FinPersona-Bench: A Benchmark for Longitudinal Psychometric Stability of Autonomous Financial Agents

2026-06-30

AI Summary

This study addresses the problem of large language models (LLMs) deviating from their initial risk preferences in long-horizon financial decision-making due to instruction salience decay (MSD). The authors formally define MSD for the first time and introduce FinPersona-Bench, a simulation benchmark that decouples market prices from fundamental values to evaluate agent behavioral stability across calm, crash, and bubble market regimes. Proposing a falsifiable longitudinal behavioral evaluation framework combined with periodic instruction re-anchoring, experiments across 18 state-of-the-art LLMs reveal that MSD exhibits temporal accumulation and strong model dependency. Notably, during late-stage market crashes, behavioral divergence between re-anchored and non-re-anchored agents amplifies by 4.4×. While re-anchoring effectively stabilizes conservative agents, it may impair the performance of aggressive agents in low-signal market environments.

Abstract

Large Language Models (LLMs) are increasingly deployed as autonomous financial agents initialized with explicit behavioral mandates such as "preserve capital" or "avoid speculative bets" that are meant to govern every decision throughout deployment. In practice, however, as market context accumulates over long horizons, these mandates gradually lose their behavioral influence, a phenomenon we formalize as Mandate Salience Decay (MSD). To measure MSD objectively, we introduce FinPersona-Bench, a simulation benchmark in which a synthetic market decouples observable price from hidden fundamental value, enabling falsifiable evaluation across three failure modes: trading without signal in calm markets, panic-selling during crashes, and ignoring fundamental value during speculative bubbles. Evaluating 18 leading frontier and open-source LLMs, each assigned one of three behavioral profiles ranging from strict capital preservation to aggressive growth, shows that MSD compounds over time and is model-dependent. In crash scenarios, the behavioral gap between static agents and those receiving periodic mandate re-grounding grows 4.4x from the first to the final quarter of the simulation. The effects of mandate re-grounding are not uniformly positive: it consistently helps conservative agents in low-signal markets but actively worsens behavior for aggressive agents in the same setting. These findings suggest that reliable long-horizon deployment requires selective, mandate-aware re-grounding based on agent profile and market regime.

Problem

• Mandate Salience Decay
• autonomous financial agents
• behavioral stability
• longitudinal evaluation
• LLM alignment

Innovation

• Mandate Salience Decay
• FinPersona-Bench
• autonomous financial agents
• behavioral stability
• LLM re-grounding