other hedging approaches.
22 sharesSource ↗
Short-term trading increases price volatility in crude oil futures markets, with different futures markets susceptible to same fall but different rise scenarios, according to a study.
19 sharesSource ↗
Only investors who can split the bid-ask spread can profitably exploit pricing errors in the Euro Short Term Rate Overnight Index Swaps market, according to a study.
16 sharesSource ↗
The study shows that NFTs have a minor link with traditional currencies but significantly impact returns and volatility during the Covid-19 crisis.
16 sharesSource ↗
The study reveals that volatility in financial markets and global macroeconomic variables impact soybeans' convenience yield, supporting the notion of commodity financialization.
12 sharesSource ↗
The study shows that insider trading strategies, specifically buying insider purchases and selling insider sales, yield higher profits over longer holding periods.
12 sharesSource ↗
The study cautions that data-mining bias can skew forecasts and that the contribution of macro-variables in forecasting total variance by GARCH-MIDAS models is exaggerated.
12 sharesSource ↗
The research suggests AI tokens can diversify portfolios under normal market conditions and provide low-cost hedging for traditional assets, except oil and cryptocurrency, but are sensitive to extreme shocks.
16 sharesSource ↗
P/E ratios: The study uses machine learning to accurately predict corporate earnings, outperforming other methods.
20 sharesSource ↗
Futures vs. Cash Index: Research shows that the CSI 300 index futures typically precede the cash index by 0-5 minutes, but this lead time can vary depending on market conditions.
21 sharesSource ↗
The study finds that smaller market capitalization, higher reversal, lower liquidity, and lower maturity in cryptocurrencies tend to yield higher returns.
12 sharesSource ↗
The research reveals that investor sentiment, risk-free interest rates, and broad market indices significantly affect IPO first-day flipping in China.
12 sharesSource ↗
The research introduces new tests for risk premia in linear factor models, offering more reliable results than traditional methods.
11 sharesSource ↗
The general machine-learning papers the letter carried in 2023-25.
5 items
A new variance-preserving aggregation function for Graph Neural Networks is proposed, which could lead to more efficient and self-normalizing networks.
77 shares7 citations todaySource ↗
A novel method for batch Bayesian Optimization (BO) is introduced, which reduces redundancy and focuses on points with high predictive means or uncertainty, showing superior performance on nonconvex test functions.
13 shares3 citations todaySource ↗
Graph Machine Learning via Bilevel Optimization: The paper views graph learning techniques as special cases of bilevel optimization, introducing a new class of energy functions for graph neural network layers, and showcasing the versatility of this approach through empirical results.
12 shares3 citations todaySource ↗
The study reveals that deep neural networks using linear complex-valued RNNs and MLPs can accurately approximate regular causal sequence-to-sequence maps, with complex eigenvalues near unit disk aiding in information storage.
91 shares41 citations todaySource ↗
A paper reveals that a model-based approach can achieve optimal sample complexity without burn-in cost in offline reinforcement learning for tabular Markov decision processes, providing an efficient solution for sample-starved applications.
31 shares113 citations todaySource ↗
Papers that shipped their code, from the Papers with Code feed (2023-25).
3 items
Named Entity Recognition (NER) is identified as a key component in various Natural Language Processing (NLP) applications.
98 shares
AgentLite is a new open-source AI library that makes it easier to develop and apply LLM agent reasoning architectures.
96 shares
Current foundation agents have difficulty adapting to different scenarios due to changes in observations and actions.
78 shares
Repositories the letter featured.
7 items
Stock Indicators for Python is a tool designed to analyze the stock market.
144 shares
Postgres Vector Search: The article provides a simple method for managing vector search on Postgres.
461 shares
Non-Transformer Time Series Models: The repository offers insights into the reasons behind the failure of transformers in time series forecasting and emphasizes the most effective non-transformer models.
232 shares
Labeling Text Data: This tool simplifies the process of embedding and semantically labeling text datasets.
109 shares
Hackable LLM: A customizable version of the latest open-source LLMs, based on nanoGPT, is accessible under the Apache 2.0 license.
