|
|
|
AUTHOR(S): Metaxoudis Eleftherios
|
|
TITLE |
 PDF |
|
ABSTRACT Financial Foundation Models (FFMs) have recently emerged as powerful architectures for forecasting, risk assessment, and decision support in complex financial environments. Despite their empirical success, existing research predominantly emphasizes performance optimization, explainability, and data efficiency, while largely overlooking a foundational question: how FFMs internally structure decision timing, priority allocation, and delay under uncertainty. This paper addresses this gap by proposing a computational-philosophical re-theorization of FFMs, conceptualizing them not merely as predictive systems but as decision-making architectures governed by implicit temporal and hierarchical logics. We introduce a structured framework grounded in six core computational concepts: procrastination (decision deferral), anti-procrastination (forced activation), metric (value encoding beyond loss functions), inverse priority (dominance of rare events), priority inheritance (temporal propagation of salience), and the trade-off between responsiveness and throughput. These concepts are analytically mapped onto FFM architectures to reveal how priorities emerge, shift, and propagate across decision layers, particularly under stress conditions. To operationalize the framework, we employ financial Digital Twins as experimental substrates for stress testing and scenario simulation. Illustrative simulations and statistical comparisons are used to demonstrate how priority-aware and delay-sensitive FFMs exhibit qualitatively different decision behaviours compared to conventional designs. The proposed approach contributes a novel theoretical lens for understanding decision latency and priority dynamics in financial AI systems, with implications for model architecture, risk management, and regulatory stress testing. By bridging Computational Philosophy and Financial Engineering, this study advances a foundational perspective on how financial AI systems decide -not only what to predict, but when and under which priorities to act. |
|
KEYWORDS Financial Foundation Models; Computational Philosophy; Decision Theory; Priority Inversion; Digital Twins; Stress Testing |
|
|
|
Cite this paper Metaxoudis Eleftherios. (2026) Decision, Priority, and Delay: A Computational-Philosophical Re-theorization of Financial Foundation Models. International Journal of Economics and Management Systems, 11, 42-62 |
|
|
