A methodology for analyzing financial needs hierarchy from social discussions using LLM

TL;DR

This paper develops a data-driven framework to test whether financial needs expressed in social media follow a hierarchical structure consistent with Maslow-based NHF and a three-tier NPF model. Using Reddit posts from 2020–2023, the authors extract age, income, financial needs, emotions, stress, risk propensity, and topics with LLMs and topic modeling, then map these needs to both frameworks. They find general support for a hierarchical progression from basic to long-term financial goals, with clear alignment of topics and needs across NHF and NPF, and show how income relates to higher-tier needs. The work demonstrates a scalable, data-driven approach to understanding real-world financial behavior, offering insights for advisors, regulators, and researchers while outlining avenues for temporal and geographic expansion.

Abstract

This study examines the hierarchical structure of financial needs as articulated in social media discourse, employing generative AI techniques to analyze large-scale textual data. While human needs encompass a broad spectrum from fundamental survival to psychological fulfillment financial needs are particularly critical, influencing both individual well-being and day-to-day decision-making. Our research advances the understanding of financial behavior by utilizing large language models (LLMs) to extract and analyze expressions of financial needs from social media posts. We hypothesize that financial needs are organized hierarchically, progressing from short-term essentials to long-term aspirations, consistent with theoretical frameworks established in the behavioral sciences. Through computational analysis, we demonstrate the feasibility of identifying these needs and validate the presence of a hierarchical structure within them. In addition to confirming this structure, our findings provide novel insights into the content and themes of financial discussions online. By inferring underlying needs from naturally occurring language, this approach offers a scalable and data-driven alternative to conventional survey methodologies, enabling a more dynamic and nuanced understanding of financial behavior in real-world contexts.

Figures (5)

• Figure 1: Framework for extraction of various features from user text data
• Figure 2: Interconnections Between Financial Topics and Hierarchical Needs Framework
• Figure 3: Mapping the Relationship Between Needs Hierarchy and Needs Prioritization
• Figure 4: Co-occurrence relationships among financial need topics
• Figure 5: Co-occurrence of Need labels from two Frameworks