5,205 shares
AI Security Risk Tool: PyRIT is an open-source software that aids in identifying potential risks in generative AI systems used by security professionals and machine learning engineers.
1,093 shares
LLM Data Indexing: LLM is a tool that enables the indexing and querying of any data using natural language, and it also displays sources and citations.
1,012 shares
Industry news: funds, hiring, markets and regulation.
5 items
A quantitative hedge fund is starting 2024 by significantly increasing its staff.
4 shares
Man Group has improved efficiency for quantitative analysts with its proprietary data platform.
4 shares
London-based hedge fund firm, Qube Research & Technologies, has built the largest ever short position against Barclays, representing 0.73% of the bank's stock.
3 shares
Hong Kong-based hedge fund firm Infini Capital Management has received preliminary approval to open an office in the Abu Dhabi Global Markets.
3 shares
Millenium Quant Trader and Fordham Speaker, Paul Krueger, is highlighted in Rebellion Research.
2 shares
Episodes on markets, quant methods and economics.
6 items
Philosophy in Market Mastery: Mark Ritchie II highlights the need for understanding market trends, mastering trading strategies, and risk management in fluctuating markets.
10 shares
U.S. Financial History: Mark Higgins' Investing in U.S. Financial History offers a detailed financial history of the US, providing insights into managing financial crises, identifying asset bubbles, and the advantages of index funds.
8 shares
Active Management and ETFs: Jack Shannon emphasizes the significance of downside protection, active equity management, and knowledge of market history for long-term investment success.
16 shares
David Lebovitz and Michael Cembalest discuss the role of private market allocations in portfolio diversification and growth, based on Cembalest's review of alternative investments.
4 shares
Will McBride and Dmitry Pargamanik of Market Chameleon emphasize the significance of data in trading in a podcast following their webinar on trade data.
4 shares
The podcast humorously discusses data science trends, AI creativity, and Hollywood movies with Kaggle Grand Master Konrad Banachewicz, sponsored by Intrepid AI and Arctic Wolf Labs.
4 shares
Posts from quant researchers on X.
19 items
The article recaps recent research on asset pricing, machine learning, and volatility.
8 shares
The article discusses a repository based on the book Machine Learning in Finance From Theory to Practice.
5 shares
The article presents a multiple-testing framework for predictive signals, outperforming existing frameworks in FX trading.
3 shares
The article features an interview with Clifford Asness on the use of price momentum and AI in natural language processing.
2 shares
The article shares data on the fluctuation of various portfolios, observed monthly from 1929 to 2023.
1 shares
The article explores a republished paper on diversified multi-asset strategies throughout different economic cycles.
1 shares
The article introduces a new paper that predicts the likelihood of order fulfillment at the best quotes in a limit order book.
1 shares
The article examines Dobromir Tzotchev's 2018 paper on market-neutral carry trades across various asset classes.
1 shares
The article presents a dataset of corporate bond factors from 1994 to 2023, based on a paper by Houweling and van Zundert.
1 shares
Research indicates that unexpected information in earnings call reports, rather than the earnings themselves, cause a more significant post-announcement price drift.
0 shares
A recent paper highlights the potential problems of information leakage and lookahead bias in large language models.
0 shares
According to ManGroup, investors are prone to making poor decisions when experiencing high levels of stress or pain.
0 shares
Zhang's research validates the existence of support and resistance levels in stock prices, suggesting stocks closing just above round numbers tend to rise, while those just below tend to fall.
0 shares
The article celebrates Jean Baptiste Joseph Fourier's invention of Fourier series in the 18th century and explores its current use in Fast Fourier Transforms (FFTs).
0 shares
Richard W. Evans offers tutorials and Python code for popular computational methods in economics.
0 shares
The article debates the ability of structured products to control the Volatility Index (Vix).
0 shares
The article examines the effect of order clustering around whole dollar levels on stock price trends.
0 shares
Cliff Asness discusses efficient markets, momentum fundamentals, and AI in natural language processing, highlighting the trend of savvy investors moving towards passive investing.
0 shares
NYU Tandon's Department of Finance and Risk Engineering hosted a lecture by Xin Zhang from the University of Vienna on the Wasserstein Space of Stochastic Processes.
0 shares