86 papers
This collection explores the cutting edge of Q-Fin — Gn, where advanced quantum computing principles intersect with financial modeling and game theory. These emerging studies investigate how quantum algorithms can solve complex economic problems and optimize strategic interactions far beyond the reach of classical computers, offering a glimpse into a future where financial markets operate with unprecedented speed and precision.
Every new preprint in this category originates from arXiv, the premier repository for physics and computer science research. At Gist.Science, we process each submission to provide both accessible plain-language explanations and detailed technical summaries, ensuring these breakthroughs are understandable to everyone from industry experts to curious students. Below are the latest papers in this rapidly evolving field, curated to keep you ahead of the curve.
This paper systematically surveys the rapidly growing private credit market by integrating theoretical frameworks on delegated monitoring and soft information with empirical evidence showing that while direct lenders earn a persistent spread premium over syndicated loans, risk-adjusted returns for average funds are largely offset by fees.
The paper introduces DatedGPT, a family of twelve 1.3B-parameter language models trained on temporally partitioned data with strict annual cutoffs from 2013 to 2024 to eliminate lookahead bias in financial forecasting while maintaining competitive performance.
This paper demonstrates that common shocks drive the tails of firm-level idiosyncratic return distributions, revealing a significant annual premium of 7–8% for stocks with high exposure to innovations in the lower-tail quantile factor, a risk that is amplified during periods of weak intermediary capital and low liquidity while also predicting aggregate market excess returns.
Using proprietary transaction-level data on gilt-backed repo trades, this paper demonstrates that dealer market power and the transmission of heterogeneous shocks within funding networks are significant drivers of bond mispricing, collectively accounting for up to 4 percentage points of yield deviation.
This paper extends Kyle's (1985) seminal insider trading model to an infinite-dimensional, many-asset setting, deriving a parsimonious equilibrium with closed-form solutions for trading strategies, cross-market price impacts, and information efficiency.
This paper presents a stochastic macro-financial model integrating the Keen framework with a jump-diffusion asset market to formalize the feedback loops between credit expansion and crash risk, demonstrating how endogenous financial fragility can generate regimes ranging from stable growth to recurrent boom-bust cycles.
This paper establishes a structural equilibrium framework for decentralized exchanges, demonstrating that liquidity providers strategically adjust market depth to manage risk through a trade-off involving risk aversion, replication costs, and private information, which collectively determine the economic viability and profitability of liquidity provision.
This paper investigates how to optimize input context for large language models in mention markets by introducing Market-Conditioned Prompting (MCP) and its mixture variant (MixMCP), demonstrating that richer contextual information and treating market probabilities as priors significantly improve the accuracy and calibration of keyword-mention forecasts.
This study introduces the Autonomous Disclosure Regulator, an AI-driven framework that reveals how companies exploit a "Strategic Gap" of confusing language and unpredictable timing to delay market truth by 60%, demonstrating that shifting from passive data repositories to active, real-time auditing agents is essential to restore market integrity and recover significant welfare losses.
This paper demonstrates a transparent and lightweight method to "brain scan" large language models, enabling researchers to identify, quantify, and steer the specific economic concepts and biases guiding their reasoning without compromising performance.